Fuzzy model fragment retrieval

Given a set of collected evidence and a knowledge base, fuzzy compositional modelling (FCM) begins by retrieving model fragments which are the most likely to be relevant to the available data. Since FCM often involves imprecise and uncertain information, a match between the available data and the knowledge base cannot in general be done precisely, partial matching may suffice. This paper proposes a more flexible fuzzy model fragment retrieval mechanism to match data items with broader, including possibly subjective information in the knowledge base. It is capable of retrieving those model fragments that can approximately match the collected evidence, when no exact match occurs. The retrieval process and its capability is illustrated by means of an application example.