A ranking algorithm of keyword search on probabilistic XML data

Discusses the problem of efficiently ranking the search results of keyword related only to content on probabilistic XML data. A new ranking model is presented according to the characteristic of probabilistic XML data. Unlike the existing ranking algorithms which only depend on the probabilities of retrieval results, the new ranking algorithm proposed fully considered the degrees of nodes discriminating and describing the documents and the characteristic of probabilistic XML data. A ranking model of retrieval results which satisfied the above features is designed and a new inverted index structure for the ranking model is proposed. The new algorithm can accomplish keyword search quickly, so as to provide the most relevant information to the users. The results of simulation experiment show that the proposed method is effective.