Classification and Retrieval of Reusable Components Using Semantic Features

A. bs t rac t A u t o m a t e d assistance for software component reuse involves supporting retrieval, adaptation and verification of software components. T h e in formal i ty of feature-based software classification schemes is an impediment t o f o r m a l l y verifying the reusability of a software component. T h e use of f o r m a l specifications t o model and retrieve reusable components alleviates the in formal i ty , but the f o r m a l reasoning required f o r retrieval introduces questions of scalability. To provide scalability, current retrieval s y s t e m s resort t o syntactic classification at s o m e level of abstraction, abandoning the s e m a n t i c i n f o r m a t i o n provided by the specification. In t h i s paper, we propose a methodology tha t s h i p s the overhead of f o r m a l reasoning f r o m t h e retrieval t o t h e classification phase of reuse. Software components are classified using semant ic fea tures tha t are derived f r o m the ir f o r m a l specification. Retrieval of func t ional ly s imi lar components can t h e n be accomplished based o n the stored feature sets. Formal verification can be applied t o precisely de termine t h e reusability of the set of s imi lar components.