earticle

영어

Disperse engineering information constitutes a domain of complex relationship, into which research using complex network could contribute significantly to its organization and preservation. A complex network which takes engineering software data formats as nodes is constructed and its statistical characteristics which include directed edge or undirected edge, the node number, the edge number, average path length, clustering coefficient and node degree distribution, etc, are analyzed in this paper. Voronoi diagram based complex network visualization method and its retrieval method are provided. According to statistical characteristics of constructed network, visualization method calculates network nodes’ position in a two-dimensional plane and divides this plane into Voronoi diagram by feature nodes of network nodes’ clustering and network nodes. Retrieval method restricts compared nodes to nodes in Voronoi domain of feature nodes which are more related to query node, so that the number of compared nodes is reduced in retrieval process. The experiment result indicates applying this method into data formats conversion path retrieval can ensure retrieval precision and improve retrieval efficiency, thus provide reliable basis for migrating numerous disperse engineering information.