I. Introduction
Historically, BPM (Business Process Management) is presented for companies as key to grow up effectively and efficiently. Consequently, large structures opt for PAIS (Process Aware Information System) such SAP (Enterprise Resource Planning System) to implement and manage their processes. There is a crucial need to set up methods and techniques that monitor process execution in a PAIS [1]. The available techniques include process mining used to discover process models, analyze process performance, and improve process execution quality. Several researches are using historical data recorded during process execution for process mining exercises that allow analyzing real process behavior [2]–[4]. In parallel, some other researches employ mining techniques to extract configurable process models from event logs, knowing that the concept of configurable processes is more gainful facing the high cost constraint of developing new separate business process for each application context in PAIS [5]–[7]. Configurable process models are marked by representation of common parts and variable parts in one consolidated model. It is also possible to use defined configuration rules in order to derive variant model that fit to the user requirements [8]. This type of processes promotes business processes reusability and reduces time of business process development. However, the quality of configurable processes discovered does not depend only on mining techniques applied but also on the quality of event logs. To overcome the challenge related to event logs data complexity, semantic technologies are introduced to prepare traces before proceeding to process mining application. It is motivated by the need to increase abstraction level of traces [5]. Semantics were also employed for multi-purposes: i) boosting the level of automation in BPM, ii) defeating the dependence between business experts and IT personnel and iii) reduce complexity in business process reengineering [9]. As for configurable process models, semantics are used to i) derive process variants [5], ii) extract configuration rules [18] or iii) semantically validate configurable process model [8]. According to our previous comparative study, we raised the need to use semantics for discovering semantically enriched configurable business processes and we proposed new framework for this purpose [10]. In this paper, we study the semantic enrichment of event logs and develop the first component of the proposed framework, which is related to semantic enrichment of event logs by linking them to the appropriate ontologies. The remainder of this paper is organized as follows. Section 2 introduces semantic in configurable process discovery, and semantic enrichment of configurable process. Section 3 provides an overview of related works on semantic enrichment of configurable process models and semantic in configurable process discovery. Section 4 that details our approach for event logs pre-processing. Finally, Section 5 concludes the present work and gives directions for future one.