Applying Dewey Encoding to Construct XML Index for Path and Keyword Query

Path and Keyword search in XML documents has recently become an active area of research, in which constructing an effective index is a key issue. Existing index structures and encoding schemes for XML query processing target the Parent-Child and Ancestor-Descendant relationships, however, they depend on large index space and complex algorithms used to process regular path expressions. In this paper, we first introduce an encoding scheme by which the relationship of two nodes on a DOM tree and the topology of n nodes of a tree can be calculated efficiently. Then, we propose a novel index approach which is separated into two parts. One, called path index, keeps the path information of XML documents. The other, called value index, stores the content. Experiments show that the approach is more effective than the traditional ones, namely DataGuide and XISS, in terms of running time and size of index space.