Developing a methodology for BRM-S

Diplomarbeit, Juli 2011

BRM-S is a framework with a graphical tool, the BRM-S Designer, for modeling business domains in a semantic as well as a syntactic way. At its current research level, a user is left alone in the decisions that lead to a BRM-S model, making this a complex and time-consuming procedure.The work at hand presents a methodology supporting the user with a semi-automatic process. First, concept candidates are extracted from domain specific documents and are being put into the Semantic and Syntactic Scope in BRM-S by expert’s acknowledgements as concepts. These concepts are then being brought into context by definitions and a full taxonomical hierarchy by linking it to an upper hierarchy, WordNet. Non-taxonomical relation candidates and their appearance in texts are extracted from domain specific documents and build the base for a heuristic which tries to give clear and manageable relation proposals to domain experts. In addition to proposing elements and relations for the Semantic and Syntactic Scope, automatic preparations are made to make modeling the Construction Scope as easy as possible for experts. Every element’s origin is stored allowing to keep the BRM-S model up-to-date following element’s creation, change and deletion. This supports traceability and by that supports a model’s lifecycle.