Ebrahim Bagheri – Publication Page

Mining Variable Fragments from Process Event Logs

Asef Pourmasoumi and Mohsen Kahani and Ebrahim Bagheri
Reference:
Asef Pourmasoumi; Mohsen Kahani and Ebrahim Bagheri Mining Variable Fragments from Process Event Logs. In Information Systems Frontiers, 2016.
Links to Publication: [www]
Abstract:
Many peer-organizations are now using process-aware information systems for managing their organizational processes. Most of these peer-organizations have shared processes, which include many commonalities and some degrees of variability. Analyzing and mining the commonalities of these processes can have many benefits from the reusability point of view. In this paper, we propose an approach for extracting common process fragments from a collection of event logs. To this end, we first analyze the process fragment literature from a theo≈retical point of view, based on which we present a new process fragment definition, called morphological fragments to support composability and flexibility. Then we propose a novel algorithm for extracting such morphological fragments directly from process event logs. This algorithm is capable of eliciting common fragments from a family of processes that may not have been executed within the same application/organization. We also propose supporting algorithms for detecting and categorizing morphological fragments for the purpose of reusability. Our empirical studies show that our approach is able to support reusability and flexibility in process fragment identification.
Bibtex Entry:
@article{isfi2016, author = {Asef Pourmasoumi and Mohsen Kahani and Ebrahim Bagheri}, title = {Mining Variable Fragments from Process Event Logs}, journal = {Information Systems Frontiers}, year = {2016}, url = {http://www.springer.com/business+%26+management/business+information+systems/journal/10796}, abstract = {Many peer-organizations are now using process-aware information systems for managing their organizational processes. Most of these peer-organizations have shared processes, which include many commonalities and some degrees of variability. Analyzing and mining the commonalities of these processes can have many benefits from the reusability point of view. In this paper, we propose an approach for extracting common process fragments from a collection of event logs. To this end, we first analyze the process fragment literature from a theo≈retical point of view, based on which we present a new process fragment definition, called morphological fragments to support composability and flexibility. Then we propose a novel algorithm for extracting such morphological fragments directly from process event logs. This algorithm is capable of eliciting common fragments from a family of processes that may not have been executed within the same application/organization. We also propose supporting algorithms for detecting and categorizing morphological fragments for the purpose of reusability. Our empirical studies show that our approach is able to support reusability and flexibility in process fragment identification.} }




Powered by WordPress