Process Mining with Semantics: Real-time Application on a Learning Process Domain

Abstract

Automated means for extraction, analysing or harmonization of various kinds of data that are stored in today’s information systems - is indispensable to perform an effective process mining (PM). In view of that, this paper introduces a semantic-based process mining approach that is capable of detecting useful patterns or trends within any given data or process base. The work illustrates the method using a case study of the learning process domain. Essentially, the paper takes into account the context of the individual learners activities within a learning knowledge-base in order to find the best possible ways to efficiently realize (meta-analysis) the individual properties or attributes the process instances share amongst themselves within the knowledge-base. The goal is to identify patterns that have an effect on users performance and then respond by making decisions based on individual properties (assertions) and the classification process. Thus, the method of this paper is grounded on the semantic modelling and process mining techniques. Practically, the method uses the semantics of the captured events logs about the learning process and discovered models to create new knowledge that is applied for enhancement of the existing information knowledge-base. Theoretically, the work focus on augmenting the information values of the resulting process models based on the individual attributes (object and data properties) that are well-defined within an ontology. On one hand, in order to ensure validity, the work looks at the extent to which the individual process elements and harmonization is met. Whereas, reliability refers to the level of consistency in providing a well-suited inference mechanism or knowledge-base management system that is useful towards drawing valuable and/or accurate conclusions as a result of the improved method of process analysis.