This is an old revision of the document!


Master Projects

Documents

Possible assignments

Detecting root causes of complaints and investigating the continuation within the customer journey

In the Dutch health care system health care insurance is obligated for all residents. The government sets the basis package and insurers compete based on price and service. Customer service is therefore very important for every health insurance company; especially in the fast changing digital world. As a result customer satisfaction is the most important KPI. Complaints are related to customer satisfaction, because they have a highly negative effect on customer satisfaction. Therefore, it is crucial to prevent customers from having any complaints. Customers do not submit a complaint out of the blue: a complaint is a reaction to one or more occurrences in the past. The best way to understand the triggers, or root causes, it to understand the touchpoints prior to a complaint.

CZ aims for a method to prevent complaints (in order to increase customer satisfaction and lower churn). When taking the touchpoints of the customer journey into account, complaint root causes can be detected and CZ can prevent future complaints. Moreover, the touchpoints after the complaint give insights about an optimal follow up after the complaint.

Prior research development

Underlined works together with several companies like CZ, Aegon and SNS to build a generic framework in which all customer contacts are brought together as a unique dataset which is further enriched by Underlined with relevant analyses that can be linked to customer events. Underlined co-created customer journey mining algorithms, together with the TU/e (Bart Hompes and Joos Buijs from the Architecture of Information Systems group). This research showed that it is possible to distinguish the different journeys per customer. Moreover, in collaboration with the TiU econometrist department, a driver model has recently developed to distinguish relevant drivers and to predict the NPS-score.

Research goal and questions:

In further collaborative research between TU/e, CZ and Underlined, the research focus will be on developing a predictive method on complaints, based on process mining techniques. The scope of the current master project will be developing a complaint driver model by taking into account the customer journey to detect root causes and follow up triggers. More precisely, after a proper preprocessing of the different types of data, the master project will start by addressing the following list of questions:

  1. Can customers be segmented? Which distance measure is suitable for clustering customers such that similar customers can be grouped based on their behavior throughout their journey? How good is this clustering method when including the domain knowledge and/or internal quality evaluation measures?
  2. What has happened ? What is the model of the current complaining process ? How does this differs among the different customers? Which sources of data are relevant to enrich this model?
  3. Why did it happen? How does data from the different cases of each segment decides different paths within the model? Is this in line with the expectations of the business owner?
  4. What will happen? Based on the above segmentation, can we predict the category of the current customer, using their characteristics and behavior ?

In the above-mentioned analysis, all available data will be leveraged. This includes logged data of the journey (workflow process data, call center data, online click trails, social media data), online and offline feedback and non-transactional data (product and background information).

About Underlined

Underlined has a proven approach and toolset for providing insights how to improve customer experience during the customer journey. Using data from customer contact channels, online environment, customer feedback and research response, we can reveal the actual customer journey and customer experience. This enables companies to measure the customer journey of their customers’ behaviour and emotions continuously and to manage and improve the journey across all channels. Underlined calls this Customer Journey Management. More information about Underlined can be found at our website: www.underlined.nl

About CZ

CZ is one of the largest health insurance companies in the Netherlands. The core activity of CZ group is the implementation of compulsory insurance against medical expenses: the basic insurance. In addition, we also offer supplementary insurance for health risks that are not covered by the basic insurance. In addition, we offer specific products for employers, such as group health insurance and occupational health programs. More information about CZ can be found at our website: www.cz.nl

Contact

For more information, contact Marwan Hassani.

Philips HUE Product Evolution Using Stream Mining of Customer Journey

Philips HUE is a connected personal lighting system. It is controlled by a range of apps and smart home devices.

To acquire Philips HUE, one starts with a starter kit that consists of a few lamps and a bridge. Subsequently, consumers decide to expand their system with additional lamps or/and physical sensors. About 50 lamps can be controlled in one system.

An example of HUE system is seen in the Figure below. It consists of a bridge, three color lamps and a dimmer switch.

Every consumer chooses a device to interact with the lights. By default, all consumers start with HUE or third-party app. There are also options in the HUE app to create routines to be able to control the lights automatically. For instance, one can create a wakeup routine so that the lights in the bedroom go on naturally in the morning. HUE can also be controlled by Philips physical sensors (motion sensor, tap, dimmer switch), voice control and other third party smart home devices. Amazon Echo, Google Home, Eneco Toon, Homekit are some of smart home devices that are linked to HUE. In addition, traditional wall switch can also be used to control the lights. For further information see Philips Lighting website.

The master thesis focusses on developing a process mining model to understand product evolution of consumers in their journey and identify the paths that contribute to a high Customer Lifetime Value (CLV). Moreover, the thesis will also explore data mining techniques that can be applied on the outcome of process mining model.

We would like to capture customer behavior in process models. This way we can understand the behavior better and systematically explore ways of influencing this. Process mining techniques can be used to extract process-related information from event logs. Process mining is the linking pin between data science and process science. Data science approaches tend to be process agonistic whereas process science approaches tend to be model-driven without considering the “evidence” hidden in the data. Process mining aims to bridge this gap. Process mining seeks the confrontation between event data (i.e., observed behavior) and process models (hand-made or discovered automatically). The interest in process mining is rising as is reflected by the growing numbers of publications, citations and commercial tools (Disco, ARIS PPM, QPR, Celonis, SNP, minit, myInvenio, Perceptive, etc.). In the academic world, ProM is the de-facto standard (www.processmining.org) and research groups all over the world have contributed to the hundreds of ProM plug-ins available.

Understanding the evolution of customer behavior during their journey requires a temporal analysis of the upgrading events. Stream process mining of the generated models guarantees observing the evolution of the interleaving sub-processes and detecting interesting concept drifts. Stream mining is a recent subject in data mining. The task of streaming process discovery includes, among other things, working on potentially unbounded event logs. Few algorithms have been developed, that are able to perform process discovery in a single pass over the data. This master project will explore applying several streaming process mining and stream data mining techniques on the event data in order to answer the following questions:

  1. What happened? Which usage, extension and upgrade patterns do we have? How did they evolve over time? Which concept drifts exist?
  2. Why did happen? Will any, and if so which, customer segmentation help in explaining the answers to (1)? Will splitting interleaving sub-process better relate these answers to e.g. some timing aspects?
  3. What will happen? Can we predict next customer usages, extensions and/or upgrades based on our previous learned models?
  4. What is the best that could happen? How can we use our prediction knowledge in recommending certain controllable actions such that Customer Lifetime Value (CLV) (or other KPIs) can be maximized?

Contact

For more information, contact Marwan Hassani.

Process mining in Logistics - 3D Visualization and Scalable Process Mining on Big Event Data (2 Topics)

Vanderlande is the global market leader for value-added logistic process automation at airports and in the parcel market. The company is also a leading supplier of process automation solutions for warehouses. Some figures:

  • Vanderlande’s baggage handling systems move 3.7 billion pieces of luggage around the world per year.
  • Our systems are active in 600 airports including 13 of the world’s top 20.
  • More than 39 million parcels are sorted by its systems every day, which have been installed for the world’s leading parcel companies.
  • Many of the largest global e-commerce players and distribution firms have confidence in Vanderlande’s efficient and reliable solutions.

Vanderlande focuses on the optimization of its customers’ business processes and competitive positions. Through close cooperation, we strive for the improvement of our customers’ operational activities and the expansion of their logistical achievements.

For Vanderlande, it is critical that we have state-of-the-art techniques to analyze and optimize our customers’ logistics processes. Reasons are (a) the constant increasing size and complexity of our material handling solutions, (b) growing complexity of our software solutions, covering larger parts of our customers’ business processes, and © the demand for more advanced service offerings, covering logistics and business services. We believe that process mining is of high value for Vanderlande. Therefore, we work with the Eindhoven University of Technology on making process mining fit for analyzing logistics processes. In this context, we offer graduation projects on process mining and the application in our business.

Topic 1: 3D Visualisation of logistic processes

Within Vanderlande, we use a 3D library to create realistic simulation and emulation models. The topic of this graduation topic is to use this existing library within our process mining tooling; we want to visualize the logistic processes in 3D within our process mining tooling.

Deliverables:

  • Become acquainted with both our 3D library and our process mining tooling
  • Come up with a conceptual solution to combine both
  • Realize the solution in a proof-of-concept, where the existing 3D library is connected to our process mining tooling
  • Application of the proof-of-concept on at least one system to test and validate

Topic 2: Implementation of scalable process discovery and conformance checking algorithms

Our systems generate big amounts of data. This is often a limiting factor; existing process mining tooling cannot work with multiple days of systems data. Because of this, we perform our analyses on a relative limited time-period, which reduces the certainty of the outcome of our analyses. For this, we are looking for algorithms that allow for scalable process discovery and conformance checking. This is a topic that we would like to work on with a graduate student.

Deliverables of the graduation study:

  • Problem analysis
  • Conceptual solution of algorithms that allow for scalable process discovery and conformance checking
  • Realize the solution in a proof-of-concept
  • Apply the proof-of-concept on at least one system to test and validate

More Information

For more information, please contact Dr. Dirk Fahland

Want to win?

Win the Process Discovery Contest (PDC) 2018!

Project setting

The Process Discovery Contest is dedicated to the assessment of tools and techniques that discover business process models from event logs. The objective is to compare the efficiency of techniques to discover process models that provide a proper balance between “overfitting” and “underfitting”. A process model is overfitting (the event log) if it is too restrictive, disallowing behavior which is part of the underlying process. This typically occurs when the model only allows for the behavior recorded in the event log. Conversely, it is underfitting (the reality) if it is not restrictive enough, allowing behavior which is not part of the underlying process. This typically occurs if it overgeneralizes the example behavior in the event log. A number of event logs will be provided. These event logs are generated from business process models that show different behavioral characteristics. The process models will be kept secret: only “training” event logs showing a portion of the possible behavior will be disclosed. The winner is/are the contestant(s) that provides the technique that can discover process models that are the closest to the original process models, in term of balancing between “overfitting” and “underfitting”. To assess this balance we take a classification perspective where a “test” event log will be used. The test event log contains traces representing real process behavior and traces representing behavior not related to the process. Each trace of the training and test log will record complete executions of instances of the business processes. In other words, each trace records all events of one process instance from the starting state till the end state.

A model is as good in balancing “overfitting” and “underfitting” as it is able to correctly classify the traces in the “test” event log:

  • Given a trace representing real process behavior, the model should classify it as allowed.
  • Given a trace representing a behavior not related to the process, the model should classify it as disallowed.

The contest is not restricted to any modelling notation and no preference is made. Any procedural (e.g., Petri Net or BPMN) or declarative (e.g., Declare) notation is equally welcome. The context is not restricted to open-source tools. Proprietary tools can also participate.

Project description

The goal of the Master Project is to participate in the 2018 edition of the Process Discovery Contest, if possible using the techniques that were developed for the 2017 edition (see the figure below for an example model) and that allowed us to classify all traces correctly. However, these techniques did not allow us to win the 2017 edition because the models we generated were considered to be less informative than the BPMN models as discovered by the winning competitor. As a result, a possibility would be to develop a conversion from our models to BPMN models. Furthermore, the 2018 edition might bring new challenges to the Contest, which might require extensions to our techniques. As the call for the 2018 edition is not out yet, it is hard to say what kind of extensions would be needed.

For the earlier editions of the Contest, the price for the winner included a flight to the BPM conference, the lodging expenses during the conference, and a full registration for the conference. Provided that the same prices will be available for the 2018 edition, and provided that the result of the Master Project wins the Contest, it will be the master student who will pick up these prices and visit the BPM 2018 Conference, which will be held in Sydney.

Time restrictions

Given that the main goal is to participate in the 2018 edition of the Contest, it would be ideal if the master student starts just before the Contest starts. This way, the student can first get acquainted with the techniques using the 2017 edition, and then start applying and extending them for the 2018 edition. As the Contest typically starts in March, it would be ideal if the student starts February/March.

Project Team

Principal Supervisor
Renata Medeiros de Carvalho
Position: UD
Room: MF 7.146
Tel (internal):
Projects:
Courses: 2IAB0, 2IMC93, 2IMC98, JM0200
Links: Scopus page, DBLP page, TU/e employee page
Daily Supervisor
Eric Verbeek
Position: Scientific Programmer
Room: MF 7.062
Tel (internal): 3755
Projects: CoseLoG
Courses:
Links: Personal home page, Google scholar page, Scopus page, ORCID page, DBLP page, TU/e employee page
Eric is the scientific programmer in the AIS group. As such, he is the custodian of the process mining framework ProM. In you want access to the ProM repository, or have any questions related to ProM and its development, as Eric. Recently, he has been working on a decomposition framework for both process discovery as conformance checking in ProM. Earlier, he also worked on ExSpect and Woflan.

When Portfolio Management meets Process Mining Challenges and Opportunities

FLIGHTMAP is Bicore’s flagship software solution for portfolio management. Since its launch in 2010, a growing group of international clients, such as DAF, Océ, and Fokker, have implemented FLIGHTMAP. With this tooling, they can perform roadmapping, budget and resource planning, scenario analysis, planning and tracking, and more. More information about FLIGHTMAP is available via www.flightmap.com. The figure above shows a screenshot of the tool obtained after the portfolio analysis

To keep its leading position, Bicore continuously develops its innovative functionality for decision support and the ease of use and embedding. Next to this core functionality, FLIGHTMAP has additional modules, such as the HUB for connecting to external systems. The major next steps in development are upgrades to the latest front-end technologies and a stepwise migration to the cloud.

As an important step to make FLIGHTMAP smarter in decision support, we are looking to leverage new insights from process mining and data analytics into FLIGHTMAP’s functionality.

Bicore is looking for candidates for a Master Thesis internship in the area of process and data mining to bring new insights and link them to FLIGHTMAP’s next releases.

The assignment should look into the best way to harvest best practices in portfolio analysis and portfolio reviews, as well as project selection. Since FLIGHTMAP is running for more than 7 years, a lot of historical data is available to work with. Where possible, we would like to link the results of mining to specific recommendations.

In discussion with academic expert sin this fields, the application of process mining techniques are much more common in the traditional transaction systems domain. Applying them in the decision support domain of FLIGHTMAP gives rise to interesting research questions, as well as to practical relevance.

You will work closely with the development and delivery team of FLIGHTMAP to align, and contribute to the FLIGHTMAP roadmap.

Contact

For more information, contact Massimiliano de Leoni.

Example completed master projects

Bram in 't Groen

VDSEIR - A graphical layer on top of the Octopus toolset

Description

In his work, Bram in 't Groen introduces a graphical representation for DSEIR (a language used in the Octopus toolset for designing embedded systems) called Visual DSEIR (VDSEIR). By using VDSEIR, users of the toolset can create specifications in DSEIR by means of creating graphical models, removing the need for those users to know how to program in the Octopus API. Bram in 't Groen provides a model transformation from VDSEIR to DSEIR that makes use of an intermediate generator model and a parser that is automatically generated from an annotated JavaCC grammar. The graphical representation for DSEIR consists of several perspectives and it contains a special form of syntactic sugar, namely hierarchy. It is possible to create hierarchical models in the graphical representation without having support for hierarchy in the original DSEIR language, because these hierarchical models can be translated into non-hierarchical DSEIR models. This way, additional expressiveness is created for the user, without modifying the underlying toolset.

Type

AIS / External / ESI

Borana Luka

Model merging in the context of configurable process models

Description

While the role of business process models in the operation of modern organizations becomes more and more prominent, configurable process models have recently emerged as an approach that can facilitate their reuse, thereby helping to reduce costs and effort. Configurable models incorporate the behavior of several model variants into one model, which can be configured and individualized as necessary. The creation of configurable models is a complicated process, and tool support for it is in its early steps. In her thesis, Borana Luka evaluates two existing approaches to process model merging which are supported by tools and test an approach to model merging based on the similarity between models. Borana’s work resulted in a paper presented in the 2011 International Workshop on Process Model Collections.

Type

AIS / Internal / CoSeLog project involving 10 municpalities

Links

Staff involved

Cosmina Cristina Niculae

Guided configuration of industry reference models

Description

Configurable process models are compact representations of process families, capturing both the similarities and differences of business processes and further allowing for the individualization of such processes in line with particular requirements. Such a representation of business processes can be adopted in the consultancy sector and especially in the ERP market, as ERP systems represent general solutions applicable for a range of industries and need further configuration before being implemented to particular organizations. Configurable process models can potentially bring several benefits when used in practice, such as faster delivery times in project implementations or standardization of business processes. Cosmina Niculae conducted her project within To-Increase B.V., a company that specializes in ERP implementations. She developed an approach to make configuration much easier, implemented it, and tested it on real-life cases within To-Increase.

Type

AIS / External / To-Increase

Links

Staff involved

Dennis Schunselaar

Configurable Declare

Description

Declarative languages are becoming more popular for modeling business processes with a high degree of variability. Unlike procedural languages, where the models define what is to be done, a declarative model specifies what behavior is not allowed, using constraints on process events. In his thesis, Dennis Schunselaar studies how to support configurability in such a declarative setting. He takes Declare as an example of a declarative process modeling language and introduces Configurable Declare. Configurability is achieved by using configuration options for event hiding and constraint omission. He illustrated our approach using a case study, based on process models of ten Dutch municipalities. A Configurable Declare model is constructed supporting the variations within these municipalities.

Type

AIS / Internal

Links

Staff involved

Erik Nooijen

Artifact-Centric Process Analysis, Process discovery in ERP systems

Description

In his thesis, Erik Nooijen developed an automated technique for discovering process models from enterprise resource planning (ERP) systems. In such systems, several different processes interact together to maintain the resources of a business, where all information about the business resources are stored in a very large relational database. The challenge for discovering processes from ERP system data, is to identify from arbitrary tables how many different processes exist in the system, to extract event data for each instance of each process in the system. Erik Nooijen identified a number of data mining techniques that can solve these challenges and integrated them in a software tool, so that he can automatically extract for a process of the ERP system an event log containing all events of that process. Then classical process discovery techniques allow to show the different process models of the system. Erik conducted his project within Sligro where he is actively using his software to improve the company's processes. The thesis resulted in a workshop paper presented at the International Conference on Business Process Management 2012 in Tallinn, Estonia.

Type

AIS / External / Sligro

Links

Staff involved

Older entries >>


If you have a master project item for this page, which may include possible master project assignments, on-going master projects, and completed master projects, please send it to Eric Verbeek.