Automatic use case flow pattern generation using verb clustering

Software requirements completeness is the one of the key elements for successful software development. Incomplete requirements can cause one to misunderstand requirements and eventually build a wrong system as a developer interprets them in an ad-hoc manner. In order to handle this issue, the use case pattern is suggested, which is a set of commonly discovered scenarios in the use case specification. However, identifying the use case pattern is not systematic and comprehensive without providing any algorithmic or statistical rationale. This paper proposes an automatic approach to identifying the use case flow pattern from various use case specification. In this paper, we gathered 83 use case specifications from eight industrial systems, and presented two use case flow patterns identified from the use case specification.