Achieve Semantic-Based Precise Component Selection Via an Ontology Model Interlinking Application Domain and MVICS.

Chengpu Li,Xiaodong Liu,Jessie Kennedy

2010-01-01

Abstract:With the continuous increases in component varieties and size, it becomes more challenging to find components perfectly applicable for an application. A major obstacle for wider and smoother component reuse is the lack of automated and effective approaches to component specification and retrieval. This paper presents a domain-integrated ontology-based approach to holistic and adaptation-aware component specification and search. The work explores the possibility of combining the MVICS model with domain ontology, with financial domain as a case study. Such a combination enhances the function and application scope of the MVICS model by bringing more domain semantics into component specification and retrieval. The approach and its prototype tool support three distinctive features integrally, including Semantic-based Dynamic Component Retrieval, Flexible Interlink with Domain Model, and Result Component Profile. Furthermore, the inter-linkage between MVICS and domain models has resulted in improved retrieval precision and scope. The effectiveness of the approach and tool has been evidenced by user feedbacks via wide web use.