Von BI zu PI: Der nächste Schritt auf dem Weg zu datengetriebenen Entscheidungen

„Alles ist stetig und fortlaufend im Wandel.“ „Das Tempo der Veränderungen nimmt zu.“ „Die Welt wird immer komplexer und Unternehmen müssen Schritt halten.“ Unternehmen jeder Art und Größe haben diese Sätze schon oft gehört – vielleicht zu oft! Und dennoch ist es für den Erfolg eines Unternehmens von entscheidender Bedeutung, sich den Veränderungen anzupassen.


Read this article in English: 
“From BI to PI: The Next Step in the Evolution of Data-Driven Decisions”


Sie müssen die zugrunde liegenden organisatorischen Bausteine verstehen, um sicherzustellen, dass die von Ihnen getroffenen Entscheidungen sich auch in die richtige Richtung entwickeln. Es geht sozusagen um die DNA Ihres Unternehmens: die Geschäftsprozesse, auf denen Ihre Arbeitsweise basiert, und die alles zu einer harmonischen Einheit miteinander verbinden. Zu verstehen, wie diese Prozesse verlaufen und an welcher Stelle es Verbesserungsmöglichkeiten gibt, kann den Unterschied zwischen Erfolg und Misserfolg ausmachen.

Unternehmen, die ihren Fokus auf Wachstum gesetzt haben, haben dies bereits erkannt. In der Vergangenheit wurde Business Intelligence als die Lösung für diese Herausforderung betrachtet. In jüngerer Zeit sehen sich zukunftsorientierte Unternehmen damit konfrontiert, Lösungen zu überwachen, die mit dem heutigen Tempo der Veränderungen Schritt halten können. Gleichzeitig erkennen diese Unternehmen, dass die zunehmende Komplexität der Geschäftsprozesse dazu führt, dass herkömmliche Methoden nicht mehr ausreichen.

Anpassung an ein sich änderndes Umfeld? Die Herausforderungen von BI

Business Intelligence ist nicht notwendigerweise überholt oder unnötig. In einer schnelllebigen und sich ständig verändernden Welt stehen die BI-Tools und -Lösungen jedoch vor einer Reihe von Herausforderungen. Hierzu können zählen:

  • Hohe Datenlatenz – Die Datenlatenz gibt an, wie lange ein Benutzer benötigt, um Daten beispielsweise über ein Business-Intelligence-Dashboard abzurufen. In vielen Fällen kann dies mehr als 24 Stunden dauern. Ein geschäftskritischer Zeitraum, da Unternehmen Geschäftschancen für sich nutzen möchten, die möglicherweise ein begrenztes Zeitfenster haben.
  • Unvollständige Datensätze – Business Intelligence verfolgt einen breiten Ansatz, sodass Prüfungen möglicherweise zwar umfassend, aber nicht tief greifend sind. Dies erhöht die Wahrscheinlichkeit, dass Daten übersehen werden; insbesondere in Fällen, in denen die Prüfungsparameter durch die Tools selbst nur schwer geändert werden können.
  • Erkennung statt Analyse – Business-Intelligence-Tools sind in erster Linie darauf ausgelegt, Daten zu finden. Der Fokus hierbei liegt vor allem auf Daten, die für ihre Benutzer nützlich sein können. An dieser Stelle endet jedoch häufig die Leistungsfähigkeit der Tools, da sie Benutzern keine einfachen Optionen bieten, die Daten tatsächlich zu analysieren. Die Möglichkeit, umsetzbare Erkenntnisse zu gewinnen, verringert sich somit.
  • Eingeschränkte Skalierbarkeit – Im Allgemeinen bleibt Business Intelligence ein Bereich für Spezialisten und Experten mit dem entsprechenden Know-how, über das Mitarbeiter im operativen Bereich oftmals nicht verfügen. Ohne umfangreiches Verständnis für die geschäftlichen Prozesse und deren Analyse innerhalb des Unternehmens bleibt die optimierte Anwendung eines bestimmten Business-Intelligence-Tools aber eingeschränkt.
  • Nicht nachvollziehbare Metriken – Werden Metriken verwendet, die nicht mit den Geschäftsprozessen verknüpft sind, kann Business Intelligence kaum positive Veränderungen innerhalb eines Unternehmens unterstützen. Für Benutzer ist es schwierig, Ergebnisse richtig auszuwerten und zu verstehen und diese Ergebnisse zweckdienlich zu nutzen.

Process Intelligence: der nächste wegweisende Schritt

Es bedarf einer effektiveren Methode zur Prozessanalyse, um eine effiziente Arbeitsweise und fundierte Entscheidungsfindung sicherzustellen. An dieser Stelle kommt Process Intelligence (PI) ins Spiel. PI bietet die entscheidenden Hintergrundinformationen für die Beantwortung von Fragen, die mit Business-Intelligence-Tools unbeantwortet bleiben.

Process Intelligence ermöglicht die durchgehende Visualisierung von Prozessabläufen mithilfe von Rohdaten. Mit dem richtigen Process-Intelligence-Tool können diese Rohdaten sofort analysiert werden, sodass Prozesse präzise angezeigt werden. Der Endbenutzer kann diese Informationen nach Bedarf einsehen und bearbeiten, ohne eine Vorauswahl für die Analyse treffen zu müssen.

Zum Vergleich: Da Business Intelligence vordefinierte Analysekriterien benötigt, kann BI nur dann wirklich nützlich sein, wenn diese Kriterien auch definiert sind. Unternehmen können verzögerte Analysen vermeiden, indem sie Process Intelligence zur Ermittlung der Hauptursache von Prozessproblemen nutzen, und dann die richtigen Kriterien zur Bestimmung des Analyserahmens auswählen.

Anschließend können Sie Ihre Systemprozesse analysieren und erkennen die Diskrepanzen und Varianten zwischen dem angestrebten Geschäftsprozess und dem tatsächlichen Verlauf Ihrer Prozesse. Und je schneller Sie Echtzeit-Einblicke in Ihre Prozesse gewinnen, desto schneller können Sie in Ihrem Unternehmen positive Veränderungen auf den Weg bringen.

Kurz gesagt: Business Intelligence eignet sich dafür, ein breites Verständnis über die Abläufe in einem Unternehmen zu gewinnen. Für einige Unternehmen kann dies ausreichend sein. Für andere hingegen ist ein Überblick nicht genug.

Sie suchen nach einer Möglichkeit um festzustellen, wie jeder Prozess in Ihrer Organisation tatsächlich funktioniert? Die Antwort hierauf lautet Software. Software, die Prozesserkennung, Prozessanalyse und Konformitätsprüfung miteinander kombiniert.

Mit den richtigen Process-Intelligence-Tools können Sie nicht nur Daten aus den verschiedenen IT-Systemen in Ihrem Unternehmen gewinnen, sondern auch Ihre End-to-End-Prozesse kontinuierlich überwachen. So erhalten Sie Erkenntnisse über mögliche Risiken und Verbesserungspotenziale. PI steht für einen kollaborativen Ansatz zur Prozessverbesserung, der zu einem bahnbrechenden Verständnis über die Abläufe in Ihrem Unternehmen führt, und wie diese optimiert werden können.

Erhöhtes Potenzial mit Signavio Process Intelligence

Mit Signavio Process Intelligence erhalten Sie wegweisende Erkenntnisse über Ihre Prozesse, auf deren Basis Sie bessere Geschäftsentscheidungen treffen können. Erlangen Sie eine vollständige Sicht auf Ihre Abläufe und ein Verständnis dafür, was in Ihrer Organisation tatsächlich geschieht.

Als Teil der Signavio Business Transformation Suite lässt sich Signavio Process Intelligence perfekt mit der Prozessmodellierung und -automatisierung kombinieren. Als eine vollständig cloudbasierte Process-Mining-Lösung erleichtert es die Software, organisationsweit zusammenzuarbeiten und Wissen zu teilen.

Generieren Sie neue Ideen, sparen Sie Aufwand und Kosten ein und optimieren Sie Ihre Prozesse. Erfahren Sie mehr über Signavio Process Intelligence.

From BI to PI: The Next Step in the Evolution of Data-Driven Decisions

“Change is a constant.” “The pace of change is accelerating.” “The world is increasingly complex, and businesses have to keep up.” Organizations of all shapes and sizes have heard these ideas over and over—perhaps too often! However, the truth remains that adaptation is crucial to a successful business.


Read this article in German: Von der Datenanalyse zur Prozessverbesserung: So gelingt eine erfolgreiche Process-Mining-Initiative

 


Of course, the only way to ensure that the decisions you make are evolving in the right way is to understand the underlying building blocks of your organization. You can think of it as DNA; the business processes that underpin the way you work and combine to create a single unified whole. Knowing how those processes operate, and where the opportunities for improvement lie, can be the difference between success and failure.

Businesses with an eye on their growth understand this already. In the past, Business Intelligence was seen as the solution to this challenge. In more recent times, forward-thinking organizations see the need for monitoring solutions that can keep up with today’s rate of change, at the same time as they recognize that increasing complexity within business processes means traditional methods are no longer sufficient.

Adapting to a changing environment? The challenges of BI

Business Intelligence itself is not necessarily defunct or obsolete. However, the tools and solutions that enable Business Intelligence face a range of challenges in a fast-paced and constantly changing world. Some of these issues may include:

  • High data latency – Data latency refers to how long it takes for a business user to retrieve data from, for example, a business intelligence dashboard. In many cases, this can take more than 24 hours, a critical time period when businesses are attempting to take advantage of opportunities that may have a limited timeframe.
  • Incomplete data sets – The broad approach of Business Intelligence means investigations may run wide but not deep. This increases the chances that data will be missed, especially in instances where the tools themselves make the parameters for investigations difficult to change.
  • Discovery, not analysis – Business intelligence tools are primarily optimized for exploration, with a focus on actually finding data that may be useful to their users. Often, this is where the tools stop, offering no simple way for users to actually analyze the data, and therefore reducing the possibility of finding actionable insights.
  • Limited scalability – In general, Business Intelligence remains an arena for specialists and experts, leaving a gap in understanding for operational staff. Without a wide appreciation for processes and their analysis within an organization, the opportunities to increase the application of a particular Business Intelligence tool will be limited.
  • Unconnected metrics – Business Intelligence can be significantly restricted in its capacity to support positive change within a business through the use of metrics that are not connected to the business context. This makes it difficult for users to interpret and understand the results of an investigation, and apply these results to a useful purpose within their organization.

Process Intelligence: the next evolutionary step

To ensure companies can work efficiently and make the best decisions, a more effective method of process discovery is needed. Process Intelligence (PI) provides the critical background to answer questions that cannot be answered with Business Intelligence tools.

Process Intelligence offers visualization of end-to-end process sequences using raw data, and the right Process Intelligence tool means analysis of that raw data can be conducted straight away, so that processes are displayed accurately. The end-user is free to view and work with this accurate information as they please, without the need to do a preselection for the analysis.

By comparison, because Business Intelligence requires predefined analysis criteria, only once the criteria are defined can BI be truly useful. Organizations can avoid delayed analysis by using Process Intelligence to identify the root causes of process problems, then selecting the right criteria to determine the analysis framework.

Then, you can analyze your system processes and see the gaps and variants between the intended business process and what you actually have. And of course, the faster you discover what you have, the faster you can apply the changes that will make a difference in your business.

In short, Business Intelligence is suitable for gaining a broad understanding of the way a business usually functions. For some businesses, this will be sufficient. For others, an overview is not enough.

They understand that true insights lie in the detail, and are looking for a way of drilling down into exactly how each process within their organization actually works. Software that combines process discovery, process analysis, and conformance checking is the answer.

The right Process Intelligence tools means you will be able to automatically mine process models from the different IT systems operating within your business, as well as continuously monitor your end-to-end processes for insights into potential risks and ongoing improvement opportunities. All of this is in service of a collaborative approach to process improvement, which will lead to a game-changing understanding of how your business works, and how it can work better.

Early humans evolved from more primitive ancestors, and in the process, learned to use more and more sophisticated tools. For the modern human, working in a complex organization, the right tool is Process Intelligence.

Endless Potential with Signavio Process Intelligence

Signavio Process Intelligence allows you to unearth the truth about your processes and make better decisions based on true evidence found in your organization’s IT systems. Get a complete end-to-end perspective and understanding of exactly what is happening in your organization in a matter of weeks.

As part of Signavio Business Transformation Suite, Signavio Process Intelligence integrates perfectly with Signavio Process Manager and is accessible from the Signavio Collaboration Hub. As an entirely cloud-based process mining solution, the tool makes it easy to collaborate with colleagues from all over the world and harness the wisdom of the crowd.

Find out more about Signavio Process Intelligence, and see how it can help your organization generate more ideas, save time and money, and optimize processes.

Von der Datenanalyse zur Prozessverbesserung: So gelingt eine erfolgreiche Process-Mining-Initiative

Den Prozessdaten auf der Spur: Systematische Datenanalyse kombiniert mit Prozessmanagement

Die Digitalisierung verändert Organisationen aller Branchen. In zahlreichen Unternehmen werden alltägliche Betriebsabläufe softwarebasiert modelliert, automatisiert und optimiert. Damit hinterlässt fast jeder Prozess elektronische Spuren in den CRM-, ERP- oder anderen IT-Systemen einer Organisation. Process Mining gilt als effektive Methode, um diese Datenspuren zusammenzuführen und für umfassende Auswertungen zu nutzen. Sie kombiniert die systematische Datenanalyse mit Geschäftsprozessmanagement: Dabei werden Prozessdaten aus den verschiedenen IT-Systemen einer Organisation extrahiert und mit Hilfe von Data-Science-Technologien visualisiert und ausgewertet.


Read this article in English: From BI to PI: The Next Step in the Evolution of Data-Driven Decisions

 


Professionelle Process-Mining-Lösungen erlauben, die Ergebnisse dieser Prozessauswertungen auf Dashboards darzustellen und nach bestimmten Prozessen, Transaktionen, Abteilungen oder Kunden zu filtern. So ist es möglich, die Performance, Durchlaufzeiten und die Kosten einzelner Betriebsabläufe zu erfassen. Prozessverantwortliche werden auf diesem Wege auf Verzögerungen, ineffiziente Abläufe und mögliche Prozessverbesserungen aufmerksam.

Praxisbeispiel: Einkaufsprozess – Prozessabweichungen als Kosten- und Risikofaktor

Ein Beispiel aus dem Unternehmensalltag ist ein einfacher Einkaufsprozess: Ein Mitarbeiter benötigt einen neuen Laptop. Im Normalfall beginnt der Prozess mit der Anfrage des Mitarbeiters, die durch seinen Manager bestätigt wird. Ist kein Laptop vorrätig, löst das für den Einkauf zuständige Team die Bestellung aus. Zu einem späteren Zeitpunkt wird der Laptop dem Mitarbeiter übergeben und das Unternehmen erhält eine Rechnung. Diese Rechnung wird geprüft und fristgemäß gemäß den vorgegebenen Konditionen beglichen. Obwohl dieser alltägliche Prozess nicht sehr komplex ist, weicht er im Unternehmensalltag häufig vom modellierten Idealzustand ab, was unnötige Kosten und möglicherweise auch Risiken verursacht.

Die Gründe sind vielfältig:

  • Freigaben fehlen
  • Während des Bestellprozesses sind Informationen unvollständig
  • Rechnungen werden aufgrund von unvollständigen Informationen mehrfach korrigiert

Process Mining ermöglicht, den gesamten Prozessverlauf alltäglicher Betriebsabläufe unter die Lupe zu nehmen und faktenbasierte Diskussionen zwischen den Fachabteilungen, Prozessverantwortlichen sowie dem Management in einer Organisation anzuregen. So werden unternehmensweite Prozessverbesserungen möglich – vorausgesetzt, die Methode wird richtig angewandt und ist strategisch durchdacht. Doch wie gelingt eine erfolgreiche unternehmensweite Process-Mining-Initiative über Abteilungsgrenzen hinaus?

Wie sich eine erfolgreiche Process-Mining-Initiative auf den Weg bringen lässt

Jedes Unternehmen ist einzigartig und geht mit unterschiedlichen Fragestellungen an eine Process-Mining-Initiative heran: ob einzelne Prozesse gezielt verbessert, Prozesslebenszyklen verkürzt oder abteilungsübergreifende Abläufe an unterschiedlichen Standorten miteinander verglichen werden. Sie alle haben etwas gemeinsam: Eine erfolgreiche Process-Mining-Initiative erfordert ein strategisches Vorgehen.

Schritt 1: Mit Weitsicht planen und richtig kommunizieren

Wie definiere ich die Ziele und den Umfang der Process-Mining-Initiative?

Die Anfangsphase einer Process-Mining-Initiative dient der Planung und entscheidet häufig über den Erfolg eines Projektes. In erster Linie kommt es darauf an, die Ziele des Projektes zu definieren und die Erfolgsfaktoren zu bestimmen. Die Ziele einer erfolgreichen Process-Mining-Initiative sind SMART definiert: spezifisch, messbar, attainable/relevant, reasonable/umsetzbar und zeitgebunden/time-bound. Mögliche Ziele für das Projekt lassen sich zum Beispiel wie folgt formulieren:

  • Prozessdauer auf 25 Tage reduzieren
  • Hauptunterschiede zwischen zwei Ländern hinsichtlich bestimmter Prozesse identifizieren
  • Prozessautomatisierung um 25% steigern

Unter diesen Voraussetzungen lässt sich auch der Rahmen der Process-Mining-Initiative festlegen: Sie halten fest, welche Prozesse, konkret betroffen sind und wie sie mit den IT-Systemen und Mitarbeiterrollen in Ihrer Organisation verknüpft sind.

Welche Rollen und Verantwortlichkeiten gibt es?

Die Ziele Ihrer Process-Mining-Initiative sollten unternehmensweit geteilt werden: Dies erfordert neben einer klaren Strategie eine transparente Kommunikation in der gesamten Organisation: Indem Sie Ihren Mitarbeitern das nötige Wissen an die Hand geben, um die Initiative erfolgreich mitzugestalten, sichern Sie sich auch ihre Unterstützung.

So verstehen sie nicht nur, warum dieses Projekt sinnvoll ist, sondern sind auch in der Lage, das Wissen auf ihre individuelle Rolle und Situation zu übertragen. Im Rahmen einer Process-Mining-Initiative sind verschiedene Projektbeteiligte in unterschiedlichen Rollen aktiv:

Während Projektträger verantwortlich für die Prozessanalyse sind (z. B. Chief Procurement Officer oder Process Owner), wissen Prozessexperten, wie ein bestimmter Prozess verläuft und kennen die verschiedenen Variationen. Sie nutzen Methoden wie Process Mining, um ihr Wissen zu vertiefen und Diskussionen über die gewonnenen Daten anzustoßen. Sie arbeiten eng mit Business-Analysten zusammen, die die Prozessanalyse vorantreiben. Datenexperten wiederum verfolgen die einzelnen Spuren, die ein Prozess in der IT-Landschaft einer Organisation hinterlässt und bereiten sie so auf, dass sie Aufschluss über die Performance eines Prozesses geben.

Wie gestaltet sich die Zusammenarbeit?

Diese unterschiedlichen Rollen gilt es im Rahmen einer erfolgreichen Process-Mining-Initiative an einen Tisch zu bringen: So können die gewonnen Erkenntnisse gemeinsam im Team interpretiert und diskutiert werden, um die richtigen Veränderungen anzustoßen. Die daraus gewonnen Prozessverbesserungen spiegeln das Know-how des gesamten Teams wider und sind das Ergebnis einer erfolgreichen Zusammenarbeit.

Schritt 2: Die technischen Voraussetzungen schaffen

Wie werden Prozessdaten systemübergreifend aggregiert und aufbereitet?

Nun wird es Zeit für die technischen Vorbereitungen: Entscheidend ist es, alle Anforderungen an die beteiligten IT-Systeme zu durchdenken und die IT-Verantwortlichen so früh wie möglich einzubeziehen. Um valide Daten für Prozessverbesserungen zu generieren, sind diese drei Teilschritte nötig:

  1.  Datenextraktion: Relevante Daten aus unterschiedlichen IT-Systemen werden aggregiert (Datenquellen sind datenbasierte Tabellen aus ERP- und CRM-Lösungen, analytische Daten wie Reports, Logdateien, CSV-Dateien usw.)
  2.  Datenumwandlung gemäß den Anforderungen für Process Mining: Die extrahierten Daten werden in Cases (Abfolge verschiedener Prozessschritte) umgewandelt, mit einem Zeitstempel versehen und in Event-Logs gespeichert.
  3.  Datenübertragung: Die Process-Mining-Software greift auf die gespeicherten Event-Logs zu.

Welche Rolle spielen Konnektoren?

Diese Teilschritte werden erfahrungsgemäß mittels eines Software-Konnektors durchgeführt und in regelmäßigen Abständen wiederholt. Ein Software-Konnektor hat die Aufgabe, die Daten aus der IT-Landschaft eines Unternehmens nach den Anforderungen der Process-Mining-Lösung zu übersetzen. Er wird speziell für die Kombination mit bestimmten IT-Systemen wie SAP, Oracle oder Salesforce entwickelt und steuert die gesamte Datenintegration von der Extraktion über die Umwandlung bis zur Datenübertragung.

Process-Mining-Lösungen wie Signavio Process Intelligence verfügen über Standardkonnektoren sowie über eine API für individuell entwickelte Konnektoren. Im Rahmen der technischen Vorbereitungen gilt es, mit Blick auf das jeweilige Szenario über die Möglichkeiten der Umsetzbarkeit zu entscheiden und andere technische Lösungen zu evaluieren.

Schritt 3: Von der Prozessanalyse zur Prozessverbesserung

Wie lassen sich die ermittelten Daten für Verbesserungen nutzen?

Sind die umgewandelten Daten in der Process-Mining-Lösung verfügbar, beginnt die Prozessauswertung. Durch IT-gestütztes Process Mining erhalten Prozessexperten die Möglichkeit, alle vorliegenden Daten zu visualisieren und einzelne Prozesse detailliert auszuwerten. Die vorliegenden Prozesse werden nun hinsichtlich unterschiedlicher Faktoren untersucht, etwa mit Blick auf Durchlaufzeiten, Performance und den Prozessfluss. Im direkten Vergleich lässt sich auf diesem Wege ermitteln, welche Faktoren sich auf die Erfolgskennzahlen auswirken und an welchen Stellen Verzögerungen oder Abweichungen auftreten.

Die so gewonnen Erkenntnisse bilden eine wichtige Grundlage für faktenbasierte Diskussionen zwischen den verschiedenen Stakeholdern der Process-Mining-Initiative. Doch erst die konkreten Schritte, die aus dieser Datenbasis abgeleitet werden, entscheiden über den Erfolg des Projektes: Entscheidend ist, wie diese Erkenntnisse in die Praxis umgesetzt werden.

 

Eine Process-Mining-Lösung, die nicht als reines Analysetool zur Verfügung steht, sondern in eine umfassende Lösung für die Modellierung, Automatisierung und Analyse professioneller Geschäftsprozesse integriert ist, erleichtert den Schritt von der Business Process Discovery zur Prozessverbesserung. Schließlich gilt es, konkrete Prozessverbesserungen und Änderungen zu planen, in den Unternehmensalltag zu integrieren und die Ergebnisse auszuwerten – auch über das Ende der Process-Mining-Initiative hinaus.

Warum ist ein Process-Mining-Projekt nie vollständig abgeschlossen?  

Wer einmal mit der Prozessverbesserung beginnt, wird feststellen: Viele weitere Stellen in den Prozessen warten nur darauf, verbessert zu werden. Daher lohnt es sich, einige Wochen nach der initialen Prozessverbesserung neue Daten zu extrahieren, um herauszufinden, welche Veränderungen nachweislich zu mehr Effizienz geführt haben. Eine kontinuierliche Messung und Auswertung erleichtert einen umfassenden Blick auf die eigene Organisation:

  • Funktionieren die überarbeiteten Prozesse wie geplant?
  • Haben Prozessveränderungen unvorhersehbare Effekte?
  • Treten Schwachstellen in anderen Prozessen auf?
  • Haben sich die Prozesse verändert, seitdem sie überarbeitet wurden?
  • Wie lässt sich ein bestimmter Prozess weiter verbessern?

Somit lässt sich zusammenfassen: Wem es gelingt, die Datenspuren in den IT-Systemen der eigenen Organisation zu verfolgen, ist auf dem richtigen Weg zur kontinuierlichen Verbesserung. Davon profitieren nicht nur die Prozesse und IT-Systeme, sondern auch die Mitarbeiter in den Organisationen.

Attribution Models in Marketing

Attribution Models

A Business and Statistical Case

INTRODUCTION

A desire to understand the causal effect of campaigns on KPIs

Advertising and marketing costs represent a huge and ever more growing part of the budget of companies. Studies have found out this share is as high as 10% and increases with the size of companies (CMO study by American Marketing Association and Duke University, 2017). Measuring precisely the impact of a specific marketing campaign on the sales of a company is a critical step towards an efficient allocation of this budget. Would the return be higher for an euro spent on a Facebook ad, or should we better spend it on a TV spot? How much should I spend on Twitter ads given the volume of sales this channel is responsible for?

Attribution Models have lately received great attention in Marketing departments to answer these issues. The transition from offline to online marketing methods has indeed permitted the collection of multiple individual data throughout the whole customer journey, and  allowed for the development of user-centric attribution models. In short, Attribution Models use the information provided by Tracking technologies such as Google Analytics or Webtrekk to understand customer journeys from the first click on a Facebook ad to the final purchase and adequately ponderate the different marketing campaigns encountered depending on their responsibility in the final conversion.

Issues on Causal Effects

A key question then becomes: how to declare a channel is responsible for a purchase? In other words, how can we isolate the causal effect or incremental value of a campaign ?

          1. A/B-Tests

One method to estimate the pure impact of a campaign is the design of randomized experiments, wherein a control and treated groups are compared.  A/B tests belong to this broad category of randomized methods. Provided the groups are a priori similar in every aspect except for the treatment received, all subsequent differences may be attributed solely to the treatment. This method is typically used in medical studies to assess the effect of a drug to cure a disease.

Main practical issues regarding Randomized Methods are:

  • Assuring that control and treated groups are really similar before treatment. Uually a random assignment (i.e assuring that on a relevant set of observable variables groups are similar) is realized;
  • Potential spillover-effects, i.e the possibility that the treatment has an impact on the non-treated group as well (Stable unit treatment Value Assumption, or SUTVA in Rubin’s framework);
  • The costs of conducting such an experiment, and especially the costs linked to the deliberate assignment of individuals to a group with potentially lower results;
  • The number of such experiments to design if multiple treatments have to be measured;
  • Difficulties taking into account the interaction effects between campaigns or the effect of spending levels. Indeed, usually A/B tests are led by cutting off temporarily one campaign entirely and measuring the subsequent impact on KPI’s compared to the situation where this campaign is maintained;
  • The dynamical reproduction of experiments if we assume that treatment effects may change over time.

In the marketing context, multiple campaigns must be tested in a dynamical way, and treatment effect is likely to be heterogeneous among customers, leading to practical issues in the lauching of A/B tests to approximate the incremental value of all campaigns. However, sites with a lot of traffic and conversions can highly benefit from A/B testing as it provides a scientific and straightforward way to approximate a causal impact. Leading companies such as Uber, Netflix or Airbnb rely on internal tools for A/B testing automation, which allow them to basically test any decision they are about to make.

References:

Books:

Experiment!: Website conversion rate optimization with A/B and multivariate testing, Colin McFarland, ©2013 | New Riders  

A/B testing: the most powerful way to turn clicks into customers. Dan Siroker, Pete Koomen; Wiley, 2013.

Blogs:

https://eng.uber.com/xp

https://medium.com/airbnb-engineering/growing-our-host-community-with-online-marketing-9b2302299324

Study:

https://cmosurvey.org/wp-content/uploads/sites/15/2018/08/The_CMO_Survey-Results_by_Firm_and_Industry_Characteristics-Aug-2018.pdf

        2. Attribution models

Attribution Models do not demand to create an experimental setting. They take into account existing data and derive insights from the variability of customer journeys. One key difficulty is then to differentiate correlation and causality in the links observed between the exposition to campaigns and purchases. Indeed, selection effects may bias results as exposure to campaigns is usually dependant on user-characteristics and thus may not be necessarily independant from the customer’s baseline conversion probabilities. For example, customers purchasing from a discount price comparison website may be intrinsically different from customers buying from FB ad and this a priori difference may alone explain post-exposure differences in purchasing bahaviours. This intrinsic weakness must be remembered when interpreting Attribution Models results.

                          2.1 General Issues

The main issues regarding the implementation of Attribution Models are linked to

  • Causality and fallacious reasonning, as most models do not take into account the aforementionned selection biases.
  • Their difficult evaluation. Indeed, in almost all attribution models (except for those based on classification, where the accuracy of the model can be computed), the additionnal value brought by the use of a given attribution models cannot be evaluated using existing historical data. This additionnal value can only be approximated by analysing how the implementation of the conclusions of the attribution model have impacted a given KPI.
  • Tracking issues, leading to an uncorrect reconstruction of customer journeys
    • Cross-device journeys: cross-device issue arises from the use of different devices throughout the customer journeys, making it difficult to link datapoints. For example, if a customer searches for a product on his computer but later orders it on his mobile, the AM would then mistakenly consider it an order without prior campaign exposure. Though difficult to measure perfectly, the proportion of cross-device orders can approximate 20-30%.
    • Cookies destruction makes it difficult to track the customer his the whole journey. Both regulations and consumers’ rising concerns about data privacy issues mitigate the reliability and use of cookies.1 – From 2002 on, the EU has enacted directives concerning privacy regulation and the extended use of cookies for commercial targeting purposes, which have highly impacted marketing strategies, such as the ‘Privacy and Electronic Communications Directive’ (2002/58/EC). A research was conducted and found out that the adoption of this ‘Privacy Directive’ had led to 64% decrease in advertising methods compared to the rest of the world (Goldfarb et Tucker (2011)). The effect was stronger for generalized sites (Yahoo) than for specialized sites.2 – Users have grown more and more conscious of data privacy issues and have adopted protective measures concerning data privacy, such as automatic destruction of cookies after a session is ended, or simply giving away less personnal information (Goldfarb et Tucker (2012) ) .Valuable user information may be lost, though tracking technologies evolution have permitted to maintain tracking by other means. This issue may be particularly important in countries highly concerned with data privacy issues such as Germany.
    • Offline/Online bridge: an Attribution Model should take into account all campaigns to draw valuable insights. However, the exposure to offline campaigns (TV, newspapers) are difficult to track at the user level. One idea to tackle this issue would be to estimate the proportion of conversions led by offline campaigns through AB testing and deduce this proportion from the credit assigned to the online campaigns accounted for in the Attribution Model.
    • Touch point information available: clicks are easy to follow but irrelevant to take into account the influence of purely visual campaigns such as display ads or video.

                          2.2 Today’s main practices

Two main families of Attribution Models exist:

  • Rule-Based Attribution Models, which have been used for in the last decade but from which companies are gradualy switching.

Attribution depends on the individual journeys that have led to a purchase and is solely based on the rank of the campaign in the journey. Some models focus on a single touch points (First Click, Last Click) while others account for multi-touch journeys (Bathtube, Linear). It can be calculated at the customer level and thus doesn’t require large amounts of data points. We can distinguish two sub-groups of rule-based Attribution Models:

  • One Touch Attribution Models attribute all credit to a single touch point. The First-Click model attributes all credit for a converion to the first touch point of the customer journey; last touch attributes all credit to the last campaign.
  • Multi-touch Rule-Based Attribution Models incorporate information on the whole customer journey are thus an improvement compared to one touch models. To this family belong Linear model where credit is split equally between all channels, Bathtube model where 40% of credit is given to first and last clicks and the remaining 20% is distributed equally between the middle channels, or time-decay models where credit assigned to a click diminishes as the time between the click and the order increases..

The main advantages of rule-based models is their simplicity and cost effectiveness. The main problems are:

– They are a priori known and can thus lead to optimization strategies from competitors
– They do not take into account aggregate intelligence on customer journeys and actual incremental values.
– They tend to bias (depending on the model chosen) channels that are over-represented at the beggining or end of the funnel, according to theoretical assumptions that have no observationnal back-ups.

  • Data-Driven Attribution Models

These models take into account the weaknesses of rule-based models and make a relevant use of available data. Being data-driven, following attribution models cannot be computed using single user level data. On the contrary values are calculated through data aggregation and thus require a certain volume of customer journey information.

References:

https://dspace.mit.edu/handle/1721.1/64920

 

        3. Data-Driven Attribution Models in practice

                          3.1 Issues

Several issues arise in the computation of campaigns individual impact on a given KPI within a data-driven model.

  • Selection biases: Exposure to certain types of advertisement is usually highly correlated to non-observable variables which are in turn correlated to consumption practices. Differences in the behaviour of users exposed to different campaigns may thus only be driven by core differences in conversion probabilities between groups whether than by the campaign effect.
  • Complementarity: it may be that campaigns A and B only have an effect when combined, so that measuring their individual impact would lead to misleading conclusions. The model could then try to assess the effect of combinations of campaigns on top of the effect of individual campaigns. As the number of possible non-ordered combinations of k campaigns is 2k, it becomes clear that inclusing all possible combinations would however be time-consuming.
  • Order-sensitivity: The effect of a campaign A may depend on the place where it appears in the customer journey, meaning the rank of a campaign and not merely its presence could be accounted for in the model.
  • Relative Order-sensitivity: it may be that campaigns A and B only have an effect when one is exposed to campaign A before campaign B. If so, it could be useful to assess the effect of given combinations of campaigns as well. And this for all campaigns, leading to tremendous numbers of possible combinations.
  • All previous phenomenon may be present, increasing even more the potential complexity of a comprehensive Attribution Model. The number of all possible ordered combination of k campaigns is indeed :

 

                          3.2 Main models

                                  A) Logistic Regression and Classification models

If non converting journeys are available, Attribition Model can be shaped as a simple classification issue. Campaign types or campaigns combination and volume of campaign types can be included in the model along with customer or time variables. As we are interested in inference (on campaigns effect) whether than prediction, a parametric model should be used, such as Logistic Regression. Non paramatric models such as Random Forests or Neural Networks can also be used though the interpretation of campaigns value would be more difficult to derive from the model results.

A common pitfall is the usual issue of spurious correlations on one hand and the correct interpretation of coefficients in business terms.

An advantage if the possibility to evaluate the relevance of the model using common model validation methods to evaluate its predictive power (validation set \ AUC \pseudo R squared).

                                  B) Shapley Value

Theory

The Shapley Value is based on a Game Theory framework and is named after its creator, the Nobel Price Laureate Lloyd Shapley. Initially meant to calculate the marginal contribution of players in cooperative games, the model has received much attention in research and industry and has lately been applied to marketing issues. This model is typically used by Google Adords and other ad bidding vendors. Campaigns or marketing channels are in this model seen as compementary players looking forward to increasing a given KPI.
Contrarily to Logistic Regressions, it is a non-parametric model. Contrarily to Markov Chains, all results are built using existing journeys, and not simulated ones.

Channels are considered to enter the game sequentially under a certain joining order. Shapley value try to The Shapley value of channel i is the weighted sum of the marginal values that channel i adds to all possible coalitions that don’t contain channel i.
In other words, the main logic is to analyse the difference of gains when a channel i is added after a coalition Ck of k channels, k<=n. We then sum all the marginal contributions over all possible ordered combination Ck of all campaigns excluding i, with k<=n-1.

Subsets framework

A first an most usual way to compute the Shapley Vaue is to consider that when a channel enters coalition, its additionnal value is the same irrelevant of the order in which previous channels have appeared. In other words, journeys (A>B>C) and (B>A>C) trigger the same gains.
Shapley value is computed as the gains associated to adding a channel i to a subset of channels, weighted by the number of (ordered) sequences that the (unordered) subset represents, summed up on all possible subsets of the total set of campaigns where the channel i is not present.
The Shapley value of the channel 𝑥𝑗 is then:

where |S| is the number of campaigns of a coalition S and the sum extends over all subsets S that do not not contain channel j. 𝜈(𝑆)  is the value of the coalition S and 𝜈(𝑆 ∪ {𝑥𝑗})  the value of the coalition formed by adding 𝑥𝑗 to coalition S. 𝜈(𝑆 ∪ {𝑥𝑗}) − 𝜈(𝑆) is thus the marginal contribution of channel 𝑥𝑗 to the coalition S.

The formula can be rewritten and understood as:

This method is convenient when data on the gains of on all possible permutations of all unordered k subsets of the n campaigns are available. It is also more convenient if the order of campaigns prior to the introduction of a campaign is thought to have no impact.

Ordered sequences

Let us define 𝜈((A>B)) as the value of the sequence A then B. What is we let 𝜈((A>B)) be different from 𝜈((B>A)) ?
This time we would need to sum over all possible permutation of the S campaigns present before  𝑥𝑗 and the N-(S+1) campaigns after 𝑥𝑗. Doing so we will sum over all possible orderings (i.e all permutations of the n campaigns of the grand coalition containing all campaigns) and we can remove the permutation coefficient s!(p-s+1)!.

This method is convenient when the order of channels prior to and after the introduction of another channel is assumed to have an impact. It is also necessary to possess data for all possible permutations of all k subsets of the n campaigns, and not only on all (unordered) k-subsets of the n campaigns, k<=n. In other words, one must know the gains of A, B, C, A>B, B>A, etc. to compute the Shapley Value.

Differences between the two approaches

We simulate an ordered case where the value for each ordered sequence k for k<=3 is known. We compare it to the usual Shapley value calculated based on known gains of unordered subsets of campaigns. So as to compare relevant values, we have built the gains matrix so that the gains of a subset A, B i.e  𝜈({B,A}) is the average of the gains of ordered sequences made up with A and B (assuming the number of journeys where A>B equals the number of journeys where B>A, we have 𝜈({B,A})=0.5( 𝜈((A>B)) + 𝜈((B>A)) ). We let the value of the grand coalition be different depending on the order of campaigns-keeping the constraints that it averages to the value used for the unordered case.

Note: mvA refers to the marginal value of A in a given sequence.
With traditionnal unordered coalitions:

With ordered sequences used to compute the marginal values:

 

We can see that the two approaches yield very different results. In the unordered case, the Shapley Value campaign C is the highest, culminating at 20, while A and B have the same Shapley Value mvA=mvB=15. In the ordered case, campaign A has the highest Shapley Value and all campaigns have different Shapley Values.

This example illustrates the inherent differences between the set and sequences approach to Shapley values. Real life data is more likely to resemble the ordered case as conversion probabilities may for any given set of campaigns be influenced by the order through which the campaigns appear.

Advantages

Shapley value has become popular in allocation problems in cooperative games because it is the unique allocation which satisfies different axioms:

  • Efficiency: Shaple Values of all channels add up to the total gains (here, orders) observed.
  • Symmetry: if channels A and B bring the same contribution to any coalition of campaigns, then their Shapley Value i sthe same
  • Null player: if a channel brings no additionnal gains to all coalitions, then its Shapley Value is zero
  • Strong monotony: the Shapley Value of a player increases weakly if all its marginal contributions increase weakly

These properties make the Shapley Value close to what we intuitively define as a fair attribution.

Issues

  • The Shapley Value is based on combinatory mathematics, and the number of possible coalitions and ordered sequences becomes huge when the number of campaigns increases.
  • If unordered, the Shapley Value assumes the contribution of campaign A is the same if followed by campaign B or by C.
  • If ordered, the number of combinations for which data must be available and sufficient is huge.
  • Channels rarely present or present in long journeys will be played down.
  • Generally, gains are supposed to grow with the number of players in the game. However, it is plausible that in the marketing context a journey with a high number of channels will not necessarily bring more orders than a journey with less channels involved.

References:

R package: GameTheoryAllocation

Article:
Zhao & al, 2018 “Shapley Value Methods for Attribution Modeling in Online Advertising “
https://link.springer.com/content/pdf/10.1007/s13278-017-0480-z.pdf
Courses: https://www.lamsade.dauphine.fr/~airiau/Teaching/CoopGames/2011/coopgames-7%5b8up%5d.pdf
Blogs: https://towardsdatascience.com/one-feature-attribution-method-to-supposedly-rule-them-all-shapley-values-f3e04534983d

                                  B) Markov Chains

Markov Chains are used to model random processes, i.e events that occur in a sequential manner and in such a way that the probability to move to a certain state only depends on the past steps. The number of previous steps that are taken into account to model the transition probability is called the memory parameter of the sequence, and for the model to have a solution must be comprised between 0 and 4. A Markov Chain process is thus defined entirely by its Transition Matrix and its initial vector (i.e the starting point of the process).

Markov Chains are applied in many scientific fields. Typically, they are used in weather forecasting, with the sequence of Sunny and Rainy days following a Markov Process of memory parameter 0, so that for each given day the probability that the next day will be rainy or sunny only depends on the weather of the current day. Other applications can be found in sociology to understand the dynamics of social classes intergenerational reproduction. To get more both mathematical and applied illustration, I recommend the reading of this course.

In the marketing context, Markov Chains are an interesting way to model the conversion funnel. To go from the from the Markov Model to the Attribution logic, we calculate the Removal Effect of each channel, i.e the difference in conversions that happen if the channel is removed. Please read below for an introduction to the methodology.

The first step in a Markov Chains Attribution Model is to build the transition matrix that captures the transition probabilities between the campaigns accross existing customer journeys. This Matrix is to be read as a “From state A to state B” table, from the left to the right. A first difficulty is finding the right memory parameter to use. A large memory parameter would allow to take more into account interraction effects within the conversion funnel but would lead to increased computationnal time, a non-readable transition matrix, and be more sensitive to noisy data. Please note that this transition matrix provides useful information on the conversion funnel and on the relationships between campaigns and can be used as such as an analytical tool. I suggest the clear and easily R code which can be found here or here.

Here is an illustration of a Markov Chain with memory Parameter of 0: the probability to go to a certain campaign B in the next step only depend on the campaign we are currently at:

The associated Transition Matrix is then (with null probabilities left as Blank):

The second step is  to compute the actual responsibility of a channel in total conversions. As mentionned above, the main philosophy to do so is to calculate the Removal Effect of each channel, i.e the changes in the number of conversions when a channel is entirely removed. All customer journeys which went through this channel are settled out to be unsuccessful. This calculation is done by applying the transition matrix with and without the removed channels to an initial vector that contains the number of desired simulations.

Building on our current example, we can then settle an initial vector with the desired number of simulations, e.g 10 000:

 

It is possible at this stage to add a constraint on the maximum number of times the matrix is applied to the data, i.e on the maximal number of campaigns a simulated journey is allowed to have.

Advantages

  • The dynamic journey is taken into account, as well as the transition between two states. The funnel is not assumed to be linear.
  • It is possile to build a conversion graph that maps the customer journey provides valuable insights.
  • It is possible to evaluate partly the accuracy of the Attribution Model based on Markov Chains. It is for example possible to see how well the transition matrix help predict the future by analysing the number of correct predictions at any given step over all sequences.

Disadvantages

  • It can be somewhat difficult to set the memory parameter. Complementarity effects between channels are not well taken into account if the memory is low, but a parameter too high will lead to over-sensitivity to noise in the data and be difficult to implement if customer journeys tend to have a number of campaigns below this memory parameter.
  • Long journeys with different channels involved will be overweighted, as they will count many times in the Removal Effect.  For example, if there are n-1 channels in the customer journey, this journey will be considered as failure for the n-1 channel-RE. If the volume effects (i.e the impact of the overall number of channels in a journey, irrelevant from their type° are important then results may be biased.

References:

R package: ChannelAttribution

Git:

https://github.com/MatCyt/Markov-Chain/blob/master/README.md

Course:

https://www.ssc.wisc.edu/~jmontgom/markovchains.pdf

Article:

“Mapping the Customer Journey: A Graph-Based Framework for Online Attribution Modeling”; Anderl, Eva and Becker, Ingo and Wangenheim, Florian V. and Schumann, Jan Hendrik, 2014. Available at SSRN: https://ssrn.com/abstract=2343077 or http://dx.doi.org/10.2139/ssrn.2343077

“Media Exposure through the Funnel: A Model of Multi-Stage Attribution”, Abhishek & al, 2012

“Multichannel Marketing Attribution Using Markov Chains”, Kakalejčík, L., Bucko, J., Resende, P.A.A. and Ferencova, M. Journal of Applied Management and Investments, Vol. 7 No. 1, pp. 49-60.  2018

Blogs:

https://analyzecore.com/2016/08/03/attribution-model-r-part-1

https://analyzecore.com/2016/08/03/attribution-model-r-part-2

                          3.3 To go further: Tackling selection biases with Quasi-Experiments

Exposure to certain types of advertisement is usually highly correlated to non-observable variables. Differences in the behaviour of users exposed to different campaigns may thus only be driven by core differences in converison probabilities between groups whether than by the campaign effect. These potential selection effects may bias the results obtained using historical data.

Quasi-Experiments can help correct this selection effect while still using available observationnal data.  These methods recreate the settings on a randomized setting. The goal is to come as close as possible to the ideal of comparing two populations that are identical in all respects except for the advertising exposure. However, populations might still differ with respect to some unobserved characteristics.

Common quasi-experimental methods used for instance in Public Policy Evaluation are:

  • Discontinuity Regressions
  • Matching Methods, such as Exact Matching,  Propensity-score matching or k-nearest neighbourghs.

References:

Article:

“Towards a digital Attribution Model: Measuring the impact of display advertising on online consumer behaviour”, Anindya Ghose & al, MIS Quarterly Vol. 40 No. 4, pp. 1-XX, 2016

https://pdfs.semanticscholar.org/4fa6/1c53f281fa63a9f0617fbd794d54911a2f84.pdf

        4. First Steps towards a Practical Implementation

Identify key points of interests

  • Identify the nature of touchpoints available: is the data based on clicks? If so, is there a way to complement the data with A/B tests to measure the influence of ads without clicks (display, video) ? For example, what happens to sales when display campaign is removed? Analysing this multiplier effect would give the overall responsibility of display on sales, to be deduced from current attribution values given to click-based channels. More interestingly, what is the impact of the removal of display campaign on the occurences of click-based campaigns ? This would give us an idea of the impact of display ads on the exposure to each other campaigns, which would help correct the attribution values more precisely at the campaign level.
  • Define the KPI to track. From a pure Marketing perspective, looking at purchases may be sufficient, but from a financial perspective looking at profits, though a bit more difficult to compute, may drive more interesting results.
  • Define a customer journey. It may seem obvious, but the notion needs to be clarified at first. Would it be defined by a time limit? If so, which one? Does it end when a conversion is observed? For example, if a customer makes 2 purchases, would the campaigns he’s been exposed to before the first order still be accounted for in the second order? If so, with a time decay?
  • Define the research framework: are we interested only in customer journeys which have led to conversions or in all journeys? Keep in mind that successful customer journeys are a non-representative sample of customer journeys. Models built on the analysis of biased samples may be conservative. Take an extreme example: 80% of customers who see campaign A buy the product, VS 1% for campaign B. However, campaign B exposure is great and 100 Million people see it VS only 1M for campaign A. An Attribution Model based on successful journeys will give higher credit to campaign B which is an auguable conclusion. Taking into account costs per campaign (in the case where costs are calculated by clicks) may of course tackle this issue partly, as campaign A could then exhibit higher returns, but a serious fallacious reasonning is at stake here.

Analyse the typical customer journey    

  • Performing a duration analysis on the data may help you improve the definition of the customer journey to be used by your organization. After which days are converison probabilities null? Should we consider the effect of campaigns disappears after x days without orders? For example, if 99% of orders are placed in the 30 days following a first click, it might be interesting to define the customer journey as a 30 days time frame following the first oder.
  • Look at the distribution of the number of campaigns in a typical journey. If you choose to calculate the effect of campaigns interraction in your Attribution Model, it may indeed help you determine the maximum number of campaigns to be included in a combination. Indeed, you may not need to assess the impact of channel combinations with above than 4 different channels if 95% of orders are placed after less then 4 campaigns.
  • Transition matrixes: what if a campaign A systematically leads to a campaign B? What happens if we remove A or B? These insights would give clues to ask precise questions for a latter AB test, for example to find out if there is complementarity between channels A and B – (implying none should be removed) or mere substitution (implying one can be given up).
  • If conversion rates are available: it can be interesting to perform a survival analysis i.e to analyse the likelihood of conversion based on duration since first click. This could help us excluse potential outliers or individuals who have very low conversion probabilities.

Summary

Attribution is a complex topic which will probably never be definitively solved. Indeed, a main issue is the difficulty, or even impossibility, to evaluate precisely the accuracy of the attribution model that we’ve built. Attribution Models should be seen as a good yet always improvable approximation of the incremental values of campaigns, and be presented with their intrinsinc limits and biases.

Sentiment Analysis of IMDB reviews

Sentiment Analysis of IMDB reviews

This article shows you how to build a Neural Network from scratch(no libraries) for the purpose of detecting whether a movie review on IMDB is negative or positive.

Outline:

  • Curating a dataset and developing a "Predictive Theory"

  • Transforming Text to Numbers Creating the Input/Output Data

  • Building our Neural Network

  • Making Learning Faster by Reducing "Neural Noise"

  • Reducing Noise by strategically reducing the vocabulary

Curating the Dataset

In [3]:
def pretty_print_review_and_label(i):
    print(labels[i] + "\t:\t" + reviews[i][:80] + "...")

g = open('reviews.txt','r') # features of our dataset
reviews = list(map(lambda x:x[:-1],g.readlines()))
g.close()

g = open('labels.txt','r') # labels
labels = list(map(lambda x:x[:-1].upper(),g.readlines()))
g.close()

Note: The data in reviews.txt we're contains only lower case characters. That's so we treat different variations of the same word, like The, the, and THE, all the same way.

It's always a good idea to get check out your dataset before you proceed.

In [2]:
len(reviews) #No. of reviews
Out[2]:
25000
In [3]:
reviews[0] #first review
Out[3]:
'bromwell high is a cartoon comedy . it ran at the same time as some other programs about school life  such as  teachers  . my   years in the teaching profession lead me to believe that bromwell high  s satire is much closer to reality than is  teachers  . the scramble to survive financially  the insightful students who can see right through their pathetic teachers  pomp  the pettiness of the whole situation  all remind me of the schools i knew and their students . when i saw the episode in which a student repeatedly tried to burn down the school  i immediately recalled . . . . . . . . . at . . . . . . . . . . high . a classic line inspector i  m here to sack one of your teachers . student welcome to bromwell high . i expect that many adults of my age think that bromwell high is far fetched . what a pity that it isn  t   '
In [4]:
labels[0] #first label
Out[4]:
'POSITIVE'

Developing a Predictive Theory

Analysing how you would go about predicting whether its a positive or a negative review.

In [5]:
print("labels.txt \t : \t reviews.txt\n")
pretty_print_review_and_label(2137)
pretty_print_review_and_label(12816)
pretty_print_review_and_label(6267)
pretty_print_review_and_label(21934)
pretty_print_review_and_label(5297)
pretty_print_review_and_label(4998)
labels.txt 	 : 	 reviews.txt

NEGATIVE	:	this movie is terrible but it has some good effects .  ...
POSITIVE	:	adrian pasdar is excellent is this film . he makes a fascinating woman .  ...
NEGATIVE	:	comment this movie is impossible . is terrible  very improbable  bad interpretat...
POSITIVE	:	excellent episode movie ala pulp fiction .  days   suicides . it doesnt get more...
NEGATIVE	:	if you haven  t seen this  it  s terrible . it is pure trash . i saw this about ...
POSITIVE	:	this schiffer guy is a real genius  the movie is of excellent quality and both e...
In [41]:
from collections import Counter
import numpy as np

We'll create three Counter objects, one for words from postive reviews, one for words from negative reviews, and one for all the words.

In [56]:
# Create three Counter objects to store positive, negative and total counts
positive_counts = Counter()
negative_counts = Counter()
total_counts = Counter()

Examine all the reviews. For each word in a positive review, increase the count for that word in both your positive counter and the total words counter; likewise, for each word in a negative review, increase the count for that word in both your negative counter and the total words counter. You should use split(' ') to divide a piece of text (such as a review) into individual words.

In [57]:
# Loop over all the words in all the reviews and increment the counts in the appropriate counter objects
for i in range(len(reviews)):
    if(labels[i] == 'POSITIVE'):
        for word in reviews[i].split(" "):
            positive_counts[word] += 1
            total_counts[word] += 1
    else:
        for word in reviews[i].split(" "):
            negative_counts[word] += 1
            total_counts[word] += 1

Most common positive & negative words

In [ ]:
positive_counts.most_common()

The above statement retrieves alot of words, the top 3 being : ('the', 173324), ('.', 159654), ('and', 89722),

In [ ]:
negative_counts.most_common()

The above statement retrieves alot of words, the top 3 being : ('', 561462), ('.', 167538), ('the', 163389),

As you can see, common words like "the" appear very often in both positive and negative reviews. Instead of finding the most common words in positive or negative reviews, what you really want are the words found in positive reviews more often than in negative reviews, and vice versa. To accomplish this, you'll need to calculate the ratios of word usage between positive and negative reviews.

The positive-to-negative ratio for a given word can be calculated with positive_counts[word] / float(negative_counts[word]+1). Notice the +1 in the denominator – that ensures we don't divide by zero for words that are only seen in positive reviews.

In [58]:
pos_neg_ratios = Counter()

# Calculate the ratios of positive and negative uses of the most common words
# Consider words to be "common" if they've been used at least 100 times
for term,cnt in list(total_counts.most_common()):
    if(cnt > 100):
        pos_neg_ratio = positive_counts[term] / float(negative_counts[term]+1)
        pos_neg_ratios[term] = pos_neg_ratio

Examine the ratios

In [12]:
print("Pos-to-neg ratio for 'the' = {}".format(pos_neg_ratios["the"]))
print("Pos-to-neg ratio for 'amazing' = {}".format(pos_neg_ratios["amazing"]))
print("Pos-to-neg ratio for 'terrible' = {}".format(pos_neg_ratios["terrible"]))
Pos-to-neg ratio for 'the' = 1.0607993145235326
Pos-to-neg ratio for 'amazing' = 4.022813688212928
Pos-to-neg ratio for 'terrible' = 0.17744252873563218

We see the following:

  • Words that you would expect to see more often in positive reviews – like "amazing" – have a ratio greater than 1. The more skewed a word is toward postive, the farther from 1 its positive-to-negative ratio will be.
  • Words that you would expect to see more often in negative reviews – like "terrible" – have positive values that are less than 1. The more skewed a word is toward negative, the closer to zero its positive-to-negative ratio will be.
  • Neutral words, which don't really convey any sentiment because you would expect to see them in all sorts of reviews – like "the" – have values very close to 1. A perfectly neutral word – one that was used in exactly the same number of positive reviews as negative reviews – would be almost exactly 1.

Ok, the ratios tell us which words are used more often in postive or negative reviews, but the specific values we've calculated are a bit difficult to work with. A very positive word like "amazing" has a value above 4, whereas a very negative word like "terrible" has a value around 0.18. Those values aren't easy to compare for a couple of reasons:

  • Right now, 1 is considered neutral, but the absolute value of the postive-to-negative rations of very postive words is larger than the absolute value of the ratios for the very negative words. So there is no way to directly compare two numbers and see if one word conveys the same magnitude of positive sentiment as another word conveys negative sentiment. So we should center all the values around netural so the absolute value fro neutral of the postive-to-negative ratio for a word would indicate how much sentiment (positive or negative) that word conveys.
  • When comparing absolute values it's easier to do that around zero than one.

To fix these issues, we'll convert all of our ratios to new values using logarithms (i.e. use np.log(ratio))

In the end, extremely positive and extremely negative words will have positive-to-negative ratios with similar magnitudes but opposite signs.

In [59]:
# Convert ratios to logs
for word,ratio in pos_neg_ratios.most_common():
    pos_neg_ratios[word] = np.log(ratio)

Examine the new ratios

In [14]:
print("Pos-to-neg ratio for 'the' = {}".format(pos_neg_ratios["the"]))
print("Pos-to-neg ratio for 'amazing' = {}".format(pos_neg_ratios["amazing"]))
print("Pos-to-neg ratio for 'terrible' = {}".format(pos_neg_ratios["terrible"]))
Pos-to-neg ratio for 'the' = 0.05902269426102881
Pos-to-neg ratio for 'amazing' = 1.3919815802404802
Pos-to-neg ratio for 'terrible' = -1.7291085042663878

If everything worked, now you should see neutral words with values close to zero. In this case, "the" is near zero but slightly positive, so it was probably used in more positive reviews than negative reviews. But look at "amazing"'s ratio - it's above 1, showing it is clearly a word with positive sentiment. And "terrible" has a similar score, but in the opposite direction, so it's below -1. It's now clear that both of these words are associated with specific, opposing sentiments.

Run the below code to see more ratios.

It displays all the words, ordered by how associated they are with postive reviews.

In [ ]:
pos_neg_ratios.most_common()

The top most common words for the above code : ('edie', 4.6913478822291435), ('paulie', 4.0775374439057197), ('felix', 3.1527360223636558), ('polanski', 2.8233610476132043), ('matthau', 2.8067217286092401), ('victoria', 2.6810215287142909), ('mildred', 2.6026896854443837), ('gandhi', 2.5389738710582761), ('flawless', 2.451005098112319), ('superbly', 2.2600254785752498), ('perfection', 2.1594842493533721), ('astaire', 2.1400661634962708), ('captures', 2.0386195471595809), ('voight', 2.0301704926730531), ('wonderfully', 2.0218960560332353), ('powell', 1.9783454248084671), ('brosnan', 1.9547990964725592)

Transforming Text into Numbers

Creating the Input/Output Data

Create a set named vocab that contains every word in the vocabulary.

In [19]:
vocab = set(total_counts.keys())

Check vocabulary size

In [20]:
vocab_size = len(vocab)
print(vocab_size)
74074

Th following image rpresents the layers of the neural network you'll be building throughout this notebook. layer_0 is the input layer, layer_1 is a hidden layer, and layer_2 is the output layer.

In [1]:
 
Out[1]:

TODO: Create a numpy array called layer_0 and initialize it to all zeros. Create layer_0 as a 2-dimensional matrix with 1 row and vocab_size columns.

In [21]:
layer_0 = np.zeros((1,vocab_size))

layer_0 contains one entry for every word in the vocabulary, as shown in the above image. We need to make sure we know the index of each word, so run the following cell to create a lookup table that stores the index of every word.

TODO: Complete the implementation of update_input_layer. It should count how many times each word is used in the given review, and then store those counts at the appropriate indices inside layer_0.

In [ ]:
# Create a dictionary of words in the vocabulary mapped to index positions 
# (to be used in layer_0)
word2index = {}
for i,word in enumerate(vocab):
    word2index[word] = i

It stores the indexes like this: 'antony': 22, 'pinjar': 23, 'helsig': 24, 'dances': 25, 'good': 26, 'willard': 71500, 'faridany': 27, 'foment': 28, 'matts': 12313,

Lets implement some functions for simplifying our inputs to the neural network.

In [25]:
def update_input_layer(review):
    """
    The element at a given index of layer_0 should represent
    how many times the given word occurs in the review.
    """
     
    global layer_0
    
    # clear out previous state, reset the layer to be all 0s
    layer_0 *= 0
    
    # count how many times each word is used in the given review and store the results in layer_0 
    for word in review.split(" "):
        layer_0[0][word2index[word]] += 1

Run the following cell to test updating the input layer with the first review. The indices assigned may not be the same as in the solution, but hopefully you'll see some non-zero values in layer_0.

In [26]:
update_input_layer(reviews[0])
layer_0
Out[26]:
array([[ 18.,   0.,   0., ...,   0.,   0.,   0.]])

get_target_for_labels should return 0 or 1, depending on whether the given label is NEGATIVE or POSITIVE, respectively.

In [27]:
def get_target_for_label(label):
    if(label == 'POSITIVE'):
        return 1
    else:
        return 0

Building a Neural Network

In [32]:
import time
import sys
import numpy as np

# Encapsulate our neural network in a class
class SentimentNetwork:
    def __init__(self, reviews,labels,hidden_nodes = 10, learning_rate = 0.1):
        """
        Args:
            reviews(list) - List of reviews used for training
            labels(list) - List of POSITIVE/NEGATIVE labels
            hidden_nodes(int) - Number of nodes to create in the hidden layer
            learning_rate(float) - Learning rate to use while training
        
        """
        # Assign a seed to our random number generator to ensure we get
        # reproducable results
        np.random.seed(1)

        # process the reviews and their associated labels so that everything
        # is ready for training
        self.pre_process_data(reviews, labels)
        
        # Build the network to have the number of hidden nodes and the learning rate that
        # were passed into this initializer. Make the same number of input nodes as
        # there are vocabulary words and create a single output node.
        self.init_network(len(self.review_vocab),hidden_nodes, 1, learning_rate)

    def pre_process_data(self, reviews, labels):
        
        # populate review_vocab with all of the words in the given reviews
        review_vocab = set()
        for review in reviews:
            for word in review.split(" "):
                review_vocab.add(word)

        # Convert the vocabulary set to a list so we can access words via indices
        self.review_vocab = list(review_vocab)
        
        # populate label_vocab with all of the words in the given labels.
        label_vocab = set()
        for label in labels:
            label_vocab.add(label)
        
        # Convert the label vocabulary set to a list so we can access labels via indices
        self.label_vocab = list(label_vocab)
        
        # Store the sizes of the review and label vocabularies.
        self.review_vocab_size = len(self.review_vocab)
        self.label_vocab_size = len(self.label_vocab)
        
        # Create a dictionary of words in the vocabulary mapped to index positions
        self.word2index = {}
        for i, word in enumerate(self.review_vocab):
            self.word2index[word] = i
        
        # Create a dictionary of labels mapped to index positions
        self.label2index = {}
        for i, label in enumerate(self.label_vocab):
            self.label2index[label] = i
        
    def init_network(self, input_nodes, hidden_nodes, output_nodes, learning_rate):
        # Set number of nodes in input, hidden and output layers.
        self.input_nodes = input_nodes
        self.hidden_nodes = hidden_nodes
        self.output_nodes = output_nodes

        # Store the learning rate
        self.learning_rate = learning_rate

        # Initialize weights

        # These are the weights between the input layer and the hidden layer.
        self.weights_0_1 = np.zeros((self.input_nodes,self.hidden_nodes))
    
        # These are the weights between the hidden layer and the output layer.
        self.weights_1_2 = np.random.normal(0.0, self.output_nodes**-0.5, 
                                                (self.hidden_nodes, self.output_nodes))
        
        # The input layer, a two-dimensional matrix with shape 1 x input_nodes
        self.layer_0 = np.zeros((1,input_nodes))
    
    def update_input_layer(self,review):

        # clear out previous state, reset the layer to be all 0s
        self.layer_0 *= 0
        
        for word in review.split(" "):
            if(word in self.word2index.keys()):
                self.layer_0[0][self.word2index[word]] += 1
                
    def get_target_for_label(self,label):
        if(label == 'POSITIVE'):
            return 1
        else:
            return 0
        
    def sigmoid(self,x):
        return 1 / (1 + np.exp(-x))
    
    def sigmoid_output_2_derivative(self,output):
        return output * (1 - output)
    
    def train(self, training_reviews, training_labels):
        
        # make sure out we have a matching number of reviews and labels
        assert(len(training_reviews) == len(training_labels))
        
        # Keep track of correct predictions to display accuracy during training 
        correct_so_far = 0

        # Remember when we started for printing time statistics
        start = time.time()
        
        # loop through all the given reviews and run a forward and backward pass,
        # updating weights for every item
        for i in range(len(training_reviews)):
            
            # Get the next review and its correct label
            review = training_reviews[i]
            label = training_labels[i]
            
            ### Forward pass ###

            # Input Layer
            self.update_input_layer(review)

            # Hidden layer
            layer_1 = self.layer_0.dot(self.weights_0_1)

            # Output layer
            layer_2 = self.sigmoid(layer_1.dot(self.weights_1_2))
            
            ### Backward pass ###

            # Output error
            layer_2_error = layer_2 - self.get_target_for_label(label) # Output layer error is the difference between desired target and actual output.
            layer_2_delta = layer_2_error * self.sigmoid_output_2_derivative(layer_2)

            # Backpropagated error
            layer_1_error = layer_2_delta.dot(self.weights_1_2.T) # errors propagated to the hidden layer
            layer_1_delta = layer_1_error # hidden layer gradients - no nonlinearity so it's the same as the error

            # Update the weights
            self.weights_1_2 -= layer_1.T.dot(layer_2_delta) * self.learning_rate # update hidden-to-output weights with gradient descent step
            self.weights_0_1 -= self.layer_0.T.dot(layer_1_delta) * self.learning_rate # update input-to-hidden weights with gradient descent step

            # Keep track of correct predictions.
            if(layer_2 >= 0.5 and label == 'POSITIVE'):
                correct_so_far += 1
            elif(layer_2 < 0.5 and label == 'NEGATIVE'):
                correct_so_far += 1
            
            sys.stdout.write(" #Correct:" + str(correct_so_far) + " #Trained:" + str(i+1) \
                             + " Training Accuracy:" + str(correct_so_far * 100 / float(i+1))[:4] + "%")
    
    def test(self, testing_reviews, testing_labels):
        """
        Attempts to predict the labels for the given testing_reviews,
        and uses the test_labels to calculate the accuracy of those predictions.
        """
        
        # keep track of how many correct predictions we make
        correct = 0

        # Loop through each of the given reviews and call run to predict
        # its label. 
        for i in range(len(testing_reviews)):
            pred = self.run(testing_reviews[i])
            if(pred == testing_labels[i]):
                correct += 1
            
            sys.stdout.write(" #Correct:" + str(correct) + " #Tested:" + str(i+1) \
                             + " Testing Accuracy:" + str(correct * 100 / float(i+1))[:4] + "%")
    
    def run(self, review):
        """
        Returns a POSITIVE or NEGATIVE prediction for the given review.
        """
        # Run a forward pass through the network, like in the "train" function.
        
        # Input Layer
        self.update_input_layer(review.lower())

        # Hidden layer
        layer_1 = self.layer_0.dot(self.weights_0_1)

        # Output layer
        layer_2 = self.sigmoid(layer_1.dot(self.weights_1_2))
        
        # Return POSITIVE for values above greater-than-or-equal-to 0.5 in the output layer;
        # return NEGATIVE for other values
        if(layer_2[0] >= 0.5):
            return "POSITIVE"
        else:
            return "NEGATIVE"
        

Run the following code to create the network with a small learning rate, 0.001, and then train the new network. Using learning rate larger than this, for example 0.1 or even 0.01 would result in poor performance.

In [ ]:
mlp = SentimentNetwork(reviews[:-1000],labels[:-1000], learning_rate=0.001)
mlp.train(reviews[:-1000],labels[:-1000])

Running the above code would have given an accuracy around 62.2%

Reducing Noise in Our Input Data

Counting how many times each word occured in our review might not be the most efficient way. Instead just including whether a word was there or not will improve our training time and accuracy. Hence we update our update_input_layer() function.

In [ ]:
def update_input_layer(self,review):
    self.layer_0 *= 0
        
    for word in review.split(" "):
        if(word in self.word2index.keys()):
            self.layer_0[0][self.word2index[word]] =1

Creating and running our neural network again, even with a higher learning rate of 0.1 gave us a training accuracy of 83.8% and testing accuracy(testing on last 1000 reviews) of 85.7%.

Reducing Noise by Strategically Reducing the Vocabulary

Let us put the pos to neg ratio's that we found were much more effective at detecting a positive or negative label. We could do that by a few change:

  • Modify pre_process_data:
    • Add two additional parameters: min_count and polarity_cutoff
    • Calculate the positive-to-negative ratios of words used in the reviews.
    • Change so words are only added to the vocabulary if they occur in the vocabulary more than min_count times.
    • Change so words are only added to the vocabulary if the absolute value of their postive-to-negative ratio is at least polarity_cutoff
In [ ]:
def pre_process_data(self, reviews, labels, polarity_cutoff, min_count):
        
        positive_counts = Counter()
        negative_counts = Counter()
        total_counts = Counter()

        for i in range(len(reviews)):
            if(labels[i] == 'POSITIVE'):
                for word in reviews[i].split(" "):
                    positive_counts[word] += 1
                    total_counts[word] += 1
            else:
                for word in reviews[i].split(" "):
                    negative_counts[word] += 1
                    total_counts[word] += 1

        pos_neg_ratios = Counter()

        for term,cnt in list(total_counts.most_common()):
            if(cnt >= 50):
                pos_neg_ratio = positive_counts[term] / float(negative_counts[term]+1)
                pos_neg_ratios[term] = pos_neg_ratio

        for word,ratio in pos_neg_ratios.most_common():
            if(ratio > 1):
                pos_neg_ratios[word] = np.log(ratio)
            else:
                pos_neg_ratios[word] = -np.log((1 / (ratio + 0.01)))

        # populate review_vocab with all of the words in the given reviews
        review_vocab = set()
        for review in reviews:
            for word in review.split(" "):
                if(total_counts[word] > min_count):
                    if(word in pos_neg_ratios.keys()):
                        if((pos_neg_ratios[word] >= polarity_cutoff) or (pos_neg_ratios[word] <= -polarity_cutoff)):
                            review_vocab.add(word)
                    else:
                        review_vocab.add(word)

        # Convert the vocabulary set to a list so we can access words via indices
        self.review_vocab = list(review_vocab)
        
        # populate label_vocab with all of the words in the given labels.
        label_vocab = set()
        for label in labels:
            label_vocab.add(label)
        
        # Convert the label vocabulary set to a list so we can access labels via indices
        self.label_vocab = list(label_vocab)
        
        # Store the sizes of the review and label vocabularies.
        self.review_vocab_size = len(self.review_vocab)
        self.label_vocab_size = len(self.label_vocab)
        
        # Create a dictionary of words in the vocabulary mapped to index positions
        self.word2index = {}
        for i, word in enumerate(self.review_vocab):
            self.word2index[word] = i
        
        # Create a dictionary of labels mapped to index positions
        self.label2index = {}
        for i, label in enumerate(self.label_vocab):
            self.label2index[label] = i

Our training accuracy increased to 85.6% after this change. As we can see our accuracy saw a huge jump by making minor changes based on our intuition. We can keep making such changes and increase the accuracy even further.

 

Download the Data Sources

The data sources used in this article can be downloaded here:

Interview – Die Herausforderungen der Sensor-Datenanalyse für die Automobilindustrie

Interview mit Andreas Festl von VIRTUAL VEHICLE

Andreas Festl ist Data Scientist bei VIRTUAL VEHICLE, ein führendes F&E Zentrum für die Automobil- und Bahnindustrie mit Sitz in Graz, Österreich. Das Zentrum konzentriert sich auf die konsequente Virtualisierung der Fahrzeugentwicklung. Wesentliches Element dabei ist die Verknüpfung von numerischer Simulation und Hardware-Testen, welche ein umfassendes HW-SW Systemdesign sicherstellt. Herr Festl forscht dort an Kontext-basierten Informationssystemen für den Einsatz im Fahrzeug und in der Entwicklung. Er ist ausgebildeter Mathematiker, der sich schon früh dem Thema Data Science verschrieben hat. Zusätzlich ist Herr Festl in der Lehre für Data and Information Science an der Fachhochschule Joanneum tätig.

Data Science Blog: Herr Festl, Sie sind technischer Data Scientist und arbeiten mit Daten, die zum großen Teil von Maschinen generiert werden. Was unterscheidet Ihren Arbeitsalltag vermutlich von den Data Scientists, die sich mit geschäftlichen Daten befassen?

Das wesentliche Merkmal an den Daten, mit denen wir arbeiten, ist die nicht vernachlässigbare zeitliche Komponente. Stellen Sie sich zum Beispiel eine Messung der Fahrzeuggeschwindigkeit vor: Dieses Messsignal kann natürlich nur dann sinnvoll interpretiert und verarbeitet werden, wenn die Zeit mitberücksichtigt wird. Die bloße Kenntnis der einzelnen Geschwindigkeitswerte hilft Ihnen ohne die korrekte Abfolge nicht weiter. Das führt dazu, dass viele Algorithmen aus dem Bereich des maschinellen Lernens nicht direkt auf diesen Daten arbeiten können.

Es existieren hier natürlich dennoch viele Möglichkeiten und Ansätze dafür, Wissen aus den Daten zu gewinnen; diese werden jedoch scheinbar noch nicht so oft verwendet, weshalb die verfügbare Software meist nicht für industrielle, sondern für akademische Nutzer ausgelegt ist. Ein wesentlicher Teil meiner Arbeit besteht deshalb darin, die passenden Libraries zu finden und diese für unsere Use-Cases anzupassen oder die Methode neu zu implementieren. Es gibt durchaus immer wieder Zeiten in denen meine Job-Beschreibung „mathematischer Programmierer“ lauten sollte und nicht “Data Scientist“. Ich denke, das ist im klassischen Bereich, der sich geschäftlichen Daten beschäftigt, vielleicht nicht mehr so häufig, da dort die verfügbare Software schon sehr ausgreift ist.

Außerdem beschreiben unsere Daten oft komplexe technische Prozesse in Fahrzeugkomponenten. Hier ist eine rege Kommunikation mit den jeweiligen Domänenexperten unerlässlich, damit ich auch als fachfremder Data Scientist den Prozess, der die Daten erzeugt, zumindest in Grundzügen verstehen kann. Dieser kommunikative Teil, in dem man sehr viel über verschiedenste Fachbereiche erfährt, ist für mich einer der schönsten Aspekte meiner Arbeit.

Data Science Blog: Wenn Data Science einem Laien erklärt wird, kommen häufig Beispiele von Kaufempfehlungen oder Gesundheitsprognosen von Fitness-Apps zur Sprache. Welches Beispiel würden Sie im Kontext von Automotive verwenden?

Die Möglichkeiten für den Einsatz von Data Science im Automotive Bereich sind extrem vielfältig – sie kann eigentlich über den gesamten Lebenszyklus eines Fahrzeugs gewinnbringend eingesetzt werden. Ein Einsatzbeispiel, das der Fahrer direkt positiv erleben kann, wäre die Predictive Maintenance von Fahrzeugteilen. Ähnlich zu den von Ihnen angesprochenen Fitness-Apps geht es hier darum eine „Gesundheitsprognose“ für die einzelnen Fahrzeugteile anhand von Messwerten zu erstellen. Im Idealfall müssen Sie Ihr Auto dann nicht mehr in fixen Service-Intervallen in die Werkstatt stellen, sondern das Auto meldet sich automatisch kurz bevor ein Teil ausgetauscht werden muss. Diese Meldung erschiene dann deshalb, weil die Messwerte darauf schließen lassen, dass es bald zu einem Defekt kommen wird und nicht einfach nach einem fixen, vorher definierten Zeitraum. Heute werden ja Teile oft einfach deswegen ausgetauscht, weil es der Wartungsplan so vorsieht – unabhängig von ihrer tatsächlichen Abnutzung.

Data Science Blog: Was sind denn gegenwärtig besonders interessante Anwendungsfälle und an welchen arbeiten Sie für die Zukunft?

Aus Sicht der Anwendung finde ich es besonders spannend durch Sensor-Signale auf Eigenschaften des Fahrers zu schließen. Die Methodik dazu entwickeln wir gerade in aktuellen Projekten. Es ist zum Beispiel durchaus denkbar, sicherheitsrelevante Ereignisse und Fahrmanöver zu identifizieren. Diese Informationen können dann vielseitig verwendet werden. Einige Beispiele dazu: Verkehrsplaner könnten damit automatisiert besonders gefährliche Kreuzungen angezeigt bekommen, Versicherer könnten ihren Kunden auf das individuelle Risikoverhalten abgestimmte Produkte anbieten oder Kunden könnten sich Ihren Taxifahrer über eine App nach seinem Fahrstil aussuchen. Denkbar wäre auch eine Diebstahlsicherung: Das Fahrzeug erkennt über den Fahrstil, dass es von einer unbefugten Person benutzt wird und löst daraufhin einen Alarm aus. Hier eröffnen sich viele Möglichkeiten.

Aus Sicht der Datenanalyse finde ich es besonders interessant, Algorithmen, die für ganz andere Aufgabenstellung entwickelt wurden, auf Probleme aus dem Automotive-Bereich anzuwenden. In einem unserer Projekte analysieren wir beispielsweise Software-Logfiles von Prüfständen und verwenden dazu Association Rules (eine Technik aus der Warenkorbanalyse) und Methoden, die normalerweise für das Untersuchen von Interaktionen in sozialen Netzwerken verwendet werden. Dass diese Übertragbarkeit gegeben ist finde ich extrem spannend.

Data Science Blog: Über welche Datenquellen verfügen Sie? Gibt es auch fahrzeugexterne Datenquellen, die sinnvoll sein könnten?

Da sprechen Sie natürlichen einen kritischen Punkt in jedem Data Science Projekt an: Ohne Daten geht nichts. Zusätzlich müssen die verwendeten Daten eine gewisse Qualität aufweisen und natürlich mit dem zu lösenden Problem in möglichst direktem Zusammenhang stehen.

Welche Datenquellen wir genau verwenden, hängt natürlich sehr stark vom konkretem Projekt ab. In industrienahen Projekten werden die Daten in der Regel vom Industriepartner bereitgestellt. Das kann dann alles Mögliche sein: Messungen von Prüfständen, Fertigungs-Protokolle, Wartungsdaten und vieles mehr.

Diese „Industrie-Daten“ unterliegen dann aber üblicherweise einer strengen Geheimhaltung und dürfen nicht in anderen Projekten verwendet werden. Deshalb haben wir im Unternehmen einen eigenen Datenlogger entwickelt, mit dem wir selber Daten aufnehmen können, die dann uns gehören. Diese Daten verwenden wir hauptsächlich in forschungsnahen Projekten, in denen die Ergebnisse publiziert werden sollen.

Fahrzeugexterne Datenquellen sind definitiv sinnvoll und werden immer mehr mit den klassischen Sensor-Daten fusioniert; oft ergibt sich dann durch eine Kombination von proprietären und offen verfügbaren Daten ein großer Mehrwert. In der vorhin angesprochenen Erkennung von sicherheitsrelevanten Ergebnissen spielt zum Beispiel das Wetter eine wesentliche Rolle: Eine zu schnell gefahrene Kurve ist bei Nässe oder Glätte deutlich gefährlicher als auf trockener Fahrbahn. Generell werden Daten über Umwelt und Infrastruktur immer wichtiger. Praktisch jeder fahrerzentrierte Dienst benötigt sie. Denken Sie zum Beispiel an Google Maps, das bereits heute die Bewegungsdaten von vielen Verkehrsteilnehmern gemeinsam analysiert um Vorhersagen über die Verkehrsdichte und damit über die optimale Route zu treffen.

Data Science Blog: Wie aufwändig gestaltet sich das Data Engineering, also die Datenbereitstellung und -zusammenführung?

Das ist definitiv ein schwieriges Unterfangen. Gerade Sensordaten erreichen schnell eine beachtliche Größe, die den Einsatz eines Big Data Technologie-Stacks erforderlich macht. Hier macht uns aber wieder die bereits angesprochene zeitliche Komponente unserer Daten zu schaffen. Die meisten Big Data Technologien skalieren ja, indem sie die Datenpunkte mehr oder weniger zufällig auf mehrere Rechner verteilen. Das ist bei unseren Daten aber nicht zulässig, die Reihenfolge der Daten ist hochrelevant! Hier müssen wir also entweder auf einer anderen Ebene parallelisieren oder Technologie mit spezieller Funktionalität für Zeitreihen verwenden.

Data Science Blog: Welche Technologien setzen Sie für die Datenbereitstellung und -analyse ein? Was halten Sie vom Einsatz von Open Source Software?

Wir implementieren unsere Analysen meist in R oder Python, manchmal kommen auch Matlab oder C# (letzteres meist für User Interfaces) zum Einsatz. Für Big Data Analysen verwenden wir meist Apache Spark über die R und Python APIs. Für die Datenablage und Bereitstellung verwenden wir hauptsächlich PostgreSQL mit Timescale Erweiterung, InfluxDB sowie Apache Hadoop. Grundsätzlich sind wir jedoch nicht auf bestimmte Technologien fixiert, sondern versuchen immer das jeweils beste Tool für den jeweiligen Einsatzzweck zu verwenden.

Ich finde es spricht nichts gegen den Einsatz von Open Source Software – wie Sie ja auch an unserem Technologie-Stack erkennen können. Ich habe aber auch nichts gegen Closed Source Software – es gibt in beiden Bereichen genug gute und schlechte Software. Worauf ich aber achte, ist keine neue Technologie zu verwenden, hinter der ein zu kleines Entwicklerteam oder gar nur ein einzelner Entwickler steht. Hier ist mir die Gefahr zu groß, dass die Entwicklung bald eingestellt wird und die Ergebnisse meiner Analysen nicht mehr nachvollziehbar sind.

Data Science Blog: Zum Abschluss noch eine Frage von jungen Nachwuchskräften, die davon träumen, eine Karriere als Data Scientist im Ingenieurwesen zu machen: Welche Voraussetzungen bzw. Eigenschaften sollte ein Data Scientist in Ihrem Bereich mitbringen?

Neben einer fundierten fachlichen Ausbildung sind Neugier und der Wille, Zusammenhänge zu verstehen, Eigenschaften, die für jeden Data Scientist sehr wichtig sind. Zusätzlich hilft es durchaus eine kommunikative Persönlichkeit zu sein: Es gilt in Workshops die richtigen Informationen über die Daten einzuholen – das ist nicht immer ganz leicht. Zusätzlich müssen natürlich regelmäßig die Resultate der jeweiligen Analysen einem oft fachfremden Publikum präsentiert werden.

Applying Data Science Techniques in Python to Evaluate Ionospheric Perturbations from Earthquakes

Multi-GNSS (Galileo, GPS, and GLONASS) Vertical Total Electron Content Estimates: Applying Data Science techniques in Python to Evaluate Ionospheric Perturbations from Earthquakes

1 Introduction

Today, Global Navigation Satellite System (GNSS) observations are routinely used to study the physical processes that occur within the Earth’s upper atmosphere. Due to the experienced satellite signal propagation effects the total electron content (TEC) in the ionosphere can be estimated and the derived Global Ionosphere Maps (GIMs) provide an important contribution to monitoring space weather. While large TEC variations are mainly associated with solar activity, small ionospheric perturbations can also be induced by physical processes such as acoustic, gravity and Rayleigh waves, often generated by large earthquakes.

In this study Ionospheric perturbations caused by four earthquake events have been observed and are subsequently used as case studies in order to validate an in-house software developed using the Python programming language. The Python libraries primarily utlised are Pandas, Scikit-Learn, Matplotlib, SciPy, NumPy, Basemap, and ObsPy. A combination of Machine Learning and Data Analysis techniques have been applied. This in-house software can parse both receiver independent exchange format (RINEX) versions 2 and 3 raw data, with particular emphasis on multi-GNSS observables from GPS, GLONASS and Galileo. BDS (BeiDou) compatibility is to be added in the near future.

Several case studies focus on four recent earthquakes measuring above a moment magnitude (MW) of 7.0 and include: the 11 March 2011 MW 9.1 Tohoku, Japan, earthquake that also generated a tsunami; the 17 November 2013 MW 7.8 South Scotia Ridge Transform (SSRT), Scotia Sea earthquake; the 19 August 2016 MW 7.4 North Scotia Ridge Transform (NSRT) earthquake; and the 13 November 2016 MW 7.8 Kaikoura, New Zealand, earthquake.

Ionospheric disturbances generated by all four earthquakes have been observed by looking at the estimated vertical TEC (VTEC) and residual VTEC values. The results generated from these case studies are similar to those of published studies and validate the integrity of the in-house software.

2 Data Cleaning and Data Processing Methodology

Determining the absolute VTEC values are useful in order to understand the background ionospheric conditions when looking at the TEC perturbations, however small-scale variations in electron density are of primary interest. Quality checking processed GNSS data, applying carrier phase leveling to the measurements, and comparing the TEC perturbations with a polynomial fit creating residual plots are discussed in this section.

Time delay and phase advance observables can be measured from dual-frequency GNSS receivers to produce TEC data. Using data retrieved from the Center of Orbit Determination in Europe (CODE) site (ftp://ftp.unibe.ch/aiub/CODE), the differential code biases are subtracted from the ionospheric observables.

2.1 Determining VTEC: Thin Shell Mapping Function

The ionospheric shell height, H, used in ionosphere modeling has been open to debate for many years and typically ranges from 300 – 400 km, which corresponds to the maximum electron density within the ionosphere. The mapping function compensates for the increased path length traversed by the signal within the ionosphere. Figure 1 demonstrates the impact of varying the IPP height on the TEC values.

Figure 1 Impact on TEC values from varying IPP heights. The height of the thin shell, H, is increased in 50km increments from 300 to 500 km.

2.2 Phase Smoothing

For dual-frequency GNSS users TEC values can be retrieved with the use of dual-frequency measurements by applying calculations. Calculation of TEC for pseudorange measurements in practice produces a noisy outcome and so the relative phase delay between two carrier frequencies – which produces a more precise representation of TEC fluctuations – is preferred. To circumvent the effect of pseudorange noise on TEC data, GNSS pseudorange measurements can be smoothed by carrier phase measurements, with the use of the carrier phase smoothing technique, which is often referred to as carrier phase leveling.

Figure 2 Phase smoothed code differential delay

2.3 Residual Determination

For the purpose of this study the monitoring of small-scale variations in ionospheric electron density from the ionospheric observables are of particular interest. Longer period variations can be associated with diurnal alterations, and changes in the receiver- satellite elevation angles. In order to remove these longer period variations in the TEC time series as well as to monitor more closely the small-scale variations in ionospheric electron density, a higher-order polynomial is fitted to the TEC time series. This higher-order polynomial fit is then subtracted from the observed TEC values resulting in the residuals. The variation of TEC due to the TID perturbation are thus represented by the residuals. For this report the polynomial order applied was typically greater than 4, and was chosen to emulate the nature of the arc for that particular time series. The order number selected is dependent on the nature of arcs displayed upon calculating the VTEC values after an initial inspection of the VTEC plots.

3 Results

3.1 Tohoku Earthquake

For this particular report, the sampled data focused on what was retrieved from the IGS station, MIZU, located at Mizusawa, Japan. The MIZU site is 39N 08′ 06.61″ and 141E 07′ 58.18″. The location of the data collection site, MIZU, and the earthquake epicenter can be seen in Figure 3.

Figure 3 MIZU IGS station and Tohoku earthquake epicenter [generated using the Python library, Basemap]

Figure 4 displays the ionospheric delay in terms of vertical TEC (VTEC), in units of TECU (1 TECU = 1016 el m-2). The plot is split into two smaller subplots, the upper section displaying the ionospheric delay (VTEC) in units of TECU, the lower displaying the residuals. The vertical grey-dashed lined corresponds to the epoch of the earthquake at 05:46:23 UT (2:46:23 PM local time) on March 11 2011. In the upper section of the plot, the blue line corresponds to the absolute VTEC value calculated from the observations, in this case L1 and L2 on GPS, whereby the carrier phase leveling technique was applied to the data set. The VTEC values are mapped from the STEC values which are calculated from the LOS between MIZU and the GPS satellite PRN18 (on Figure 4 denoted G18). For this particular data set as seen in Figure 4, a polynomial fit of  five degrees was applied, which corresponds to the red-dashed line. As an alternative to polynomial fitting, band-pass filtering can be employed when TEC perturbations are desired. However for the scope of this report polynomial fitting to the time series of TEC data was the only method used. In the lower section of Figure 4 the residuals are plotted. The residuals are simply the phase smoothed delay values (the blue line) minus the polynomial fit line (the red-dashed line). All ionosphere delay plots follow the same layout pattern and all time data is represented in UT (UT = GPS – 15 leap seconds, whereby 15 leap seconds correspond to the amount of leap seconds at the time of the seismic event). The time series shown for the ionosphere delay plots are given in terms of decimal of the hour, so that the format follows hh.hh.

Figure 4 VTEC and residual plot for G18 at MIZU on March 11 2011

3.2 South Georgia Earthquake

In the South Georgia Island region located in the North Scotia Ridge Transform (NSRT) plate boundary between the South American and Scotia plates on 19 August 2016, a magnitude of 7.4 MW earthquake struck at 7:32:22 UT. This subsection analyses the data retrieved from KEPA and KRSA. As well as computing the GPS and GLONASS TEC values, four Galileo satellites (E08, E14, E26, E28) are also analysed. Figure 5 demonstrates the TEC perturbations as computed for the Galileo L1 and L5 carrier frequencies.

Figure 5 VTEC and residual plots at KRSA on 19 August 2016. The plots are from the perspective of the GNSS receiver at KRSA, for four Galileo satellites (a) E08; (b) E14; (c) E24; (d) E26. The y-axes and x-axes in all plots do not conform with one another but are adjusted to fit the data. The y-axes for the residual section of each plot is consistent with one another.

Figure 6 Geometry of the Galileo (E08, E14, E24 and E26) satellites’ projected ground track whereby the IPP is set to 300km altitude. The orange lines correspond to tectonic plate boundaries.

4 Conclusion

The proximity of the MIZU site and magnitude of the Tohoku event has provided a remarkable – albeit a poignant – opportunity to analyse the ocean-ionospheric coupling aftermath of a deep submarine seismic event. The Tohoku event has also enabled the observation of the origin and nature of the TIDs generated by both a major earthquake and tsunami in close proximity to the epicenter. Further, the Python software developed is more than capable of providing this functionality, by drawing on its mathematical packages, such as NumPy, Pandas, SciPy, and Matplotlib, as well as employing the cartographic toolkit provided from the Basemap package, and finally by utilizing the focal mechanism generation library, Obspy.

Pre-seismic cursors have been investigated in the past and strongly advocated in particular by Kosuke Heki. The topic of pre-seismic ionospheric disturbances remains somewhat controversial. A potential future study area could be the utilization of the Python program – along with algorithmic amendments – to verify the existence of this phenomenon. Such work would heavily involve the use of Scikit-Learn in order to ascertain the existence of any pre-cursors.

Finally, the code developed is still retained privately and as of yet not launched to any particular platform, such as GitHub. More detailed information on this report can be obtained here:

Download as PDF

The importance of domain knowledge – A healthcare data science perspective

Data scientists have (and need) many skills. They are frequently either former academic researchers or software engineers, with knowledge and skills in statistics, programming, machine learning, and many other domains of mathematics and computer science. These skills are general and allow data scientists to offer valuable services to almost any field. However, data scientists in some cases find themselves in industries they have relatively little knowledge of.

This is especially true in the healthcare field. In healthcare, there is an enormous amount of important clinical knowledge that might be relevant to a data scientist. It is unreasonable to expect a data scientist to not only have all of the skills typically required of a data scientist, but to also have all of the knowledge a medical professional may have.

Why is domain knowledge necessary?

This lack of domain knowledge, while perfectly understandable, can be a major barrier to healthcare data scientists. For one thing, it’s difficult to come up with project ideas in a domain that you don’t know much about. It can also be difficult to determine the type of data that may be helpful for a project – if you want to build a model to predict a health outcome (for example, whether a patient has or is likely to develop a gastrointestinal bleed), you need to know what types of variables might be related to this outcome so you can make sure to gather the right data.

Knowing the domain is useful not only for figuring out projects and how to approach them, but also for having rules of thumb for sanity checks on the data. Knowing how data is captured (is it hand-entered? Is it from machines that can give false readings for any number of reasons?) can help a data scientist with data cleaning and from going too far down the wrong path. It can also inform what true outliers are and which values might just be due to measurement error.

Often the most challenging part of building a machine learning model is feature engineering. Understanding clinical variables and how they relate to a health outcome is extremely important for this. Is a long history of high blood pressure important for predicting heart problems, or is only very recent history? How long a time horizon is considered ‘long’ or ‘short’ in this context? What other variables might be related to this health outcome? Knowing the domain can help direct the data exploration and greatly speed (and enhance) the feature engineering process.

Once features are generated, knowing what relationships between variables are plausible helps for basic sanity checks. If you’re finding the best predictor of hospitalization is the patient’s eye color, this might indicate an issue with your code. Being able to glance at the outcome of a model and determine if they make sense goes a long way for quality assurance of any analytical work.

Finally, one of the biggest reasons a strong understanding of the data is important is because you have to interpret the results of analyses and modeling work. Knowing what results are important and which are trivial is important for the presentation and communication of results. An analysis that determines there is a strong relationship between age and mortality is probably well-known to clinicians, while weaker but more surprising associations may be of more use. It’s also important to know what results are actionable. An analysis that finds that patients who are elderly are likely to end up hospitalized is less useful for trying to determine the best way to reduce hospitalizations (at least, without further context).

How do you get domain knowledge?

In some industries, such as tech, it’s fairly easy and straightforward to see an end-user’s prospective. By simply viewing a website or piece of software from the user’s point of view, a data scientist can gain a lot of the needed context and background knowledge needed to understand where their data is coming from and how their model output is being used. In the healthcare industry, it’s more difficult. A data scientist can’t easily choose to go through med school or the experience of being treated for a chronic illness. This means there is no easy single answer to where to gain domain knowledge. However, there are many avenues available.

Reading literature and attending presentations can boost one’s domain knowledge. However, it’s often difficult to find resources that are penetrable for someone who is not already a clinician. To gain deep knowledge, one needs to be steeped in the topic. One important avenue to doing this is through the establishment of good relationships with clinicians. Clinicians can be powerful allies that can help point you in the right direction for understanding your data, and simply by chatting with them you can gain important insights. They can also help you visit the clinics or practices to interact with the people that perform the procedures or even watch the procedures being done. At Fresenius Medical Care, where I work, members of my team regularly visit clinics. I have in the last year visited one of our dialysis clinics, a nephrology practice, and a vascular care unit. These experiences have been invaluable to me in developing my knowledge of the treatment of chronic illnesses.

In conclusion, it is crucial for data scientists to acquire basic familiarity in the field they are working in and in being part of collaborative teams that include people who are technically knowledgeable in the field they work in. This said, acquiring even an essential understanding (such as “Medicine 101”) may go a long way for the data scientists in being able to become self-sufficient in essential feature selection and design.

 

Process Mining: Innovative Analyse von Datenspuren für Audit und Forensik

Step-by-Step:

Neue Möglichkeiten zur Aufdeckung von Compliance-Verstößen mit Process Analytics

Im Zuge der fortschreitenden Digitalisierung findet derzeit ein enormer Umbruch der alltäglichen Arbeit hin zur lückenlosen Erfassung aller Arbeitsschritte in IT-Systemen statt. Darüber hinaus sehen sich Unternehmen mit zunehmend verschärften Regulierungsanforderungen an ihre IT-Systeme konfrontiert.

Der unaufhaltsame Trend hin zur vernetzten Welt („Internet of Things“) wird die Möglichkeiten der Prozesstransparenz noch weiter vergrößern – jedoch werden bereits jetzt viele Prozesse im Unternehmensbereich über ein oder mehrere IT-Systeme erfasst. Jeder Mitarbeiter, aber auch jeder automatisiert ablaufende Prozess hinterlässt viele Datenspuren in IT-Backend-Systemen, aus denen Prozesse rückwirkend oder in Echtzeit nachgebildet werden können. Diese umfassen sowohl offensichtliche Prozesse, wie etwa den Eintrag einer erfassten Bestellung oder Rechnung, als auch teilweise verborgene Prozesse, wie beispielsweise die Änderung bestimmter Einträge oder Löschung dieser Geschäftsobjekte. 


english-flagRead this article in English:
“Process Analytics – Data Analysis for Process Audit & Improvement”


1 Das Verständnis von Process Analytics

Process Analytics ist eine datengetriebene Methodik der Ist-Prozessanalyse, die ihren Ursprung in der Forensik hat. Im Kern des dieser am Zweck orientierten Analyse steht das sogenannte Process Mining, eine auf die Rekonstruktion von Prozessen ausgerichtetes Data Mining. Im Zuge der steigenden Bedeutung der Computerkriminalität wurde es notwendig, die Datenspuren, die potenzielle Kriminelle in IT-Systemen hinterließen, zu identifizieren und zu analysieren, um das Geschehen so gut wie möglich zu rekonstruieren.

Mit dem Trend hin zu Big Data Analytics hat Process Analytics nicht nur neue Datengrundlagen erhalten, sondern ist als Analysemethode weiterentwickelt worden. Zudem ermöglicht die Visualisierung dem Analysten oder Berichtsempfänger ein tief gehendes Verständnis auch komplexerer Geschäftsprozesse.

Während in der konventionellen Prozessanalyse vor allem Mitarbeiterinterviews und Beobachtung der Mitarbeiter am Schreibtisch durchgeführt werden, um tatsächlich gelebte Prozesse zu ermitteln, ist Process Analytics eine führende Methode, die rein faktenbasiert und damit objektiv an die Prozesse herangeht. Befragt werden nicht die Mitarbeiter, sondern die IT-Systeme, die nicht nur alle erfassten Geschäftsobjekte tabellenorientiert abspeichern, sondern auch im Hintergrund – unsichtbar für die Anwender – jegliche Änderungsvorgänge z. B. an Bestellungen, Rechnungen oder Kundenaufträgen lückenlos mit einem Zeitstempel (oft Sekunden- oder Millisekunden-genau) protokollieren.

2 Die richtige Auswahl der zu betrachtenden Prozesse

Heute arbeitet nahezu jedes Unternehmen mit mindestens einem ERP-System. Da häufig noch weitere Systeme eingesetzt werden, lässt sich klar herausstellen, welche Prozesse nicht analysiert werden können: Solche Prozesse, die noch ausschließlich auf Papier und im Kopf der Mitarbeiter ablaufen, also typische Entscheiderprozesse auf oberster, strategischer Ebene, die nicht in IT-Systemen erfasst und dementsprechend nicht ausgewertet werden können. Operative Prozesse werden hingegen in der Regel nahezu lückenlos in IT-Systemen erfasst und operative Entscheidungen protokolliert.

Zu den operativen Prozessen, die mit Process Analytics sehr gut rekonstruiert und analysiert werden können und gleichermaßen aus Compliance-Sicht von höchstem Interesse sind, gehören beispielsweise Prozesse der:

  • Beschaffung
  • Logistik / Transport
  • Vertriebs-/Auftragsvorgänge
  • Gewährleistungsabwicklung
  • Schadensregulierung
  • Kreditgewährung

Process Analytics bzw. Process Mining ermöglicht unabhängig von der Branche und dem Fachbereich die größtmögliche Transparenz über alle operativen Geschäftsprozesse. Für die Audit-Analyse ist dabei zu beachten, dass jeder Prozess separat betrachtet werden sollte, denn die Rekonstruktion erfolgt anhand von Vorgangsnummern, die je nach Prozess unterschiedlich sein können. Typische Vorgangsnummern sind beispielsweise Bestell-, Auftrags-, Kunden- oder Materialnummern.

3 Auswahl der relevanten IT-Systeme

Grundsätzlich sollte jedes im Unternehmen eingesetzte IT-System hinsichtlich der Relevanz für den zu analysierenden Prozess untersucht werden. Für die Analyse der Einkaufsprozesse ist in der Regel nur das ERP-System (z. B. SAP ERP) von Bedeutung. Einige Unternehmen verfügen jedoch über ein separates System der Buchhaltung (z.B. DATEV) oder ein CRM/SRM (z. B. von Microsoft), die dann ebenfalls einzubeziehen sind.

Bei anderen Prozessen können außer dem ERP-/CRM-System auch Daten aus anderen IT-Systemen eine entscheidende Rolle spielen. Gelegentlich sollten auch externe Daten integriert werden, wenn diese aus extern gelagerten Datenquellen wichtige Prozessinformationen liefern – beispielsweise Daten aus der Logistik.

4 Datenaufbereitung

Vor der datengetriebenen Prozessanalyse müssen die Daten, die auf Prozessaktivitäten direkt oder indirekt hindeuten, in den Datenquellen identifiziert, extrahiert und aufbereitet werden. Die Daten liegen in Datenbanktabellen und Server-Logs vor und werden über ein Data Warehousing Verfahren zusammengeführt und in ein Prozessprotokoll (unter den Process Minern i.d.R. als Event Log bezeichnet) umformuliert.

Das Prozessprotokoll ist in der Regel eine sehr große und breite Tabelle, die neben den eigentlichen Prozessaktivitäten auch Parameter enthält, über die sich Prozesse filtern lassen, beispielsweise Informationen über Produktgruppen, Preise, Mengen, Volumen, Fachbereiche oder Mitarbeitergruppen.

5 Prüfungsdurchführung

Die eigentliche Prüfung erfolgt visuell und somit intuitiv vor einem Prozessflussdiagramm, das die tatsächlichen Prozesse so darstellt, wie sie aus den IT-Systemen extrahiert werden konnten.

Process Mining – Beispielhafter Process Flow mit Fluxicon Disco (www.fluxicon.com)

Das durch die Datenaufbereitung erstellte Prozessprotokoll wird in eine Datenvisualisierungssoftware geladen, die dieses Protokoll über die Vorgangsnummern und Zeitstempel in einem grafischen Prozessnetzwerk darstellt. Die Prozessflüsse werden also nicht modelliert, wie es bei den Soll-Prozessen der Fall ist, sondern es „sprechen“ die IT-Systeme.

Die Prozessflüsse werden visuell dargestellt und statistisch ausgewertet, so dass konkrete Aussagen über die im Hinblick auf Compliance relevante Prozess-Performance und -Risiken getroffen werden können.

6 Abweichung von Soll-Prozessen

Die Möglichkeit des intuitiven Filterns der Prozessdarstellung ermöglicht auch die gezielte Analyse von Ist-Prozessen, die von den Soll-Prozessverläufen abweichen.

Die Abweichung der Ist-Prozesse von den Soll-Prozessen wird in der Regel selbst von IT-affinen Führungskräften unterschätzt – mit Process Analytics lassen sich nun alle Abweichungen und die generelle Prozesskomplexität auf ihren Daten basierend untersuchen.

6 Erkennung von Prozesskontrollverletzungen

Die Implementierung von Prozesskontrollen sind Bestandteil eines professionellen Internen Kontrollsystems (IKS), die tatsächliche Einhaltung dieser Kontrollen in der Praxis ist jedoch häufig nicht untersucht oder belegt. Process Analytics ermöglicht hier die Umgehung des Vier-Augen-Prinzips bzw. die Aufdeckung von Funktionstrennungskonflikten. Zudem werden auch die bewusste Außerkraftsetzung von internen Kontrollmechanismen durch leitende Mitarbeiter oder die falsche Konfiguration der IT-Systeme deutlich sichtbar.

7 Erkennung von bisher unbekannten Verhaltensmustern

Nach der Prüfung der Einhaltung bestehender Kontrollen, also bekannter Muster, wird Process Analytics weiterhin zur Neuerkennung von bislang unbekannten Mustern in Prozessnetzwerken, die auf Risiken oder gar konkrete Betrugsfälle hindeuten und aufgrund ihrer bisherigen Unbekanntheit von keiner Kontrolle erfasst werden, genutzt. Insbesondere durch die – wie bereits erwähnt – häufig unterschätzte Komplexität der alltäglichen Prozessverflechtung fallen erst durch diese Analyse Fraud-Szenarien auf, die vorher nicht denkbar gewesen wären. An dieser Stelle erweitert sich die Vorgehensweise des Process Mining um die Methoden des maschinellen Lernens (Machine Learning), typischerweise unter Einsatz von Clustering, Klassifikation und Regression.

8 Berichterstattung – auch in Echtzeit möglich

Als hocheffektive Audit-Analyse ist Process Analytics bereits als iterative Prüfung in Abständen von drei bis zwölf Monaten ausreichend. Nach der erstmaligen Durchführung werden bereits Compliance-Verstöße, schwache oder gar unwirksame Kontrollen und gegebenenfalls sogar Betrugsfälle zuverlässig erkannt. Die Erkenntnisse können im Nachgang dazu genutzt werden, um die Schwachstellen abzustellen. Eine weitere Durchführung der Analyse nach einer Karenzzeit ermöglicht dann die Beurteilung der Wirksamkeit getroffener Maßnahmen.

In einigen Anwendungsszenarien ist auch die nahtlose Anbindung der Prozessanalyse mit visuellem Dashboard an die IT-Systemlandschaft zu empfehlen, so dass Prozesse in nahezu Echtzeit abgebildet werden können. Diese Anbindung kann zudem um Benachrichtigungssysteme ergänzt werden, so dass Entscheider und Revisoren via SMS oder E-Mail automatisiert über aktuellste Prozessverstöße informiert werden. Process Analytics wird somit zum Realtime Analytics.

Fazit

Process Analytics ist im Zuge der Digitalisieurng die hocheffektive Methodik aus dem Bereich der Big Data Analyse zur Aufdeckung Compliance-relevanter Tatbestände im gesamten Unternehmensbereich und auch eine visuelle Unterstützung bei der forensischen Datenanalyse.

 

Ways AI & ML Are Changing How We Live

From Amazon’s Alexa, a personal assistant that can do anything from making your to-do list to giving a wide range of real-time information about the world around you, to Google’s DeepMind that has very recently made headlines for possibly being able to predict the future, AI and ML are the biggest development in human history.

Machine Learning Used by Hospitals

We hear a lot about Artificial Intelligence (AI) in the realm of insurance Big Data, but there isn’t much buzz around how AI and ML are revolutionising hospitals. The national health expenditures were around $3.4 trillion and estimated to increase from 17.8 percent of GDP to 19.9 percent between 2015 and 2025. By 2021, industry analysts have predicted that the AI health market will reach $6.6 billion. By 2026, such increases in AI technology in the healthcare sector will save the economy around $150 billion annually.

Some of the most popular Artificial Intelligence applications used in hospitals now are:

  • Predictive Health Trackers – Technology that has the ability to monitor patients’ health status using real-time data collection. One such technology is the Health and Environmental Tracker (HET) which can predict if someone is about to have an asthma attack.
  • Chatbots – It isn’t only retail customer service that uses chatbots to deal with consumers. Now hospitals have automated physicians that inquire and route clinicians to the right specialists.
  • Predictive AnalyticsCleveland Clinics have partnered with Microsoft (Cortana) while John Hopkins has partnered up with GE in order to create Machine Learning technology that has the ability to monitor patients and prevent patient emergencies before they happen. It does this by analysing data for primary indicators of potential risks.

Cognitive Marketing – Content Marketing on Steroids

Customer experience and content marketing are terms often tossed around in the world of business and advertising these days. Why do we bring them up now, you ask? Well, things are about to be kicked into sixth gear, thanks to Cognitive Marketing. To explain what that is, let’s go back a bit: remember when Google’s DeepMind AlphaGo bested the top human player at the game? This wasn’t some computer beating a bored office clerk at the game of Solitaire. In order to achieve that victory, Google’s AI had to “actually show its cognitive capability to ‘think’ like humans, because to win the game, ‘intuition’ was needed rather than just ‘logical reasoning’.” Similar algorithm-powered AI’s are enabling machines to learn and grow on their own. Soon, they’ll reach the potential to create content for marketeers at a massive scale. Not only that, but they’ll always deliver the right content, to the right kind of audience, at just the right time.

More Ways Than One: How Retail Is Harnessing AI & ML

  1. Developing Store That Don’t Need Checkout Lines

Tech companies and online retail giants such as Amazon want to create cashier-free stores, at least they are trying to. Last year Amazon launched its Amazon Go which uses sensors and hundreds of cameras to track what customers pick up and then charge the amount to an application on their smart phone, put simply. But only months into the experiment Amazon has said they need to work out some kinks in the system. As of now, Amazon Go’s system can only handle 20 or so customers at a time.

Among other issues, The Guardian, citing an unnamed source, wrote in an article, stated “…if an item has been moved from its specific spot on the shelf.”  Located in Seattle, Washington, Amazon Go is now running in “beta mode” only for Amazon employees as it tests its systems. And these tests are showing that Amazon’s attempt at a cashier-free brick-and-mortar convenience store is far from ready for the real world. A Journal report stated, “For now, the technology functions flawlessly only if there are a small number of customers present, or when their movements are slow.”

  1. Could Drones Be Delivering Goods to Your Home One Day?

Imagine ordering something online from, let’s say, Amazon, and it arrives at your door in 30 minutes or so via drone. Does that sound like something out of the movie The Fifth Element? Maybe, but this technology is already is already here.

Amazon Prime Air made its first delivery to a customer via a GPS-guided flying drone on December 7th, 2016. It only took 13 minutes for the drone to deliver the merchandise to the customer. This sort of technology will be a huge game changer for retail. The supply chain industry is headed for a revolution – drone delivery is coming, and retailers who want to keep up really should adopt such technologies.

Even in 2016, consumers were totally ready to accept drone delivery. The Walk Sands Future of Retail 2016 Study showed that 79 percent of US consumers said they would be “very likely” or “somewhat likely” to choose drone delivery if their product could be delivered within an hour. For me, I’d choose it just to see how cool it was. I think it would be pretty rad to have a drone land in my yard with my package, don’t you? Furthermore, other consumers stated they would pay up to $10 for a drone delivery. Lastly, 26 percent of consumers are already expecting to have their packages delivered to them in the next two years or so.

Driverless Delivery Vehicles Already Here as Well

There was a movie I watched some months ago – you most likely heard of it or even watched it. It was the latest movie about Wolverine titled Logan. There was a certain scene that never left my memory (basically because I found it awesome) where Logan and his companions were driving along a freeway full of driverless tractor trailers that had no tractor.

In an article written for pastemagazine.com, Carlos Alvarez of Getty wrote: “… Logan’s writer and director James Mangold’s inclusion of the self-driving trucking machines make it clear that the filmmaker understands the writing on the wall about the future of shipping. It’s a future without truck drivers.” He continues to explain that the movie takes place a little over 10 years from now in 2029.

“The change may well be here long before 2029. It’s only 2017, and already we’re seeing the beginnings of automated trucking taking over the industry. At the 2017 Consumer Electronics Show this January, Peloton Technology demonstrated “platooning,” where trucks are kept in a row on the highway to reduce wind resistance and save fuel. The trucks are controlled by computers on a “Level One” of autonomous driving,” Alvarez continued in his article.

Now in Germany, Mercedes-Benz is has been developing and testing their Actros truck which is fitted with a ‘highway pilot’ system, which acts like an auto-pilot and includes a radar and stereo camera system. So far, German carmaker Daimler has restricted testing on a German autobahn. The autobahn is generally safer than testing in city conditions since the curves are not as steep. Since the tests have started, this autonomous truck has already driven over 20,000 kilometres.

Did I Say Flying Taxis? Huh, Yeah I Did!

But, if you are still not amazed, then I am about to blow your socks off. Dubai has promised to build a fully autonomous public transportation system by 2030, including autonomous flying drone taxis! Now that is really something. And it isn’t a matter of when they’ll be produced and in use because they already are.

Manufactured in China by the drone-making firm EHang, these really freaking cool quad drones on steroids can carry one person weighing up to 100 kilogrammes (I weigh over that, guess I’m walking) plus maybe a backpack or suitcase. They can fly about 30 kilometres (or 19 miles), at a speed of 60 miles per hour, give or take. And, if that isn’t the cool part, you won’t need any lessons on how to fly it. Simply push a button and it flies you from point A to point B. Whether or not you have to give it directions, don’t know. Either way, this is mostly likely the coolest piece of tech out there right now.

Copyright @ CBS Interactive Inc.