Identifier

etd-0929103-120652

Degree

Doctor of Philosophy (PhD)

Department

Computer Science

Document Type

Dissertation

Abstract

Software undergoes change at all stages of the software development process. Changing requirements represent risks to the success and completion of a project. It is critical for project management to determine the impact of requirement changes in order to control the change process. We present a requirements traceability based impact analysis methodology to predictively evaluate requirement changes for software development projects. Trace-based Impact Analysis Methodology (TIAM) is a methodology utilizing the trace information, along with attributes of the work products and traces, to define a requirement change impact metric for determining the severity of a requirement change. We define the Work product Requirements trace Model (WoRM) to represent the information required for the methodology, where WoRM consists of the models Work product Information Model (WIM) for the software product and Requirement change Information Model (RIM) for requirement changes. TIAM produces a set of classes of requirement changes ordered from low to high impact. Requirement changes are placed into classes according their similarity. The similarity between requirement changes is based on a fuzzy compatibility relation between their respective requirement change impact metrics. TIAM also identifies potentially impacted work products by generating a set of potentially impacted work products for each requirement change. The experimental results show a favorable comparison between classes of requirement changes based on actual impact and the classes based on predicted impact.

Date

2003

Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Doris L. Carver

Share

COinS