Synthetic Log Generation Under Control-Flow Conditions Using Autoregressive Models. Kuhn, M., Trinh, T., Grüger, J., & Bergmann, R. In Posenato, R. & Vanderfeesten, I., editors, Advanced Information Systems Engineering Workshops, pages 12–24, Cham, 2026. Springer Nature Switzerland. abstract bibtex Data-driven process analytics, including machine learning and process mining, rely on event logs that capture how business processes are executed. In many organizations, these logs are sensitive, incomplete, or difficult to share. This has increased interest in synthetic event log generation, especially conditional generation that creates synthetic logs under user-defined constraints. Thus, we propose the Petri Net Conditioned Synthesizer (PNC-SYN) to generate synthetic logs that are conformant with a given normative Petri net. At each step, an autoregressive model predicts a probability distribution over the next possible events, and a Petri net playout layer restricts this distribution to the transitions enabled by the Petri net. Furthermore, the autoregressive model remains responsible for generating realistic event sequences and attributes such as timestamps, costs, and resources. We study (1) how to integrate Petri net playout with an autoregressive model to ensure conforming traces while enabling attribute generation and (2) how well Petri net conditioned synthetic logs match the original attribute distributions as well as the utility for process mining tasks compared to an unconstrained autoregressive synthesizer.
@InProceedings{kuhn_autoregressive_2026,
author="Kuhn, Martin
and Trinh, Tony
and Grüger, Joscha
and Bergmann, Ralph",
editor="Posenato, Roberto
and Vanderfeesten, Irene",
title="Synthetic Log Generation Under Control-Flow Conditions Using Autoregressive Models",
booktitle="Advanced Information Systems Engineering Workshops",
year="2026",
publisher="Springer Nature Switzerland",
address="Cham",
pages="12--24",
abstract="Data-driven process analytics, including machine learning and process mining, rely on event logs that capture how business processes are executed. In many organizations, these logs are sensitive, incomplete, or difficult to share. This has increased interest in synthetic event log generation, especially conditional generation that creates synthetic logs under user-defined constraints. Thus, we propose the Petri Net Conditioned Synthesizer (PNC-SYN) to generate synthetic logs that are conformant with a given normative Petri net. At each step, an autoregressive model predicts a probability distribution over the next possible events, and a Petri net playout layer restricts this distribution to the transitions enabled by the Petri net. Furthermore, the autoregressive model remains responsible for generating realistic event sequences and attributes such as timestamps, costs, and resources. We study (1) how to integrate Petri net playout with an autoregressive model to ensure conforming traces while enabling attribute generation and (2) how well Petri net conditioned synthetic logs match the original attribute distributions as well as the utility for process mining tasks compared to an unconstrained autoregressive synthesizer.",
isbn="978-3-032-28160-9"
}
Downloads: 0
{"_id":"fRi4nSaT6J3bmg4b7","bibbaseid":"kuhn-trinh-grger-bergmann-syntheticloggenerationundercontrolflowconditionsusingautoregressivemodels-2026","author_short":["Kuhn, M.","Trinh, T.","Grüger, J.","Bergmann, R."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","author":[{"propositions":[],"lastnames":["Kuhn"],"firstnames":["Martin"],"suffixes":[]},{"propositions":[],"lastnames":["Trinh"],"firstnames":["Tony"],"suffixes":[]},{"propositions":[],"lastnames":["Grüger"],"firstnames":["Joscha"],"suffixes":[]},{"propositions":[],"lastnames":["Bergmann"],"firstnames":["Ralph"],"suffixes":[]}],"editor":[{"propositions":[],"lastnames":["Posenato"],"firstnames":["Roberto"],"suffixes":[]},{"propositions":[],"lastnames":["Vanderfeesten"],"firstnames":["Irene"],"suffixes":[]}],"title":"Synthetic Log Generation Under Control-Flow Conditions Using Autoregressive Models","booktitle":"Advanced Information Systems Engineering Workshops","year":"2026","publisher":"Springer Nature Switzerland","address":"Cham","pages":"12–24","abstract":"Data-driven process analytics, including machine learning and process mining, rely on event logs that capture how business processes are executed. In many organizations, these logs are sensitive, incomplete, or difficult to share. This has increased interest in synthetic event log generation, especially conditional generation that creates synthetic logs under user-defined constraints. Thus, we propose the Petri Net Conditioned Synthesizer (PNC-SYN) to generate synthetic logs that are conformant with a given normative Petri net. At each step, an autoregressive model predicts a probability distribution over the next possible events, and a Petri net playout layer restricts this distribution to the transitions enabled by the Petri net. Furthermore, the autoregressive model remains responsible for generating realistic event sequences and attributes such as timestamps, costs, and resources. We study (1) how to integrate Petri net playout with an autoregressive model to ensure conforming traces while enabling attribute generation and (2) how well Petri net conditioned synthetic logs match the original attribute distributions as well as the utility for process mining tasks compared to an unconstrained autoregressive synthesizer.","isbn":"978-3-032-28160-9","bibtex":"@InProceedings{kuhn_autoregressive_2026,\nauthor=\"Kuhn, Martin\nand Trinh, Tony\nand Grüger, Joscha\nand Bergmann, Ralph\",\neditor=\"Posenato, Roberto\nand Vanderfeesten, Irene\",\ntitle=\"Synthetic Log Generation Under Control-Flow Conditions Using Autoregressive Models\",\nbooktitle=\"Advanced Information Systems Engineering Workshops\",\nyear=\"2026\",\npublisher=\"Springer Nature Switzerland\",\naddress=\"Cham\",\npages=\"12--24\",\nabstract=\"Data-driven process analytics, including machine learning and process mining, rely on event logs that capture how business processes are executed. In many organizations, these logs are sensitive, incomplete, or difficult to share. This has increased interest in synthetic event log generation, especially conditional generation that creates synthetic logs under user-defined constraints. Thus, we propose the Petri Net Conditioned Synthesizer (PNC-SYN) to generate synthetic logs that are conformant with a given normative Petri net. At each step, an autoregressive model predicts a probability distribution over the next possible events, and a Petri net playout layer restricts this distribution to the transitions enabled by the Petri net. Furthermore, the autoregressive model remains responsible for generating realistic event sequences and attributes such as timestamps, costs, and resources. We study (1) how to integrate Petri net playout with an autoregressive model to ensure conforming traces while enabling attribute generation and (2) how well Petri net conditioned synthetic logs match the original attribute distributions as well as the utility for process mining tasks compared to an unconstrained autoregressive synthesizer.\",\nisbn=\"978-3-032-28160-9\"\n}\n\n\n","author_short":["Kuhn, M.","Trinh, T.","Grüger, J.","Bergmann, R."],"editor_short":["Posenato, R.","Vanderfeesten, I."],"key":"kuhn_autoregressive_2026","id":"kuhn_autoregressive_2026","bibbaseid":"kuhn-trinh-grger-bergmann-syntheticloggenerationundercontrolflowconditionsusingautoregressivemodels-2026","role":"author","urls":{},"metadata":{"authorlinks":{}}},"bibtype":"inproceedings","biburl":"https://web.wi2.uni-trier.de/publications/WI2Publikationen.bib","dataSources":["MSp3DzP4ToPojqkFy"],"keywords":[],"search_terms":["synthetic","log","generation","under","control","flow","conditions","using","autoregressive","models","kuhn","trinh","grüger","bergmann"],"title":"Synthetic Log Generation Under Control-Flow Conditions Using Autoregressive Models","year":2026}