A technique for automatically organizing software libraries for software reuse

As software reuse becomes more prominent and accepted in industry, systems and tools for software reuse become a key aspect in achieving successful reuse of software artifacts. A major problem with such tools is the retrieval and classification of the software modules. To search for and retrieve the conceptually closest software component from a library of software modules, components need to be classified in some manner. We address this problem by showing how the software library can be organized using a clustering scheme. We also show graphically using a simplified example, how our scheme clusters software components containing descriptions of software modules. We contend that this organization easily supports a retrieval method for software components to be reused. Our scheme is automatic and classifies components that have been represented using a knowledge representation-based language.