A New Knowledge Fusion Method Based on Semantic Rules

In this paper, we present a new algorithm to fuse distributed knowledge objects based on semantic relationships and rules. Knowledge objects are the representation with ontology and meta-knowledge set. A set of comparing rules to restrict the fusion condition is described formally. Filtering knowledge objects by both structural formats and semantic rules can classify them into different granularity levels. They also instruct the fusion process to avoid illogic results. Experimental results of a case study show that the proposed knowledge fusion algorithm is extensible and it can effectively improve the adaptation degree of fused knowledge objects