h1

% IMPORTANT: The following is UTF-8 encoded.  This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.

@PHDTHESIS{Pourbafrani:973806,
      author       = {Pourbafrani, Mahsa},
      othercontributors = {van der Aalst, Wil M. P. and Depaire, Benoit},
      title        = {{F}orward-{L}ooking {P}rocess {M}ining : {D}ata-{D}riven
                      {S}imulation},
      school       = {RWTH Aachen University},
      type         = {Dissertation},
      address      = {Aachen},
      publisher    = {RWTH Aachen University},
      reportid     = {RWTH-2023-11014},
      pages        = {1 Online-Ressource : Illustrationen},
      year         = {2023},
      note         = {Veröffentlicht auf dem Publikationsserver der RWTH Aachen
                      University; Dissertation, RWTH Aachen University, 2023},
      abstract     = {Business processes, whether related to production,
                      services, or the Internet ofThings (IoT), are ubiquitous in
                      today’s businesses. Information systems are integral to
                      these processes and serve as valuable sources of
                      information. By leveraging process data stored in
                      information systems, data-driven techniques can support
                      business processes. Process mining techniques, which focus
                      on event data, arehighly effective in providing insights
                      into the current state of processes. The majority of process
                      mining techniques are backward-looking and create
                      descriptive process models based on historical data to
                      identify performance and compliance issues. However,
                      forward-looking process mining aims to convert insights from
                      backward-looking approaches into predictions and actions.
                      This involves using simulation and prediction approaches to
                      support organizations in future analysis and
                      decision-making. Forward-looking approaches are the ultimate
                      objective ofprocess mining. Although existing approaches
                      rely on detailed event data, different perspectives on event
                      data can capture process behavior and underlying
                      relationships between process variables for future analysis,
                      such as the effect of daily arrival rates on resource
                      efficiency. In this thesis, we aim to create a
                      forward-looking framework for business processes to replay
                      their processes and analyze the impact of actions taken
                      using process mining insights. The focus of this thesis is
                      the simulation of processes in process mining as prescribing
                      models. Process simulation models are effectivetools for
                      analyzing processes in the future because they are open
                      about the simulation’scomponents and parameters. To do so,
                      we employ both detailed event data and aggregated event data
                      of processes across time (fine-grained event logsand
                      coarse-grained process logs). We investigate the current
                      simulation techniques and design a reference meta model to
                      make the most of the potential inside event data, which is
                      accompanied by multiple stand alone tools. Simulation models
                      should be data-driven in order to be reliable and close to
                      reality. Asa result of assessing the current detailed
                      process simulation and designed meta model, we propose a
                      novel approach to design process simulation at different
                      levels of granularity. We propose different techniques,
                      including changing the perspective and levelof event data
                      for processes. The different perspectives enable discovering
                      the hidden aspects of processes at detailed levels. Changing
                      the data perspective and the systematic transformation is
                      one of the main contributions of the thesis. It allows for
                      different levels of process simulation, such as designing
                      System Dynamics models and coarse-grained diagnostics, e.g.,
                      pattern and concept drift detection, as well as hybrid
                      simulation models. To determine the validity and
                      applicability of the proposed approaches, we evaluated them
                      using synthetic and real-world event data. Furthermore, the
                      presented approaches in this thesis are implemented as tools
                      and supported in practice.},
      cin          = {122510 / 120000},
      ddc          = {004},
      cid          = {$I:(DE-82)122510_20140620$ / $I:(DE-82)120000_20140620$},
      typ          = {PUB:(DE-HGF)11},
      doi          = {10.18154/RWTH-2023-11014},
      url          = {https://publications.rwth-aachen.de/record/973806},
}