Semi-automatic Software Feature-Relevant Information Extraction from Natural Language User Manuals: An Approach and Practical Experience at Roche Diagnostics GmbH

摘要

Context and motivation: Mature software systems comprise a vast number of heterogeneous system capabilities which are usually requested by different groups of stakeholders and which evolve over time. Software features describe and bundle low level capabilities logically on an abstract level and thus provide a structured and comprehensive overview of the entire capabilities of a software system. Question/problem: Software features are often not explicitly managed. Quite the contrary, feature-relevant information is often spread across several software engineering artifacts (e.g., user manual, issue tracking systems). It requires huge manual effort to identify and extract feature-relevant information from these artifacts in order to make feature knowledge explicit. Principal ideas/results: Our semi-automatic approach allows to identify and extract atomic software feature-relevant information from natural language user manuals by means of a domain glossary, structural sentence information, and natural language processing techniques with a precision and recall of over 94% and 96% respectively. Contribution: We provide an implementation of the atomic software feature-relevant information extraction approach together with this paper as well as corresponding evaluations based on example sections of a user manual taken from industry.