process mining bachelor thesis

• Contact/Imprint

Humboldt-Universität zu Berlin - Faculty of Mathematics and Natural Sciences - Process Management and Information Systems

Bachelor and master thesis.

General Information

Our team offers bachelor and master thesis topics as well as student projects to be written in English. 

Student may apply for a thesis or a study project during within two application windows in a year, in which new topics are made available. The first window is open from February 1st until April 1st. The second window is open from July 1st until October 1st.

Here you can find the information on how to write a thesis with us. Slides are available here ( part I ,  part II ), and recordings here ( part I ,  part II ). 

Furthermore, find below a summary of guidelines for working on your thesis with us.

Expression of Interest in a Topic (Thesis or Study Project)

If you are interested in one of the topics, please send an email expressing your interest to Dr. Saimir Bala (firstname[dot]lastname[at]hu-berlin.de).  Please explain  why this topic is interesting for you and how it fits your prior studies. Also explain what are your strengths in your studies and in which semester of your studies you are.

Process Overview

• There are two main time windows in which the team proposes new topics: Feb 1st ­– Apr 1st and Jul 1st – Oct 1st
• Within these windows students can apply for an open topic (see list of open topics below)
• Application is done by sending an email to Dr. Saimir Bala (firstname[dot]lastname[at]hu-berlin.de).
• We collect your applications and make a topic-student assignment in two rounds. First round on March, second round after the deadline. For the winter session, we have two rounds (Sep, Oct).
• Once a student has been matched to a supervisor, a kick-off meeting is scheduled to scope the topic.
• Then, students must submit a research proposal to the supervisor within a month.
• If the proposal is graded as passed, the supervision is officially registered
• Once the thesis work is concluded, the thesis defense is scheduled within a dedicated defense slot.

Important Dates

01.07.2024: New topics released. Students can express their interest.

02.09.2024: Topic assignment (1st round)

01.10.2024: Expression of interest deadline

02.10.2024: Topic assignment (2nd round)

Milestones :

- Kick-off: shortly after assignment round

- Research proposal submission deadline first round (1 month after official kick-off)

- Official start (if proposal sufficient)

- Thesis delivery 

- Grading & Defence

Please consider the following hints and guidelines for working on your thesis:

• Templates for thesis and proposal: https://www.informatik.hu-berlin.de/de/studium/formulare/vorlagen
• Page limits are as follows
• page limit is for Bachelor Informatik 40 pages and for Kombibachelor Lehramt Informatik 30 pages
• page limit is for Master Informatik 80 pages and for Master Information Systems 60 pages
• The limits do not include cover, table of content, references, and appendices.

Prerequisites

The candidate is expected to be familiar with the general rules of writing a scientific paper. Some general references are helpful for framing any thesis, no matter which topic:

• Wil van der Aalst:  How to Write Beautiful Process and Data Science Papers?  Archive Report (2022).
• Jan Recker:  Scientific Research in Information Systems: A Beginner's Guide  . Springer, Heidelberg, Germany (2021).
• Jan Mendling, Benoit Depaire, Henrik Leopold: Methodology of Algorithm Engineering . Archive Report (2023).
• Claes Wohlin, Pär Runeson, Martin Höst, Magnus Ohlsson, Björn Regnell, Anders Wesslén  Experimentation in software engineering  . Springer Science & Business Media (2012).
• Ken Peffers, Tuure Tuunanen, Marcus A. Rothenberger, Samir Chatterjee:  A Design Science Research Methodology for Information Systems Research  . J of Management Information Systems 24(3): 45-77 (2008).
• Barbara Kitchenham, Rialette Pretorius, David Budgen, Pearl Brereton, Mark Turner, Mahmood Niazi, Stephen G. Linkman:  Systematic literature reviews in software engineering - A tertiary study  . Information & Software Technology 52(8): 792-805 (2010).
• Lagendijk, Ad.  Survival Guide for Scientists: Writing, Presentation, Email  . Amsterdam University Press (2008).
• Adam LeBrocq: Journal of the Association for Information Systems Style Guide.  https://aisel.aisnet.org/cais/cais_style_guide.pdf

In agreement with the supervisor an individual list of expected readings should be studied by the student in preparation of the actual work on the thesis.

The grading of the thesis takes various criteria into account, relating both to the thesis as a product and the process of establishing its content. These include, but are not limited to:

• Correctness of spelling and grammar
• Aesthetic appeal of documents and figures
• Compliance with formal rules
• Appropriateness of thesis structure
• Coverage of relevant literature
• Appropriateness of research question and method
• Diligence of own research work
• Significance of research results
• Punctuality of work progress
• Proactiveness of handling research progress

Recent Topics

The following topics are available within the current application window.

Topic 1: Process prediction using object-centric event log (Bachelor/Master)

Business process prediction involves forecasting specific details, such as the next activity to be performed, the time remaining for the completion of a process instance, or key process indicators, for an ongoing process instance. Currently, the techniques rely on XES event logs as input data. However, the field of process mining is shifting towards utilizing object-centric event logs, which offer a comprehensive multidimensional view of the data. Despite this advancement, object-centric event logs have been underutilized as input for process prediction.

Research problem : The core research problem addressed is: How can process prediction benefit from an object-centric event log? The aim is to propose a method to process prediction using object-centric event log.

Requirements : The candidate must have previous knowledge of process mining and software development. Further desirable requirements are pro-activity and self-organization.

Initial references

• An Empirical Investigation of Different Classifiers, Encoding, and Ensemble Schemes for Next Event Prediction Using Business Process Event Logs. ACM Trans. Intell. Syst. Technol. 11(6): 68:1-68:34 (2020)
• Uncovering Object-Centric Data in Classical Event Logs for the Automated Transformation from XES to OCEL. BPM 2022: 379-396
• Benedikt Knopp, Wil M. P. van der Aalst:Order Management Object-centric Event Log in OCEL 2.0 Standard. Zenodo, 2023

Supervisor: Kate Revoredo

Topic 2: Causation discovery for process prediction (Bachelor/Master)

Business process prediction involves forecasting specific details, such as the next activity to be performed, the time remaining for the completion of a process instance, or key process indicators, for an ongoing process instance. Currently, most techniques rely on the order in which the events happened without considering the cause-effect relation among them.

Research problem: The core research problem addressed is: How can process prediction benefit from the cause-effect relation among the events? The aim is to propose a method to discover the cause relation among events and use this information for process prediction.

Requirements: The candidate must have previous knowledge of process mining, statistics, and software development. Further desirable requirements are pro-activity and self-organization.

• Jens Brunk, Matthias Stierle, Leon Papke, Kate Revoredo, Martin Matzner, Jörg Becker: Cause vs. effect in context-sensitive prediction of business process instances. Inf. Syst. 95: 101635 (2021)
• Pearl,J.(2011).Bayesiannetworks.

Topics 3: Uses of Models in Agile Software Development (Bachelor/Master)

Motivation & problem : Modeling is a key topic in software engineering. In software development projects, among other aspects, modeling supports the developer in understanding the design by providing an overview and a tool for communication with fellow developers and other stakeholders. The benefits of models for supporting system analysis and design activities have been highlighted regarding their cognitive effectiveness, often in the context of traditional methodologies. However, these benefits have also been discussed in the agile scene, but it is still not clear to what extent models are used in agile software development projects.

Objectives : conduct a systematic review of the literature, identify the uses of models in agile software development, categorize and prioritize them, and propose a framework to support agile software development based on these findings. The findings shall be evaluated according to the perspective of practitioners.

Prerequisites : (1) Basic knowledge of agile software development methodologies; (2) Intermediate knowledge of models used in software development; (3) Pro-activity, self-organization, attention to detail (desirable).

Initial References:

• Ambler, Scott W. The object primer: Agile model-driven development with UML 2.0. Cambridge University Press, 2004.
• Alfraihi, Hessa Abdulrahman A., and Kevin Charles Lano. "The integration of agile development and model driven development: A systematic literature review." The 5th International Confrence on Model-Driven Engineeing and Software Development (2017).
• Wagner, Stefan, Daniel Méndez Fernández, Michael Felderer, Antonio Vetrò, Marcos Kalinowski, Roel Wieringa, Dietmar Pfahl et al. "Status quo in requirements engineering: A theory and a global family of surveys." ACM Transactions on Software Engineering and Methodology (TOSEM) 28, no. 2 (2019): 1-48.
• Petre, Marian. "UML in practice." In 2013 35th international conference on software engineering (icse), pp. 722-731. IEEE, 2013.

Supervisor: Cielo González

Topic 4: Collaborative business-model-driven tool for agile software development projects (Bachelor/Master)

Motivation & problem: Agile software development methodologies and frameworks have changed the way software is created and are widely supported and used. However, this does not mean there are no challenges that jeopardize the principles of agile methodologies, increasing the failure rate of agile software development projects. This situation highlights the need for cohesive solutions. The use of business process models in the agile software development context emerges as a promising option due to their ability to facilitate communication and share knowledge.

Objectives : analyze, design, implement and evaluate a collaborative business-model-driven tool for agile software development projects. The objectives will be adapted to align with the student's study goals.

Prerequisites :(1) Knowledge of agile software development methodologies; (2) Knowledge of business process models; (3) Knowledge in frontend (e.g., JavaScript and TypeScript); (4) Knowledge in Java; (5) Knowledge in databases (e.g., PostgreSQL, mongoDB); (6) Pro-activity and self-organization.

Initial references:

• Moyano, Cielo González, et al. "Uses of business process modeling in agile software development projects." Information and Software Technology 152 (2022): 107028.
• Trkman, Marina, Jan Mendling, and Marjan Krisper. "Using business process models to better understand the dependencies among user stories." Information and software technology 71 (2016): 58-76.

Topic 5: Fair and Diverse Sampling of Event Logs (Bachelor and Master)

The sampling of large event logs, i.e. the selection of subsets of data, has been proposed as one possible solution to tackle the runtime requirements of process analysis tasks and to aid the understandability of data sets, that are too complex to analyze as a whole. In this context a diverse sample is one, which properly reflects the diversity or complexity of the event log, while a fair sample is one, which ensures, that each value is represented properly. For both quality criteria, approaches have been proposed, to solve these problems, for instance by optimizing subset selection functions or employing determinantal point processes.

In this thesis the student will:

* conduct research on existing approaches for diverse and fair sampling * implement selected approaches for the diverse and fair sampling for event logs * evaluate the implemented algorithms comparatively in terms of quality and efficiency

The student is expected to have existing knowledge in optimization or statistics or sampling.

Initial References

• Kabierski, M., van der Aa, H., and Weidlich, M. (2020). Sampling and approximation techniques for efficient process conformance checking. Information Systems. 104. 101666. http://dx.doi.org/10.1016/j.is.2020.101666
• Moumoulidou, Z., McGregor, A., and Meliou A.. Diverse Data Selection under Fairness Constraints. In 24th International Conference on Database Theory (ICDT 2021). Leibniz International Proceedings in Informatics (LIPIcs), Volume 186, pp. 13:1-13:25, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2021), https://doi.org/10.4230/LIPIcs.ICDT.2021.13
• Celis, L., Vijay, K., Straszak, D., Deshpande, A., Kathuria, T., and Vishnoi, N. (2018). Fair and Diverse DPP-based Data Summarization. https://arxiv.org/abs/1802.04023

Supervisor: Martin Kabierski

Topic 6: Estimating Saturation in Qualitative Studies (Master)

Grounded theory is a research methodology usually applied in qualitative analysis. It involves the collection of data (usually through interviews, surveys, ...), and the deduction of concepts, categories, and ultimately theories that emerge from the collected data. A central question to this iterative data collection-evaluation process is when one should stop collecting data, which ideally is at the point of saturation, i.e. when no new information is gained from new interviews. Determining when exactly this point has been reached is an ongoing topic of discussion and research. Species richness estimators, that estimate the completeness of samples, could be utilized to give saturation estimates that are data-driven and grounded in statistics.

In this thesis, the student will:

- assess the applicability of species richness estimation for determining saturation in qualitative studies - implement and apply the estimators to qualitative interview data - evaluate the feasibility of the approach and discuss potential limitations

The student is expected to have understanding of statistics and and optionally preliminary experience in the analysis of qualitative data. We note, that the student is not expected to collect data for the thesis, as this data will be provided by us.

• Strauss, A., & Corbin, J. (1994). Grounded theory methodology: An overview. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (pp. 273–285). Sage Publications, Inc. https://www.depts.ttu.edu/education/our-people/Faculty/additional_pages/duemer/epsy_5382_class_materials/Grounded-theory-methodology.pdf
• Saunders, Benjamin, et al. (2018). Saturation in qualitative research: exploring its conceptualization and operationalization. In: Qual Quant 52 (pp. 1893-1907). Springer. https://doi.org/10.1007/s11135-017-0574-8
• Gotelli, Nicholas & Chao, Anne. (2013). Measuring and Estimating Species Richness, Species Diversity, and Biotic Similarity from Sampling Data. 10.1016/B978-0-12-384719-5.00424-X. https://www.uvm.edu/~ngotelli/manuscriptpdfs/Gotelli_Chao_Encyclopedia_2013.pdf

Topic 7: Runtime Prediction of Alignment Construction Algorithms (Bachelor/Master)

Conformance Checking relates a process model to recorded instances of the execution of the process, typically stored in event logs, to determine where expected and actual behaviour deviate from each other. In this context alignment algorithms are regarded as the de facto standard method, due to their interpretability and accuracy in highlighting precise problem areas in the process. Yet, typically run times for alignment construction are prohibitively large, typically caused by a handful of traces in the log, for which the construction of an alignment is especially complex. One possible solution to this problem could lie in predicting the expected runtime of aligning a trace to the model, for instance using regression-based methods and then ignoring traces, that are expected to take long.

In this thesis, the student will: - assess the factors that influence the runtime of alignments - derive a methodology for predicting the runtime of alignment construction between event logs and process models - evaluate the accuracy of the predictor

The student is expected to have existing knowledge of process mining, conformance checking, and basic knowledge of regression analysis, or willingness to learn about these topics under guidance of the supervisor.

• Chapter 1, 2 and 7 in Carmona, J., van Dongen, B., Solti, A., & Weidlich, M. (2018). Conformance checking. Switzerland: Springer. https://doi.org/10.1007/978-3-319-99414-7
• Backhaus, K., Erichson, B., Weiber, R., Plinke, W. (2016). Regressionsanalyse. In: Multivariate Analysemethoden. Springer Gabler, Berlin, Heidelberg. https://doi.org/10.1007/978-3-662-08893-7_1

Topic 8: Event Log Privacy Auditing (Master)

Event Logs are often anonymized to ensure privacy. Most anonymization algorithms use formal privacy guarantees such as differential privacy. The issue with these techniques is that the algorithm or implementation could contain errors. Consequently, anonymized event logs might not have the targeted privacy guarantee and the individuals involved in the dataset might have a higher privacy loss than expected. Differential Privacy Auditing allows to check if an algorithm fulfils the differential privacy guarantee. The aim of this thesis is to adjust know Differential Privacy Auditing techniques to Event Log Anonymization and evaluate anonymization techniques from the process mining domain.

• Stephan A. Fahrenkrog-Petersen, Martin Kabierski, Han van der Aa, Matthias Weidlich: Semantics-aware mechanisms for control-flow anonymization in process mining. Inf. Syst.114: 102169 (2023)
• https://research.google/blog/dp-auditorium-a-flexible-library-for-auditing-differential-privacy/

Supervisor: Stephan Fahrenkrogh-Petersen

Topic 9: Analysis of theoretical explanations and scientific theories on transitioning from dashboards to decision making in organizational contexts (Bachelor)

This bachelor thesis seeks to analyze theoretical explanations and scientific theories concerning the transition from dashboards to decision-making processes. Dashboards are widely used tools in organizational contexts for decision-making. The study aims to examine the levels of management where dashboards are employed and how they contribute to the decision-making process within organizations. Literature :

• Burstein, F., & Holsapple, C. W. (2008). Handbook on Decision Support Systems 2. https://www.academia.edu/83497312/Handbook_on_Decision_Support_Systems_2
• Maynard, S., Burstein, F., & Arnott, D. (2001). A multi-faceted decision support system evaluation approach. Journal of Decision Systems, 10(3–4), 395–428. Mintzberg, H., Raisinghani, D., & Theoret, A. (1976). The Structure of “Unstructured”
• Decision Processes. Administrative Science Quarterly, 21(2), 246. https://doi.org/10.2307/2392045

Supervisor: Kristina Sahling

Topic 10: Visualizing Cyclic Time Arrangements in Process Graphs (Bachelor/Master)

Time is essential to understanding processes, yet most process mining approaches are limited to depicting time within a process graph as textual cues or color schemes. Adapting the visual appearance of process graphs to various time arrangements may enhance the accessibility for finding bottlenecks or delays. An example is aligning process graphs along a linear timeline [1]. In cases where processes involve repetitive patterns, such as in chronic health care or crop management, a cyclic arrangement may be useful. However, for the latter, an adequate solution in process mining is needed.

This thesis aims to develop and exemplify a design method for a visual solution in process mining that allows for exploring a cyclic time arrangement in a process graph. We will adapt the research objectives to align with the experience and study goals of the student.

Initial References :

• H. Kaur, J. Mendling, C. Rubensson, and T. Kampik, “Timeline-based Process Discovery,” CoRR, abs/2401.04114, 2024. Available: https://doi.org/10.48550/arXiv.2401.04114
• A. Yeshchenko and J. Mendling, “A Survey of Approaches for Event Sequence Analysis and Visualization using the ESeVis Framework.,” CoRR, abs/2202.07941, 2022. Available: https://arxiv.org/abs/2202.07941
• W. Aigner, S. Miksch, H. Schumann, and C. Tominski, Visualization of Time-Oriented Data. in Human-Computer Interaction Series. London: Springer London, 2011. Available: https://doi.org/10.1007/978-0-85729-079-3.

Supervisor: Christoffer Rubensson

Topic 11: Advanced Resource Analysis in Process Mining (Bachelor/Master)

In the last decade, process mining techniques have been developed to study human behavior in event data, such as the strength of collaboration between co-workers or even stress levels at a workplace. Since measuring human behavior is complex, this is a welcoming alternative to more labor-intensive methods like surveys. Still, most techniques are relatively simple but could be improved by applying theoretical frameworks from social science.

This thesis aims to develop a resource analysis approach (e.g., a metric, a concept, or a framework) in process mining grounded in an existing theory from social science. We will adapt the research objectives to align with the experience and study goals of the student.

• J. Nakatumba and W. M. P. van der Aalst, “Analyzing Resource Behavior Using Process Mining,” in Business Process Management Workshops. BPM 2009. Lecture Notes in Business Information Processing, S. Rinderle-Ma, S. Sadiq, and F. Leymann, Eds., Berlin, Heidelberg: Springer Berlin Heidelberg, 2010. Available: https://doi.org/10.1007/978-3-642-12186-9_8.
• A. Pika, M. Leyer, M. T. Wynn, C. J. Fidge, A. H. M. Ter Hofstede, and W. M. P. Van der Aalst, “Mining Resource Profiles from Event Logs,” in ACM Transactions on Management Information Systems, vol. 8, no. 1, 1:1-30, 2017. Available: https://doi.org/10.1145/3041218.
• Z. Huang, X. Lu, and H. Duan, “Resource behavior measure and application in business process management,” in Expert Systems with Applications, vol. 39, no. 7, 6458–6468, 2012. Available: https://doi.org/10.1016/j.eswa.2011.12.061.

Topic 12: Anthropomorphic Perceptions of Large Language Models: what is the gender of ChatGPT and its Counterparts? (Bachelor/Master)

Description : In today's digital era, Large Language Models (LLMs) like ChatGPT are transforming the way we interact with technology, often blurring the boundaries between machine and human cognition. This thesis delves into the intriguing realm of anthropomorphism, the human tendency to attribute human-like qualities to non-human entities. Specifically, this research aims to uncover laypeople's underlying beliefs and implicit conceptions about ChatGPT and similar models concerning an implicit gender attribution. By designing and conducting a survey, the thesis will gain insights into individuals' perception of these cutting-edge technologies. The findings can potentially illuminate not only our relationship with LLMs but also the broader implications of human-machine interactions in an increasingly AI-driven world.

• Deshpande, A., Rajpurohit, T., Narasimhan, K., & Kalyan, A. (2023). Anthropomorphization of AI: Opportunities and Risks (arXiv:2305.14784). arXiv. https://doi.org/10.48550/arXiv.2305.14784
• Farina, M., & Lavazza, A. (2023). ChatGPT in society: Emerging issues. Frontiers in Artificial Intelligence, 6. https://www.frontiersin.org/articles/10.3389/frai.2023.1130913
• Aşkın, G., Saltık, İ., Boz, T. E., & Urgen, B. A. (2023). Gendered Actions with a Genderless Robot: Gender Attribution to Humanoid Robots in Action. International Journal of Social Robotics, 15(11), 1915–1931. https://doi.org/10.1007/s12369-022-00964-0

Supervisor: Jennifer Haase

HU on the internet

• Humboldt University on Facebook
• Die Humboldt-Universität bei BlueSky
• Humboldt University on Instagram
• Humboldt University on YouTube
• Humboldt University on LinkedIn
• RSS-Feeds of the Humboldt University
• Humboldt University on Twitter

• Skip to content  (access key: 1)
• Skip to Search  (access key: 2)

Institutes, schools, other departments, and programs create their own web content and menus.

To help you better navigate the site, see here where you are at the moment.

Process mining as an enabler for business process modeling

Description.

Due to global competition and increasingly complex business models, the pressure on companies is constantly increasing. Value chains are becoming longer, delivery bottlenecks have to be taken into account and customers' demands for high product quality are increasing. Apart from the requirements resulting from the different business areas, information systems are becoming more and more important in companies and business processes are increasingly relying on the data generated by such information systems.

The aim of this thesis is to answer the question of how business processes can be created in a data-driven manner using process mining in order to gain a better insight into the company-wide target processes. In particular, the points that need to be considered when introducing data-driven process models and how such a process model can look in practice were discussed. Apart from the realization of such process models, they were compared with conventional process modeling and the corresponding advantages were worked out.

The process model was carried out on the basis of a fictitious implementation using the example of an order-to-cash process. Important topics in connection with the actual implementation were worked out and presented. In this case, the scenario describes a concrete application example.

In order to assess whether process mining actually provides added value compared to conventional process modeling methods, challenges in this context were identified by means of a literature review. Subsequently, suggestions were made as to how these challenges can be solved using process mining.

Philipp Werning

o.Univ.-Prof. Dipl.-Ing. Dr. Christian Stary

Johannes Kepler University Linz

Altenberger Straße 69

4040 Linz, Austria

Use of cookies

Our website uses cookies to ensure you get the best experience on our website, for analytical purposes, to provide social media features, and for targeted advertising. This it is necessary in order to pass information on to respective service providers. If you would like additional information about cookies on this website, please see our data privacy policy .

Required cookies

These cookies are required to help our website run smoothly.

Web statistics cookies

These cookies help us to continuously improve our services and adapt our website to your needs. We statistically evaluate the pseudonymized data collected from our website.

Marketing cookies

These cookies help us make our services more attractive to you as well as optimize our advertising and website content. We analyze and evaluate pseudonymised data collected from our website.

• Student Publications

Files in This Item:

Page view(s)

Download(s).

Google Scholar TM

Haw katalog, note about this record, send a note about this record.

Your Message was sent successfully.

Couldn't send your message. Please contact the Support.

Items in REPOSIT are protected by copyright, with all rights reserved, unless otherwise indicated.

Bachelor Thesis – Predictive Monitoring in Process Mining

Predictive Monitoring in Process Mining is the discipline of predicting business processes according to the outcome or the remaining time. The relevant data source is an event log which includes all activity data regarding a process from an ERP system. Either an abstraction of the entire workflow is created for generative models, or individual traces for each entity are extracted for descriptive models.

Whereas much research has been completed in comparing quantitative results, I provide a qualitative overview of the most important literature in this field. The thesis further focuses on the technical implementation and parameters one could apply to create such procedures and prediction models. The purpose is to ensure a first level of understanding on the different realization methods of predictive process monitoring (PPM) and to understand how the authors conducted their research procedure.

As the following summary can not go into each technical detail, it will only revolve around the most comprehensive parts of my thesis . This summary will thus include an explanation about generative and descriptive models, including the pre-processing of transaction data necessary for descriptive models, as well as the integration of different information contexts. I will finally outline my findings.

Generative Models

Generative models produce a representation or abstraction of business processes that can be calculated from observations, i.e. past process transactions. These generative methods are process-aware in the predictive process monitoring context. They belong to the traditional statistics. From data, a world state is estimated that is finally taken as input in a decision theory to predict results.

Many generative models have been found in the PPM literature. Especially those approaches perform best that, firstly, put an emphasis on including activity data, which will be covered in the section about contextual information, and secondly, that can represent the control flow i.e., the process as accurately as possible. Generative models are of particular importance in research, as handling missing information is an overall problem in ERP production environments, and at the same time, a particular strength of generative approaches.

Descriptive Models

Introduction.

Discriminative models focus on the boundaries of decisions. Their goal is to predict the property of the data. In contrast to generative PPM methods, they all follow a comprehensive workflow structure: first process sequences are extracted and filtered, then, they must be bucketed, then these traces are encoded, and finally a model is applied that can work with the data. This can be explained as they belong to the modern machine learning techniques. These can directly predict results based on preprocessed data, without the detour of creating a comprehensive abstraction of the real world.

Pre-Processing of Transaction Data

Trace bucketing is a procedure to divide traces with different methods. The resulting buckets serve as separate training sets for different models. Concerning the bucketing procedures, the simplest one is the so-called single bucket approach. It should be pursued if the initial training set is limited to the extent that the model could not be appropriately trained. Particularly, neural networks also do not require bucketing, as large amounts of data can be used for training. As another example, prefix length bucketing specialize a model for different points in time of process execution.

Sequence encoding is about treating traces as complex symbolic sequences, each carrying a data payload. The sequence encoding can be situated after trace bucketing, whereas the traces need to be encoded in machine readable language i.e., in structured data. In their simplest form, they are simple symbolic sequences which do not include activity attributes. In their more sophisticated form, they describe related data in a static or dynamic manner. Last state encoding, for example, only includes the attributes of the last event of a trace, ignoring the evolution to that point.

There is no comprehensive literature review that compares all possible combinations of trace bucketing and sequence encoding according to their performance. Nevertheless, it must be said that a holistic comparison of all different combinations can be quite an extensive task. Additionally, since encoding techniques do not exclude themselves, the significance of such a paper could be limited as combinations of multiple encoding techniques would also need to be included to gain a complete overview. Lastly, there is no universal solution as different models require different processing workflows.

Classification, Regression and Structured Algorithms as Predictors

Classifiers categorize data into distinct labels. As categorization problems are the most fundamental, they will serve as the basic building blocks for the more complex issues. Next event and outcome predictions necessitate classifiers and both revolve around the same method.

Regressors have the commonality to predict a continuous-valued output instead of one from a finite set, in contrast to classifiers. In PPM, regressors are applied to predict the remaining time.

Structured predictors output entire sequences as predictions i.e., remaining paths of a sequence in our case. Structured predictors only include Long Short-Term Memory (LSTM) models, according to the current state of research. LSTM models belong to the Recurrent Neural Networks that are used in contexts in which the input or output range of data is not known beforehand.

Admittedly, in the implementation, there are classifiers that can serve as either classifier or regressor and vice versa. Most advanced approaches use a combination of classification and regression, as well as descriptive and generative models. For other discriminative models besides LSTM, data preprocessing requires a greater effort because they must be adjusted to work with sequence information. Furthermore, neural networks are relevant for large datasets with high dimensions. In general, discriminative models seem to be performing better than generative ones.

Integration of Contextual Information

To classify the scientific work on contextual factors as comprehensively as possible, a classification can be made based on the extent of information, whereby cause and effect get increasingly unclear the larger the scope becomes. Contextual information includes intra-case, inter-case, social, external, and spatial information to a process. At least one paper has been found for each aspect.

Intra-case contextual information is information captured in an activity itself. Selected attributes to an event include the vendor, the document type, and the item category which can all be incorporated into the event log. Including intra-case information is the common approach for descriptive models.

In contrast, inter-case features, or process context information, include dependencies between different activities. The idea behind it is that predictions are also dependent on the execution of events in the same period e.g., competing for the same resource in the process context. Generative models take advantage of these features.

Regarding the social context , these factors encompass the way humans and automated agents interact within a particular organization to execute process-related activities. Friction between individuals may delay process instance, and the speed at which people work may vary.

Furthermore, external context information, as the news, can be included. Regarding this example, the information includes the fields of both the social and the external context, as the external information like the economic climate is having an impact on the people’s sentiments, such as their level of stress.

The spatial context can be seen as middle child between process, social, and external context. One successful model has been achieved by including the locations of process traces.

Beyond the intra- and inter-case contextual information, the other specialized approaches have not yet been implemented in the larger scope of PPM papers. Thus, the social, external and spatial contexts necessitate further research. Among others, external contextual information like social media, blogs etc. could be evaluated. In addition, a spatio-temporal model could be applied to further increase the remaining time accuracy.

Predictive Process Monitoring, a forecasting business operation tool, can be situated between Process Mining, as a form of comprehending the current state of a process to undertake strategic measures, and Prescriptive Process Monitoring, that interprets the prediction so that measures are described to pilot a process outcome. The review consists of both generative and discriminative models; the latter can be further broken down into prefix bucketing and encoding, implementing a machine learning model, and deciding which information context to include in the design.

All papers, that were just part of a related field of study and not necessary to implement a prediction model, were excluded. The thesis is aimed at providing a counterbalance for other quantitative reviews in this field. A large variety of findings have emerged: both generative and descriptive models play a role in current research, there is a lack of papers in the field of process suffix prediction as well as structured predictors, and the integration of a larger information context, e.g., from external sources, should be an integral part of future research.

• AI powered Analytics
• P2P & O2C
• Cash Recovery
• Fraud Detection
• Trade Compliance
• SAP Authorization Check
• VAT ID Check
• Master Data Quality Check

Process Mining

• Internal Audit
• Internal Controls
• Risk and Compliance
• Procurement
• IT & Data Science
• Public Sector
• Diligent SAP Connector
• Diligent ACL Analytics
• Diligent One Platform
• dab AnalyticSuite
• dab Nexus Loom

Our analytics platform for SAP data - smart, integrated, user-friendly

With dab Link we seamlessly integrate analytic results with your SAP system.

Perform data analytics from a portfolio of around 300 questions at the push of a button.

The ultimate SAP data analytics solution using Microsoft Fabric for seamless integration and advanced insights.

“Working at dab means working with professional and talented individuals who make great friends. ”

Daniela (Teamlead Marketing)

• ————————————
• Events & Webinars
• +49 991 991 358-0
• [email protected]

Challenges and Potentials of Process Mining - Bachelor Thesis

May 15, 2020

Philipp Kiencke

Global Partner Manager

As most of you know, dab: Daten – Analysen & Beratung GmbH is a spin-off from the Deggendorf Institute of Technology (THD). We have therefore always had working students who work for us and sometimes also write their final thesis with us.

Mathias Kirsch, for example, wrote his bachelor’s thesis last year on the topic of ‘Challenges and potentials of process mining’. This thesis sheds light on the topic of process mining on a scientific basis. Among other things, the methods, challenges and limitations, the potential of process mining and the degree of maturity of various algorithms are analysed.

The work is of interest to anyone who, in addition to colourful graphs, would like to understand how exactly process mining works, which algorithms are used in the background and which problems arise with this form of process analysis. We did not want to withhold this very good work from you and have attached it at the end of this blog post.

We would like to thank Mathias Kirsch for making his work available to us and wish him all the best for his Master’s degree.

Content Listing

Share This Article

Other articles

Data Analytics

Throwback Webinar AI-Powered Cash Back Analytics

Data Extraction

dab Nexus Content Suite Plus

Authorization Assurance

dab:AuthorizationAssurance

To stay up to date with all the news, subscribe to our newsletter!

Request a demo.

I have read the  privacy policy .

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply .

Registration

Become a partner.

We use cookies to optimize our website for you and to continuously improve it.

We and our partners use cookies to enhance your experience and understand how our digital platforms are used. See our privacy policy for more information. Privacy policy      Imprint

Skip to Content Skip to Main Navigation Skip to Footer Skip to Search

Chair of Process and Data Science

Faculties and institutions.

You Are Here: Thesis Projects

Faculties and Institutes

• Mathematics, Computer Science and Natural Sciences Faculty 1
• Architecture Faculty 2
• Civil Engineering Faculty 3
• Mechanical Engineering Faculty 4
• Georesources and Materials Engineering Faculty 5
• Electrical Engineering and Information Technology Faculty 6
• Arts and Humanities Faculty 7
• Business and Economics Faculty 8
• Medicine Faculty 10

Find Institute:

Institutions.

• University Library
• Center for Teaching and Learning Services
• Athletics and Recreation
• Central University Administration
• All Institutions

You Are Here:

Thesis projects, sub-navigation.

• PADS - Teaching
• PADS Excellence Class
• Pro-Seminars
• Bachelor Thesis - Extracting Sub WF-nets with Desirable Properties
• Bachelor Thesis - Guided Process Discovery Through Activity Scoring
• Master Thesis - ProcessGraphQA: Encoding Process Graphs for Large Language Models (LLMs)
• Master Thesis - Decomposing (Modulizing) Sound Free-choice WF-nets
• Bachelor Thesis - Mining Patterns in Performance Spectrum
• Master Thesis - Integrating Graph Partitioning Techniques to Reduce the Computational Cost of Cut Search Algorithms
• Master Thesis - Generalised Stochastic Petri Nets
• Master Thesis - Leveraging LLM for Translating Process Models into Text: A Study on Prompting Strategies and Quality Evaluation
• Master Thesis - Investigating Novel Techniques for Case Reconstruction in SLURM Logs
• Bachelor Thesis - Visualizing Object-Centric Petri Nets
• Master Thesis - Discover Medication Change Paths
• Master Thesis - Strategic Sampling and Abstraction in Process Mining: Navigating Complexity for Cohesive Operational Insights
• Bachelor Thesis - Analyzing Project Workflows through Git History and Gitlab Tickets
• Bachelor Thesis - Extending the GAIA-X Eclipse Data Space Connector with an Event Logger
• Completed Theses
• Depth Oral Colloquium

HiWi Positions

• New Process Mining Course by PADS and Celonis
• Coursera Data Science in Action
• PADS/Celonis Academic Center of Excellence

We are continuous looking for students interested in doing a thesis project in the PADS group. We are eager to supervise both Bachelor and Master Thesis projects. However, given the many requests, we need to be selective and align such projects to our expertise and research goals. Therefore, we require people to have a data-science mindset and an interest in processes and dynamic behavior. Most of the thesis projects are in the field of process mining. Therefore, we require potential students to show that they have an understanding of the existing field. Having process mining knowledge, e.g., obtained in the Business Process Intelligence (BPI) course at RWTH running in the second semester (SS) or the Coursera MOOC on Process Mining, is desirable. This makes it easier to discuss possible thesis projects in areas such as process discovery, conformance checking, performance analysis, predictive process analytics, automated process improvement, responsible process mining, etc. If you can still elect your courses, we also recommend the master courses Introduction to Data Science (WS) and Advanced Process Mining (SS). Also, check out the Seminars and practical assignments running every semester.

If you are interested, please look through the list of available theses on the website. If you have a good resume, good grades, and special expertise and experience in the above-mentioned area but were unable to find a suitable thesis online, please fill out the thesis inquiry form and send it to Viki Peeva [email protected] along with a brief motivation, your CV, and grades. She is in charge of the thesis applications for the PADS group. She will ask further questions to determine whether we can provide you with a thesis that meets your requirements and expertise.

External Thesis Projects

We do NOT supervise external thesis projects unless there is an existing collaboration (e.g., with Celonis) and the topic is related to what we do (e.g., process mining). We welcome bright students who want to specialize in the topics we cover (selected topics on the interface between data science and process science, in particular, process mining). However, we only host students who know about these topics and have shown commitment to dive deeper. Moreover, we only supervise students in our expertise area, that are working on assignments that are carefully defined by us (or in a collaborative effort). Students deserve good supervision. Note that this is also the general policy of the Computer Science department . Therefore, do not be misled by groups outside of Computer Science offering such thesis projects.

Note that the chair also has many possibilities for HiWi jobs related to process mining (either within Fraunhofer FIT or RWTH). However, this is also reserved for excellent students that have acquired a background in process mining and that want to continue a career in process mining (in industry or academia). We discourage people to apply for HiWi jobs without being able to show relevant experience. HiWi opportunities typically follow (or run in parallel) with a thesis project or excellent performance in one of our courses.

last updated: 03/06/2024

• RWTH Main Page
• Faculty of Mathematics, Computer Science and Natural Sciences
• Department of Computer Science
• Contact and Maps
• Site Credits
• Privacy Policy
• Accessibility Statement

Social Media

• Fraunhofer Gesellschaft
• Process Mining
• Data Science Center Eindhoven

Javascript is required

You must have JavaScript enabled in order to access the Celonis website. The web browser you're using either doesn't support JavaScript, or it's being blocked. Please check your browser's help resources to learn whether your browser supports JavaScript and how to allow JavaScript.

Celonis Student Thesis and Research (STaR) Program

We highly welcome bachelor or master students, or doctoral candidates interested in writing their thesis, project study, developer study or research seminar papers on process mining and related aspects. It is our mission to empower the next generation of process miners, and this means that we aim to support you and your research as well as we can.

We believe that process mining can make an impact anywhere in the world, and in almost any discipline. Therefore, we explicitly encourage students of all nationalities and backgrounds to join our Celonis STaR Program!

Join the Celonis STaR Program

Sign up for the Celonis STaR Program by completing the sign-up form below and receive all the information you may need to successfully write your thesis, project or developer study or research seminar papers:

Directions on how to get your hands on the academic version of the Celonis Software for free

Training and material on how to use, modify, work and upload data within our Celonis Software in the Celonis Academy

Access to our vibrant community of process mining enthusiasts on our Celopeers Forum

Help from our Academic Research & Innovation Coordinator

 Moreover, if you need additional resources from our side, such as access to further experts, interviews, or if you are interested in starting a project with another academic partner or one of our customers, we could help you. In this case, we need a One-Page pitch from you, explaining what you need from our side, the scientific excellence of the topic, and its relevance for Celonis (and/or other involved stakeholders). You will receive a template for this pitch after signing up for Celonis STaR Program. The pitches are evaluated internally at least twice a month, and you will hear from us shortly. 

Sign up here for the Celonis STaR Program!

Sign up here to receive instant access to our resources.

Frequently Asked Questions

Are there any limitations as to which countries students can come from or which universities they are working with.

We especially encourage students from outside of Europe to take part in the Celonis STaR Program. All our materials are available in English.

However, Celonis is legally unable to transact (directly or indirectly) with the following countries: Cuba, Iran, North Korea, Sudan, and Syria. In addition, Celonis does not do business in Russia. If the institution you are studying/working at or with is located in/belongs to one of these countries, then you will need to wait until the situation changes to apply for Celonis STaR Program.

Can I schedule a personal meeting with the Research & Innovation Coordinator?

Due to capacity reasons, we do not schedule calls with students. For every question you have regarding your project with Celonis, please first read the FAQs and the information and instructions we have given you in our welcome package (you get this after signing up for our STaR program).

If you have questions that are not answered by the FAQs, please check our Celonis Community (you will get the information how to engage here in your welcome letter after signing up)

If you have still not received the information you are looking for, you can write an email to [email protected] and state as a Subject “STaR Program + (topic of your question)”.

• Please tell us specifically what your question is and where you need support. Our team will then get back to you. Because we get many requests, following up on your request can take up some days.

Can you help me find an academic supervisor?

If you already know your topic but do not have an academic supervisor yet, we can check within our network whether we know a suitable researcher who can support you. We cannot promise you that we will find somebody, but in most cases this is feasible. It depends on your university whether external supervisors are accepted.

Please note that all communication and the checking of all regulations is in your responsibility. We can only support in making the first introduction.

Do I need to sign any documents to work with you on my project? What about Data Declarations and Non-Disclosure?

If you need to sign a Non Disclosure Agreement (NDA) depends on your topic. If you only use our standard material and don’t need anything further you don’t need to sign any documents.

If you need access to internal information at Celonis and your thesis contains confidential information (technical data, customer data), you have to sign a Non-Disclosure-Agreement (NDA).

If your project will contain this internal information, your academic supervisor will have to sign an NDA as well, and the project must not be published.

Please also check the Thesis Guidelines in your Welcome Letter for specific information on Non-Disclosure Agreements and Data Security.

How can Celonis support my project?

Celonis can provide researchers and thesis students free access to the academic version of our software and help you analyze your own data. We can provide our expertise and help you get started by giving you access to our digital training platform. We will also connect you with fellow students and researchers through the Celonis community platform. This support is entirely provided online by signing up to our thesis program.

Please do not contact us directly, but first follow the steps of this program. In some cases, we can also provide further support such as contact to experts or interviews. In these cases, we perform a pre-selection of relevant topics through a pitch, for which you can access a template after registration.

How do I find a suitable topic?

You as a student are responsible for finding a topic you want to research on. As a company we have no resources to guide you in finding a research topic. Please frame your topic in agreement with your academic supervisor. If you are writing your thesis with a company, please also make sure that the research topic is in line with your company supervisor.

How closely are you working with external students on their projects?

External students (those not working at Celonis) need to be mainly supervised by their academic and/or company supervisors. We do not have the capacity to support every project extensively.

I am a working student at a company and I would like to use the Celonis Academic Intelligent Business Cloud for my project, which is at the same time a commercial project at the company I work for. Is this possible?

Yes, but only if you use the academic version of our software just for the scope of the project, and only you personally have access to the Celonis account. Any primarily commercial use of our academic version is prohibited, as stated in the terms of use.

Is it also possible to write a doctoral/PhD thesis with support by Celonis?

Yes! If you want to write your PhD Thesis using Celonis materials, please sign up for our STaR programme. You may book an appointment with our Research & Innovation Coordinator, who will be able to discuss the scope of our involvement. However, we do not give out any PhD scholarships or working positions only with the purpose of writing a PhD thesis.

Is there a possibility to work with Celonis as a working student while I write my thesis?

At the moment, we do not offer working student positions only for the purpose of writing a thesis. However, you can check our career page for open internship or working student positions and then contact us to find out if a combination with your thesis topic is possible.

What does the Celonis STaR Program cover and what not?

What we provide after you have signed up for our Thesis Program:

• We provide students and researchers (after signing up for our STaR program) with free access to our Celonis Software the Execution Management System (EMS).
• We provide free access to our Online Academy and train you to become a process mining expert.
• Within our StaR program package you also get access to some literature and our Celosphere Conference contents.
• If you need further support e.g. interviews, use cases etc. you can apply for this by handing in a pitch for your project. All information on this is coming with our welcome package after signing up for the program.
• We connect you to other students via our Celonis Community.
• We provide you with all information you need to do your project using the Celonis Software with our welcome package and the FAQs that are regularly updated.
• If questions are not covered by the FAQs you can check out our Celonis Community (you will get the information how to access with your welcome letter after you signed up for the Thesis Program).
• If your question is still open after those steps, email us via [email protected] and state as a Subject “STaR Program + (topic of your question).

What we do not provide. Due to capacity reasons we can not…

• support you if you have not signed up for our STaR Program.
• help you find a topic.
• help you frame your project - unless you are working at Celonis (please reach out to your academic supervisor for topic collection and framing).
• provide personal calls (does not apply for people working at Celonis).
• provide you with interviews or data (unless you apply for support via a thesis pitch - only the best students with a high impact for Celonis will be accepted).
• proofread your paper.
• answer any question that is already answered by the FAQs.

Monthly news and updates

Thank you for your subscription, any question, more celonis, check us in.

Final Theses

Topic areas.

The Service-centric Networking group offers bachelor and master theses in different subject areas:

Applications

Topic Areas: Business Process Management, Internet of Things, Supply Chain Management, Machine Learning & AI, Causal Analysis

Topic Areas: Privacy Engineering, Identity Management & Digital Trust, Cloud/Edge Computing, Digital Trust in the Internet of Things

Topic Areas: 6G Networking, IoT Networking, Blockchain and Distributed Ledger Analytics

Please note!

If you are interested in a subject area, please contact our team assistants Andrea Hahn or Sandra Wild ( [email protected] ) and attach your CV, an grades overview and include some basic background information such as your degree program, main areas of study, and most importantly: two or three sentences about what general topic/ direction you would be interested. Our team assistants will put you in touch with matching SNET group members (topic coordinators) which you can then meet in order to discuss open topics in detail.

Supervised Theses

Graduate Seminar

The graduate seminar is a forum for scientific discussions. Students have the possibility to discuss their theses amongst fellow students, graduates and the professor of SNET. In the early phase of their work, their thesis approach is discussed, while at the end the results are presented. Students who are currently working on their thesis at our chair are required to attend each of the meetings, especially if other related topics are being presented. We are also looking forward to welcoming other students who are interested in the seminar or are about to write a thesis at our chair.

Additional Information

Students who decide to write a thesis at our chair are required to talk about their topic and the approach they are going to follow in an initial presentation in the early phase of their thesis time. This presentation should take 10 minutes with 5 minutes of questions and answers afterwards. After finishing their thesis, students have to defend it by giving a talk, in which they demonstrate the results achieved in the thesis. Bachelor students should talk 15 minutes with 5 minutes of questions and answers , whereas Master students are required to present 20 minutes and discuss it afterwards for 10 minutes .

Planned Graduate Seminars

September 4, 2024.

August 7, 2024

July 3, 2024

June 5, 2024

April 24, 2024 (Online Meeting)

April 3, 2024

March 6, 2024

February 7, 2024

January 10, 2024

This section covers general topics and frequently asked questions about the organizational process of bachelor, master and diploma theses at our department. Please read this page carefully before contacting one of the supervisors so that you are well prepared when you express your interest and discuss the topics with the supervisors.

The organizational workflow of a bachelor's, master's or diploma thesis at SNET is described below:

•  Please, contact the secretariat of the department. The secretariat will check possible options within the team.
• You will receive feedback from one of our team members or from the secretariat.
• As soon as the contact with a possible supervisor could be established, you discuss the desired topic with him. 
• You will have 2-4 weeks to consider and familiarize yourself with the topic.
• If you choose this topic, we expect you to write an outline of your thesis in these 2-4 weeks, which we will evaluate. This outline will help you and us clarify the work that needs to be done on your topic.
•  In the next graduate seminar, you will have to give a first presentation of your work: 10 minutes + 5 minutes question and answer. When you get the final OK from the professor, you can proceed to step 5.
• Register your topic with the Examination Office; from now on, attendance at the Graduate Seminar is mandatory (your email address will be added to a mailing list for invitations, please also check our website regularly).
• There may be several graduate seminars and meetings with your advisor while you are working.
• Submission of the dissertation; now attendance at the graduate seminar is not mandatory except for the dissertation defense. Bachelor's students give a 15 minute presentation + 5 minutes question and answer, while Master's students give 20 minutes + 10 minutes question and answer.
• After the defense, a thesis report with the grade will be sent to the examining authority.

Please note that it must be submitted at least 6 weeks before the end of the semester to ensure that you receive your grade in the same semester. Therefore, you must register your thesis with the Examination Office at least 4 months + 6 weeks for Bachelor's theses and 6 months + 6 weeks for Master's theses prior to submission.

Also when preparing a thesis, please observe the following instructions:

• Please check the APO of Faculty IV, especially § 13 (Final Theses)
• Please write your thesis in German (BA)/ English (BA/ MA)
• For Bachelor's theses, you should write approx. 30-50 pages and for Master's theses approx. 50-90 pages
• The abstract must be written twice, in German and English
• Please use the LaTeX template of the chair
• Do not copy text passages without quoting the author (plagiarism)
• SNET Thesis - LaTeX Template
• SNET Slide Layout - Presentation Template (Powerpoint)
• SNET Slide Layout - Presentation Template (Keynote)