Logo des Repositoriums

Auflistung nach Schlagwort "process mining"

1 - 4 von 4
  • Konferenzbeitrag
    Generating Data from Highly Flexible and Individual Process Settings through a Game-based Experimentation Service
    (Datenbanksysteme für Business, Technologie und Web (BTW 2017), 2017) Kaes, Georg; Rinderle-Ma, Stefanie
    The ability to adapt process instances to changing requirements has long been recognized as a fundamental research topic. While in some settings, process flexibility is only required in exceptional situations, in other settings it is the key component which drives the process design. Examples can be found in multiple domains, including the nursing domain, where each patient requires his own, individual therapy process which may change on a regular basis. In this paper, such flexible and individual process settings (FIPS) are analyzed and the basic building blocks are defined based on expert interviews and relevant literature. The building blocks are then mapped onto a game-based experimentation service which o ers a simulation and evaluation environment for FIPS. The data generated in this game are evaluated by a comparison with data from a real world FIPS.
  • Textdokument
    PhD Proposal
    (EMISA 2024, 2024) Andreswari, Rachmadita
    Process mining techniques provide operational insights into work processes in various types of organizations. These processes handle sensitive data of customers, patients, students, or citizens and their results impact the lives and careers of these affected persons. So far, much of process mining research has focused on classical dimensions of performance such a cycle time or operational cost. What is missing is a primal consideration of ethical concerns such as fairness. Fairness is a recently researched concept in machine learning, which requires a deeper integration into process mining algorithms. This research addresses this requirement. To this end, it aims to analyze fairness concerns in process mining, to develop new process mining algorithms that integrate fairness concerns, and to evaluate them for their effectiveness. Methodologically, our research will build on guidelines for design science and algorithm engineering research. In this way, we will combine engineering research with empirical evaluations.
  • Zeitschriftenartikel
    Regelbasierte Steuerung von Geschäftsprozessen – Konzeption eines Ansatzes auf Basis von Process Mining
    (Wirtschaftsinformatik: Vol. 50, No. 4, 2008) Grob, Heinz Lothar; Bensberg, Frank; Coners, André
    Der Einsatz von Geschäftsregeln zur Prozesssteuerung hat mit der Zielsetzung zu erfolgen, die Prozesseffektivität und -effizienz auf Instanzebene zu sichern. Die Erzeugung und Anpassung von Regeln an dynamische Umweltbedingungen stellt einen Aufgabenbereich des Prozessmanagements dar, der durch analytische Techniken zu unterstützen ist. Zu diesem Zweck wird im Rahmen dieses Beitrags das Konzept des Process Minings zugrunde gelegt, das eine methodische Unterstützung dieser Aufgabenfelder mithilfe von Verfahren des maschinellen Lernens leisten kann. Zur Generierung steuerungsrelevanter Regeln werden Entscheidungsbaumverfahren aufgegriffen und ein Anwendungsbeispiel aus dem Umfeld des Forderungsmanagements in der Finanzindustrie präsentiert. Hierauf aufbauend erfolgt die Konzeption eines Integrationsansatzes zur Verzahnung der operativen und analytischen Prozesse, für die auf der Implementierungsebene unterschiedliche Realisierungsmöglichkeiten zur Verfügung stehen. Abschließend wird ein mehrstufiges Selektionsmodell eingeführt, das die Auswahl von Prozessen zur regelbasierten Prozesssteuerung auf Grundlage eines Prozessportfolios unterstützt.AbstractBusiness rules are used to automate processes under the constraint of retaining process effectivity and efficiency. A salient question is how to create and adjust business rules to situational conditions of business processes in order to meet economic targets. Therefore, this contribution proposes the process mining concept to generate explicit process knowledge from available process data. In particular, decision tree induction methods permit the generation of descriptive rule sets which are able to predict process quality. These rule sets can be used as operational knowledge base to ensure effectivity and efficiency of process executions. The application of this rule based control technique for business processes is substantiated within an integrated model which combines analytical and operational processes. Finally, we propose a multi-level decision model to support the selection of adequate processes for rule-based process control.
  • Zeitschriftenartikel
    A self-portrayal of GI Junior Fellow Matthias Weidlich: Event-driven analysis of service processes
    (it - Information Technology: Vol. 60, No. 1, 2018) Weidlich, Matthias
    In domains such as e-commerce, logistics, or healthcare, the conduct of service processes is widely supported by information systems and event data is generated continuously during process execution. Such event data constitutes a valuable source of information to monitor and improve the respective service processes. My research focuses on models and methods to support event-driven analysis of service processes. Specifically, I study how event logs produced by information systems are used to automatically construct models for qualitative and quantitative analysis. Aiming at online assessment and predictive analysis of a process' behaviour, I develop monitoring techniques that utilise streams of event data produced by diverse sources. Architectures that enable efficient handling of event streams are another focal point of my research. In this article, I outline some of the related research questions and highlight my recent results in these areas.