Semi-automated Design Guidance Enhancer (SADGE): A Framework for Architectural Guidance Development

Architectural decision making is a non-trivial task for architects in the software development projects. Researchers have developed several concepts, methods and tools to assist practitioners in their decision making and decision capturing activities. One of these approaches is a decision identification technique that creates architectural guidance models from decisions made in previous projects and from knowledge about a domain found in the literature. To apply this technique, significant manual knowledge engineering effort has to be invested initially. In this paper, we introduce a framework that automatically extracts architectural knowledge entities from architectural related documents by applying natural language processing. A knowledge engineer then manually post processes and fine-tunes the extracted knowledge entities. We applied evaluation techniques from the information retrieval community to measure the sensitivity and accuracy of the framework. Our results show that the automatic approach has the highest recall and shortest processing time while the semi-automatic approach has the highest precision.