Master Thesis - Hybrid Process Model Discovery (Joint project with FBK in Trento, Italy)

Kontakt

Chiara Di Francescomarino

Name

Chiara Di Francescomarino

Telefon

work
+39 0461 314052

E-Mail

E-Mail

Kontakt

Chiara Ghidini

Name

Chiara Ghidini

Telefon

work
+39 0461 314395

E-Mail

E-Mail
 

Note: This thesis project involves working at Fondazione Bruno Kessler (FBK), one of the top research institutes in ICT in Italy (located in Trento)

Description

Discovering process models from the data contained in event logs is a highly relevant, but also notoriously difficult, problem. While process model discovery techniques reported in scientific literature tend to produce formal process models (precisely describing the possible behaviors), commercial process mining tools return informal models (a “picture” of the model not allowing for any form of formal reasoning). At first sight, it seems more attractive to discover formal models rather than informal models. However, formality may also be counterproductive when it comes to (i) the understandability of the models by the end-users; (ii) the lack of vagueness of the models when it is difficult to precisely capture the behavior; (iii) limited scalability of the discovery algorithms aiming to discover huge models from huge logs.

A possibility to overcome the limitations of the existing approaches is discovering hybrid process models from event logs, that is combining the best of both worlds (the academic and the commercial one). The idea behind these hybrid models, as described in the paper “Learning Hybrid Process Models from Events - Process Discovery Without Faking Confidence” https://doi.org/10.1007/978-3-319-65000-5_4, is to show informal dependencies (like in commercial tools) that are deliberately vague and at the same time provide formal semantics for the parts that have a clear and precise semantics. Whenever data provides enough evidence to understand whether activities are in sequence, in mutual exclusion or in parallel, the corresponding constructs (e.g., the sequential connection, the mutually exclusive decision point or a parallel branch split) can be made explicit in the model. If, on the contrary, dependencies are weak or too complex, then they are not left out, but depicted in an informal manner. The idea is to create models that have a solid backbone that can be used for conformance checking, bottleneck analysis, etc. and to add the more unclear parts informally (these are the parts for which there is not enough data or activities are happening seemingly random). The proposed hybrid process model discovery approach has been implemented as a plugin of the ProM Framework.

The purpose of the Master thesis project is to (i) develop new techniques and extend existing techniques for hybrid process model discovery, (ii) enhance the existing implementation to include the new techniques and to make it more usable for end-users; (iii) evaluate the effectiveness of the hybrid model notation as well as of the implementation with end-users.

The project aims to combine the expertise at RWTH on process discovery algorithms and the expertise at FBK in conducting empirical research studies with human subjects. The idea is to spend at least half of the time in Trento using an Erasmus+ scholarship.

About FBK

PDI (Process & Data Intelligence) is a Research Unit of the Information Technology Center, at Fondazione Bruno Kessler (FBK). Dr. Chiara Ghidini and Dr. Chiara Di Francescomarino are experts in predictive process mining. PDI provides master projects supported through Erasmus+. FBK is located in Trento, Italy. FBK has over 400 permanent researchers, 140 doctoral students, 200 visiting fellows and thesis students. FBK is very strong in the field Information and Communication Technology, also specializing in topics such as AI and Data Science.

Erasmus+ traineeship program

This thesis project is offered in the context of the Erasmus+ traineeship program. This means that students needs to apply her/himself. Some basic information:

Time period:

  • To be eligible for funding, the internship must last at least 60 days (preferably 3-6 months).
  • Students can be funded by Erasmus for a total of twelve months per academic cycle.

Who can Participate:

  • German and international students in any course of study can apply as long as they are enrolled in a degree program at RWTH Aachen and have an internship place.

Target Countries by Country Group

  • Country Group 1, 555 Euros funding per month: Denmark, Finland, France, Great Britain, Italy, Ireland, Liechtenstein, Norway, Austria, Sweden

Which internships are funded?

  • Erasmus+ funds qualified industry internships from two months (60 days) to twelve months. Eligible host companies include businesses in the public and private sector, regardless of size, legal form, or industrial sector, including social economy. Research internships at universities, including as part of a final project or thesis, can also be funded through the current Erasmus+ partial program. FBK is considered to be in the first category.

How can you apply?

  • First make sure there is an agreement with the supervisors on both the RWTH and FBK side.
  • After the commitment is there from the supervisors, the student should apply online https://rwthaachen.moveon4.de/locallogin/5b97aec185fb96c949d78223/eng# and upload the requested documents – as listed below – as PDF files. Please have the enterprise that you'll be visiting – the receiving organization – and your faculty – the sending institution – fill out or sign the Learning Agreement for Traineeships. (Please note, Sufficient language skills in the working language must be documented - at least Level B1 of the European Framework of Reference.) The best option is to use English.
  • The application will only be successful if it is properly motivated. The motivation should be content-related (why this project and why visit FBK), show how this adds to your study program, explain why you want to do this, and tell why you are the right person to do this.
  • You must sign a contract stating that you are not receiving double funding.

Required Documents and Learning Agreement for Traineeships to Provide when applying.

  • Learning Agreement for Traineeships
  • Current transcript of records
  • Certificate of enrollment
  • Resume
  • Proof of language skills
  • Statement of motivation, maximum of two DIN A4 pages
  • You are personally responsible for sufficient insurance coverage, that is health, accident, and liability insurance, during your stay abroad. Check early on to make sure whether or not your insurance is valid abroad and for what.
  • DAAD offers Erasmus+ funded students affordable group insurance tailored towards internships abroad. Here you can find more information about the DAAD group insurance Tarif 720.
  • You are personally responsible for sufficient insurance coverage, that is health, accident, and liability insurance, during your stay abroad. Check early on to make sure whether or not your insurance is valid abroad and for what.
  • DAAD offers Erasmus+ funded students affordable group insurance tailored towards internships abroad. Here you can find more information about the DAAD group insurance Tarif 720.

Please check the most current information. The above information is just to provide an idea of the require materials and constraints.

For more information please use the following link: https://www.rwth-aachen.de/go/id/bwyt/lidx/1. Study the information carefully. If you get stuck, contact . is the coordinator of the Erasmus+ traineeship program for Informatik, RWTH Aachen.

Supervisors

Prof.Dr.ir. Wil van der Aalst (RWTH), Dr. Chiara Ghidini, and Dr. Chiara Di Francescomarino (FBK)