Information retrieval from electricity distribution enterprise heterogeneous data source based on XML and index

Many electricity distribution enterprises have their own MIS and accumulate much data, but due to the decentralized distribution and heterogeneity, these data cannot be shared by other systems. This paper presents a XML information semantic retrieval method in decentralized and heterogeneous electricity distribution enterprises. To get the semantic related result nodes, it uses interconnection relationship concept and star R-related relation concept. It presents a new index structure that is composed of element tag index, element content index and node layer-path index (NLPI). Our index structure doesn't include an index that stores the interconnection relation between every two nodes for XML document. Based on the index structure, it presents a new semantic search algorithm that calculates the interconnection relation between two nodes just according to these two nodes' encoding and path information and needs not compute other nodes' interconnection relation so as to reduce searching time. The experiment states that we get the results exactly and completely while using shorter time. (6 pages)