手机知网 App
24小时专家级知识服务
打 开
手机知网|搜索

概念图:从数据图抽取模式(英文)

Su Yatfan;Wong Kamfai

  As the World Wide Web grows dramatically in recent years,there is increasing interest in semi -structured data on the web.It is inefficient to use traditional database technology to store and manipulate them.Thus,graph representations have been proposed.Similar to conventional databases,to facilitate data manipulation on a data graph,graph schemas are used.Different algorithms for schema extraction have been proposed.Noticeably Stanford University has proposed an accurate and concise graph schema.Nevertheless,it is generally expensive to construct and to maintain an accurate graph schema due to the structural irregularity and modification frequency of semi-structured data.As an alternative, approximate graph schemata have been introduced.Relatively,approximate graph schemta have been shown to be more compact.But all existing approximate graph schemata require predetermined parameters,e- g.To overcome this predicament,we introduce the conceptual graph.Conceptual graph construction uses an incremental clustering algorithm to extract schema from a given data graph.The incremental concept avoids pre-determined parameters.Experiments have shown that conceptual graphs maintain the compactness of approximate graph schemata,and are efficient for query processing in applications involving large retrieval.……