Enabling system artefact exchange and selection through a linked data layer

Research Projects
Organizational Units
Journal Issue
The use of different techniques and tools is a common practice to cover all stages in the systems development lifecycle, generating a very good number of system artefacts. Moreover, these artefacts are commonly encoded in different formats and can only be accessed, in most cases, through proprietary and non-standard protocols. This scenario can be considered a real nightmare for software or systems reuse. Possible solutions imply the creation of a real collaborative development environment where tools can exchange and share data, information and knowledge. In this context, the OSLC (Open Services for Lifecycle Collaboration) initiative pursues the creation of public specifications (data shapes) to exchange any artefact generated during the development lifecycle, by applying the principles of the Linked Data initiative. In this paper, the authors present a solution to provide a real multi-format system artefact reuse by means of an OSLC-based specification to share and exchange any artefact under the principles of the Linked Data initiative. Finally, two experiments are conducted to demonstrate the advantages of enabling an input/output interface based on an OSLC implementation on top of an existing commercial tool (the Knowledge Manager). Thus, it is possible to enhance the representation and retrieval capabilities of system artefacts by considering the whole underlying knowledge graph generated by the different system artefacts and their relationships. After performing 45 different queries over logical and physical models stored in Papyrus, IBM Rhapsody and Simulink, results of precision and recall are promising showing average values between 70-80%.
linked data, oslc, data shapes, interoperability, retrieval systems, software, reuse, support
Bibliographic citation
Alvarez-Rodríguez JM, Mendieta R, Vara JL, Fraga A, Llorens J (2018) Enabling System Artefact Exchange and Selection through a Linked Data Layer. JUCS - Journal of Universal Computer Science 24(11): 1536-1560.