A Universal Significant Reference Model Set for Process Mining Evaluation Framework

Process mining has caught the attention of researchers and practioners. Because a wide variety of process mining techniques have been proposed, it is difficult to choose a suitable process mining algorithm for a given enterprise or application domain. Model rediscoverability of process mining algorithms has been proposed as a benchmark to address this issue. Given a process model (we call it original model) and its corresponding event log, the model rediscoverability is to measure how similar between the original model and the process model mined by the process mining algorithm. As evaluating available process mining algorithms against a large set of business process models is computationally expensive, some recent works have been done to accelerate the evaluation by only evaluating a portion of process models (the so-called reference models) and recommending the others via a regression model. The effect of the recommendation is highly dependent on the quality of the reference models. Nevertheless, choosing the significant reference models from a given model set is also time-consuming and ineffective. This paper generalizes a universal significant reference model set. Furthermore, this paper also proposes a selection of process model features to increase the accuracy of recommending process mining algorithm. Experiments using artificial and real-life datasets show that our proposed reference model set and selected features are practical and outperform the traditional ones.

[1]  Jianmin Wang,et al.  On Recommendation of Process Mining Algorithms , 2012, 2012 IEEE 19th International Conference on Web Services.

[2]  Jianmin Wang,et al.  A Behavioral Similarity Measure between Labeled Petri Nets Based on Principal Transition Sequences - (Short Paper) , 2010, OTM Conferences.

[3]  Tharam S. Dillon,et al.  On the Move to Meaningful Internet Systems, OTM 2010 , 2010, Lecture Notes in Computer Science.

[4]  Tao Jin,et al.  Efficient and Accurate Retrieval of Business Process Models through Indexing - (Short Paper) , 2010, OTM Conferences.

[5]  Tao Jin,et al.  Querying business process model repositories , 2014, World Wide Web.

[6]  Jianmin Wang,et al.  An empirical evaluation of process mining algorithms based on structural and behavioral similarities , 2012, SAC '12.

[7]  Jianmin Wang,et al.  Efficient Selection of Process Mining Algorithms , 2013, IEEE Transactions on Services Computing.

[8]  Wil M. P. van der Aalst,et al.  Workflow mining: discovering process models from event logs , 2004, IEEE Transactions on Knowledge and Data Engineering.

[9]  Wil M. P. van der Aalst,et al.  Genetic process mining: an experimental evaluation , 2007, Data Mining and Knowledge Discovery.

[10]  Wil M. P. van der Aalst,et al.  Process mining: a research agenda , 2004, Comput. Ind..

[11]  Remco M. Dijkman,et al.  Graph Matching Algorithms for Business Process Model Similarity Search , 2009, BPM.

[12]  Cw Christian Günther,et al.  Towards an evaluation framework for process mining algorithms , 2007 .

[13]  Jan Mendling,et al.  Metrics for Process Models: Empirical Foundations of Verification, Error Prediction, and Guidelines for Correctness , 2008, Lecture Notes in Business Information Processing.

[14]  Jianmin Wang,et al.  A workflow net similarity measure based on transition adjacency relations , 2010, Comput. Ind..

[15]  Jianmin Wang,et al.  Mining process models with non-free-choice constructs , 2007, Data Mining and Knowledge Discovery.

[16]  Jianmin Wang,et al.  Mining process models with prime invisible tasks , 2010, Data Knowl. Eng..