Results Clustering for Keyword Search over Relational Database
Shuxin Yang, Lanying Shi · Journal of Software · 2013
Keyword Search over Relational Database (KSORD) has been a hot research topic in the field of the database. The existing prototype systems present the results to user in a linear list. The user has to browse individually. Therefore, it is still very difficult to find the information users really need. To solve this problem, this study is carried out on results clustering for Keyword Search over Relational Database. Learning from the concept of vector in physics, this study proposes a new model of result tree, which is called result-tree characteristic vector. This study also proposes a new clustering strategy based on result-tree characteristic vector. It firstly gets the result-tree characteristic information, and describes the joint tuple tree using vector representation, and then classifies the retrieval results according to the corresponding vector representation. The experimental results verify the feasibility and effectiveness of the clustering strategy in this study and manifest that the method in this study can efficiently help users navigate through and improve the users’ browsing efficiency.