The Approach Of Data Provenance In Software Traceability Visualization

Requirement changes is an unavoidable topic in software development,which cause subsequent changes in all aspects of development phases.Software traceability is to describe the process and impact of requirement changes,and the relationships between software requirements and designs.In other words,is to record the generation,destruction or conversion of software artifacts during the lifecycle of software development,which is data provenance.In order to display the information intuitively and to understand the evolution of software artifacts easily,a better visualization approach is now required.Nowadays,the agent and time dimension is not intuitive enough in software traceability visualization,which increase the cost of team management and schedule management.The reason for this problem is that the information of these two dimensions is not abstracted when the information of software traceability is modeled.Otherwise,it is a challenge for researchers to efficiently visualize traceability links for big software systems because of visual clutter issues.PROV is a highly abstract conceptual model of data provenance with good scalability,which can show tracking process in a multi-dimensional way.Expanding multi-granularity relationships between software artifacts can adjust the granularity of information according to user demand.Therefore,this paper introduces PROV model to software traceability,and shows the evolution of software artifact in an intuitive and multi-dimensional way to support management activities.First,this paper summarizes the related work of software traceability and data provenance,analyzes the existing problems of software traceability visualization approaches,and expounds the feasibility of using data provenance to solve it.Second,in this paper,basing on the PROV model,a provenance model in software traceability can be established,which also extends the corresponding provenance vocabulary.What's more,according to the specific application environment,the framework of the provenance information management is designed,including its collection,storage,validation,query and visualization.The source information is collected from the version control system,and is disposed to RDF triples.Then,ProvValidator tool is used to verify the information.After that,query the information with SPARQL and upload the results to the ProvStore server to visualize the information graphically.Last,this paper takes the scene of Hubei Provincial Intellectual Property Services Platform as an example to verify the effectiveness of the method.The graphical visualization of the scene depicts the evolution of software artifacts from its requirements to its design,which provide better support of management activities.