Research On Schema-Based Keyword Search In Databases

With the rapid development of Internet and extensive use of relational databases, there is growing recognition that if the information retrieval can combine technology and database technology, will bring a better user experience.In this paper we specify the implementations of schema-based keyword search in relational database, wo also do further study and propose improved algorithm.First, this paper presents the implementation of keyword search in relational database, including the definition of the results and searching algorithm, and on this basis, we put forward two changes: First, the candidate network join algorithm, this paper raise a candidate network evaluation algorithm based on semi-join, experiments show that the algorithm can generate a significant reduction in the number of tuples.Second, we propose reduction rules that can be used in the process of generating result tree.With these rules unnecessary tuples will be removed as much as possible and execution efficiency improved.we also propose improved algorithm based on these rules.Finally, we did the experiments on Oracle database, the different experimental parameters in the case of the new algorithm is compared with the previous algorithm, and results show that: the proposed algorithm did improve query efficiency.