Technique to Evaluate a Comprehension-based Interface for Multiple Experts

This paper deals with an interface between expert knowledge and group of experts from whom this knowledge has been acquired and who are going to be users of it. Method is based on deriving the most supported opinion of the group of experts acting as knowledge sources with further applying of it as an interface to each of experts acting as user. Knowledge is acquired using three-valued voting-type knowledge representation (agree, disagree, no-op) of interactions between objects of the domain and the concepts used by experts to describe these objects. During the co-operative knowledge base is being built, the voting-type technique is used to derive peculiarities of each expert to use concepts to describe the domain objects. This paper presents the method of evaluation the similarity between experts which is used later when experts become users of co-operative knowledge. When user is apply to a co-operative knowledge base the evaluation is used to select knowledge acquired from group of the most similar experts to that user. This can be implemented as a system to be used to navigate access of a user to such knowledge which he can interpret in the most accurate way.