Gamification is a method that integrates elements from games into non-game contexts in order to promote user motivation and engagement. Gamification has become increasingly important in teaching, as it can potentially help to optimize the learning process, increase student interest and promote knowledge comparison in a playful way.

In the context of business process modeling, students are confronted with complex systems that need to be described through structured representations and diagrams. Learning these concepts can be challenging and the question arises whether the use of gamification has positive effects on the understanding and application of modeling techniques.

Aim of the thesis:

The aim of the thesis is to investigate the influence of gamification on learning and application with a focus on business process modeling. In particular, the question should be answered as to whether the use of gamification has positive or negative effects on students' learning performance and understanding.

Background

During the modeling of business processes, a large amount of data is generated that provides information on how users proceed, which steps they perform and where difficulties arise during modeling. This includes, for example, information about when a model is created, when changes are made or elements are deleted. These records offer great potential for identifying typical problems and optimization potential in process modelling. An initial implementation for recording this data has already been developed as part of the KEA-mod project. However, the rapid increase in the amount of data led to the underlying database being significantly overloaded and ultimately deactivated.

Aim of the thesis

The aim of this thesis is to investigate how the data generated during business process modeling can be stored so that the database is not overloaded. The focus is on the development and prototypical implementation of a solution within the digital modeling platform KEA-mod.

Background

Modelling tools are becoming increasingly important in the context of software development or the digitalization and automation of business processes, for example. In the BMBF-funded project KEA-Mod, a digital modeling tool was developed as a web application in which evaluation services for the automated evaluation of student models, e.g. from exercises or exams, can be integrated. Modeling languages from the areas of software development and databases (UML, ER) and from the area of business process management (BPMN, Petri nets, EPK) are currently supported. Large Language Models (LLMs) offer great potential to be used profitably in automated assessment by providing students with feedback on their models generated by LLMs.

Task definition

The aim of the project is to investigate various possibilities for integrating LLM technologies into the KEA-Mod modeling platform and to develop a prototype implementation.

Possible focal points:

• How can student models be analyzed and evaluated with regard to the task using an LLM? How can, for example, syntactic correctness, content consistency (semantics) and pragmatic aspects (clear layout design) be taken into account according to the teacher's specifications?
• How can an LLM-based chatbot support students directly during the creation of models? How much information can a chatbot give students to support the learning process?

Students can also contribute their own ideas on the topic in consultation with the supervisors.

Process mining has established itself as a method for the data-driven analysis and improvement of business processes. Process mining makes it possible to reconstruct, analyze and improve real process flows from digital recordings in systems (event logs). A distinction is made between three main aspects: Process Discovery to automatically create process models, Conformance Checking to compare actual and target processes, and Process Enhancement to improve existing processes.

At the same time, large language models (LLMs) have opened up new possibilities in the processing and generation of natural language. The combination of both technologies promises innovative approaches for business process management.

The aim of this work is to investigate LLMs for the extension and improvement of process mining. The focus will be on the development and evaluation of new methods that combine the structured process analysis of process mining with the natural language processing capabilities of LLMs.

Possible research priorities:

• Generation of process descriptions from event logs by LLMs
• Automated and natural language interpretation of process mining results
• Integration of unstructured data in process mining through LLM-based pre-processing
• Conformance checking using LLM-supported rule extraction
• Generation of process improvement proposals based on process mining results

If you are interested, please apply via the following link: https: //portal.wiwi.kit.edu/forms/form/Bewerbung_Abschlussarbeit_AIFB-BIS

Aufgabenstellung

Moderne Routingdienste verfügen über eine Vielzahl proprietärer technischer Parameter. Um die Routen zu personalisieren, können Nutzerpräferenzen auf diese Parameter der Routingdienste abgebildet werden. Im Rahmen dieser Arbeit sollen verschiedene Algorithmen für die Kalibrierung des Valhalla Routingdienstes verglichen werden, mit dem Ziel, ein Tool zur Personalisierung der Routen zu entwickeln.

Zunächst sollen unterschiedliche Algorithmen untersucht und analysiert werden, um den besten Ansatz zur Kalibrierung des Valhalla Routingdienstes zu identifizieren. Darauf aufbauend wird ein Tool entwickelt, das die Routenqualität basierend auf den kalibrierten Parametern verbessert.

Es besteht die Möglichkeit, die kalibrierten Routinginstanzen in Mercedes-Benz Fahrzeugen zu testen. Diese Abschlussarbeit bietet die Chance, an der Schnittstelle zwischen Forschung und praktischer Anwendung zu arbeiten und direkt zur Verbesserung des Routings in der Automobilindustrie beizutragen.

In business process management, the documentation of business processes is still mainly carried out manually in a time-consuming and error-prone manner. However, there are already some approaches for improving process recording and process modeling through automation. One possibility is to generate a Petri net from XML-based documents. With higher Petri nets (e.g. Pr/T net), switching conditions can be further specified on the basis of more distinguishable markers. For example, if a marker in the "new invoice" position represents a specific invoice and the position is followed by a "grant discount" transition, it can be specified in the transition that only invoices over €500 should receive a discount. In this assignment, you will therefore develop a concept for how these business rules can be proposed to the modeler based on the documents. For example, the documents must be compared with each other and the cases in which a document only occurs must be checked. In the thesis, you will look at different approaches in order to subsequently develop a concept and implement it as a prototype. Language models, for example, could be used to extract information. However, since semi-structured data is assumed, non-NLP-based methods could also be used.

Please apply under the following link: https://portal.wiwi.kit.edu/forms/form/Bewerbung_Abschlussarbeit_AIFB-BIS