Open AccessA task-level parallelism approach for process discovery
Author(s) -
Muktikanta Sahu,
Rupjit Chakraborty,
Gopal Krishyak
Publication year - 2018
Publication title -
international journal of engineering and technology
Language(s) - English
Resource type - Journals
ISSN - 2227-524X
DOI - 10.14419/ijet.v7i4.14748
Subject(s) - speedup , computer science , parallelism (grammar) , parallel computing , data parallelism , task (project management) , process (computing) , task parallelism , event (particle physics) , algorithm , programming language , physics , management , quantum mechanics , economics
Building process models from the available data in the event logs is the primary objective of Process discovery. Alpha algorithm is one of the popular algorithms accessible for ascertaining a process model from the event logs in process mining. The steps involved in the Alpha algorithm are computationally rigorous and this problem further manifolds with the exponentially increasing event log data. In this work, we have exploited task parallelism in the Alpha algorithm for process discovery by using MPI programming model. The proposed work is based on distributed memory parallelism available in MPI programming for performance improvement. Independent and computationally intensive steps in the Alpha algorithm are identified and task parallelism is exploited. The execution time of serial as well as parallel implementation of Alpha algorithm are measured and used for calculating the extent of speedup achieved. The maximum and minimum speedups obtained are 3.97x and 3.88x respectively with an average speedup of 3.94x.