Group-by and Aggregate Functions in XML Keyword Search

In this paper, we study how to support group-by and aggregate functions in XML keyword search. It goes beyond the simple keyword query, and raises several challenges including: (1) how to address the keyword ambiguity problem when interpreting a keyword query; (2) how to identify duplicated objects and relationships in order to guarantee the correctness of the results of aggregation functions; and (3) how to compute a keyword query with group-by and aggregate functions. We propose an approach to address the above challenges. As a result, our approach enables users to explore the data as much as possible with simple keyword queries. The experimental results on real datasets demonstrate that our approach can support keyword queries with group-by and aggregate functions which are not addressed by the LCA-based approaches while achieving a similar response time to that of LCA-based approaches.