Results Diversification for XML Keyword Search Based on the Semantic Category of Central Entity

Results diversification for keyword search on XML documents has attracted considerable attentions from research community in recent years. Though search results were diversified from different perspectives in the existing methods, the effects were still far away from satisfactory. This paper proposes a new way to diversify search results according to the semantic information of central entities which the search results belong to. Entities can be clustered by measuring semantic similarity among each other and search results from entities of the same cluster would be also put together to construct several groups, each of which has a particular label. So, users can locate the results of interest easily and precisely with navigation. Extensive experimental studies show that our methods are effective.