Business Process Semantic Annotation Suggestions in ERP Systems

Abstract

Enterprise resource planning (ERP) systems are usually extremely large and combine various views of the business and the technical implementation perspective. Also, a very specific vocabulary has evolved into ERP systems. This vocabulary is not clearly mapped to business management concepts. The semantic web technology can be used in the context of an ERP to enable the absence of the automation process. The proposed solution exploits the use of the semantic web that significantly contributes to overcome the deficiencies rooting from the heterogeneity of information contents and semantics generated from various sources. In addition, ERP databases should be extended with semantic capabilities, specifically explicit descriptions of relationships and conceptualization of the entities detected in the relational model. In this paper, a core reference ontology (CRO) is built to provide metadata mapping between the various business process modeling notations. Also, domain ontology is constructed to be used in business process alignment. Ontologies include shared concepts about business process and data objects. Semantic similarity functions are applied to business process models and domain ontology concepts to increase the degree of automation of business process semantic annotation. Receiver Operating Characteristic (ROC) Curve analysis is performed between two collaborative business process models having a different level of abstraction. The implementation of the proposed solution results in significant enhancement of the Area under Curve (AUC) from .4 to .8.