Creating a Data Generator and Implementing Algorithms in Process Analysis


Bakır Ç., Yüzkat M., Karabiber F.

ELEKTRONIKA IR ELEKTROTECHNIKA, cilt.28, sa.5, ss.68-79, 2022 (SCI-Expanded)

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 28 Sayı: 5
  • Basım Tarihi: 2022
  • Doi Numarası: 10.5755/j02.eie.31126
  • Dergi Adı: ELEKTRONIKA IR ELEKTROTECHNIKA
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Aerospace Database, Applied Science & Technology Source, Central & Eastern European Academic Source (CEEAS), Communication Abstracts, Computer & Applied Sciences, INSPEC, Metadex, Directory of Open Access Journals, Civil Engineering Abstracts
  • Sayfa Sayıları: ss.68-79
  • Yıldız Teknik Üniversitesi Adresli: Evet

Özet

Process mining is a new field of work that aims to meet the need of the business world to improve efficiency and productivity. This field focuses on analysing, discovering, managing, and improving business processes. Process mining uses event logs as a resource and works on this resource. Hence, the system is developed by analysing the event logs, including each step in the process model. Our study is made up of two significant stages: a data generator for processes and algorithms applied for discovering the created processes. In the first stage, the aim was to develop a simulator with the ability to generate data that could help process modelling and development. Within the framework of this study, a system was created that could work with various process models and extract meaningful information from these models. More productive and efficient processes can be developed as a result of his system. The simulator consists of three modules. The first module is the part where users create a process model. In this module, the user can create his own business process model in the system’s interface or select from other registered models. In the second module, team-based data are simulated through these process models. These generated data are used in the third module, called “analysis”, and meaningful information is extracted. In conclusion, the process can be improved considering the information about time, resource, and cost in the generated data. At the second stage, processes were discovered using alpha, heuristic, and genetic algorithms, which are process mining discovery algorithms and synthetic and real event logs. The discovered processes were demonstrated with Petri nets, and the algorithms’ performances were compared using the fitness function, accuracy rates, and running times. In our study, the heuristic algorithm is more successful because it improves the noise in the data and incomplete processes, which are the disadvantages of the alpha algorithm. However, the genetic algorithm yielded more successful results than the alpha and heuristic algorithms due to its genetic operators.