Keyword-based Deep Web Query

Deep Web contain vast amounts of information, existing search engines are difficult to search the contents due to its hidden. So far, people has not come up with good methods and models to capture content from it, it restricts people get more and more valuable information, therefore, how to fully make the access to the information in Deep Web becomes a problem.This paper proposes a keyword-based Deep Web database query method, the method makes a consideration of Deep Web its own characteristics, and draws on the ideas of tradition search engines. The paper compares various classification algorithms, eventually using the Naive Bayes algorithm to classify the keywords, finding domain it belongs to, the algorithm is suitable for the model of Deep Web query, simple and clear, the accuracy is high. Keywords and attributes associated with the introduction of the concept of ontology, using WordNet dictionary to generate concept hierarchy tree, puts forward the semantic similarity calculation method based on ontology. The method gets rid of people’s subjective awareness, it is simple and intuitive, the cost is not very big. Decide the domain of the query keywords and its corresponding relational table attribute. Eventually generate the SQL statement for the query, the query page information feedback to the user.The part of experiment also takes examples to validate and test the feasibility and accuracy of our proposed method. Our method not only solves Deep Web in various fields of the database query, but also can be integrated with the existing search engine to help users to query quickly and efficiently.