Entropy-Based Similarity for Efficient Keyword Search In XML Databases
With the rapid emergence of XML as a data exchange and data transfer medium over the Web, an effective keyword-based search approach to locate the relevant information is indeed essential. The major challenge with keyword-based search is to ensure that; it returns only the relevant results to the query, and how the query results could be ranked properly. In this paper, the authors propose a two-level indexing scheme based on mutual information. To demonstrate how their proposed indexing works, they use the University Course listing dataset as an example.