| Along with the development of Web service and the exponentially increasing of Web data, it is widely realized that integrating relational database with the technique of information retrieval could be more powerful.In this paper, we discuss processing an q-keyword query, k1, k2, k3, ..., kq on a relational database which would be modeled as a weighted graph G(V, E). The V represents a set of nodes (tuples) and E is a set of edges. It represents the foreign key references between tuples. Let Si?V be a set of nodes that contain the keyword ki, we study finding top-k minimum cost result trees that contain at least one node in every Si. When k equals to one, the problem could be reduced to minimum Steiner Tree Problem, which is NP-Complete.We observed that in this area, the number of keywords, q, is small and the number of nodes. Besides, the number of edges usually satisfied with the inequality m< |
PDF Full Text Request