Doctoral Seminar: Ellis Emmanuel Eghan

Speaker: Ellis Emmanuel Eghan

Supervisor: Dr. J. Rilling

Supervisory Committee: Drs. D. Goswami, F. Khendek, N. Tsantalis

Title: Using Project Build Semantics to Facilitate Traceability and Reuse

Date: Thursday, November 16, 2017

Time: 10:15 a.m.

Place: EV 3.309

ABSTRACT

Build systems are used to automate the transformation of source code into executable programs and deliverables. Build systems are not only a crucial aspect of software development processes, but they are also a source of knowledge, which can be mined to help the comprehension and evolution of software systems. A key challenge current research in mining software repositories is facing is that knowledge about software artifacts is stored in specialized repositories, which often remain information silos – disconnected from each other. In this research, we introduce a modeling approach, which eliminates these information silos by linking software build knowledge with other software artifacts to improve traceability and support various software development tasks, including security vulnerability analysis, source code change impact analysis, and code reuse. More specifically, we introduce an ontology-based model of software build knowledge to support knowledge integration in the software domain. The proposed approach takes advantage of the Semantic Web, in the form of ontologies and reasoning, to formalize knowledge related to build processes and to infer new knowledge from these existing knowledge resources. This report provides an overview of the problem we tackle, the proposed solution, status of the work and the schedule for the remaining tasks.