| In recent years, the database technology has been widely used in a few fields.The keyword search based on the relational database technology does not require users to master any knowledge of SQL grammar and the database schema. Users only need to input keywords,can conveniently search information via keyword like Internet search engines. As results,users are pleased to use the technology.This paper designs a relational database system based on agricultural information. In the preprocessing stage, the key words inputted by user are processed by the method of Chinese word segmentation. Moreover, the key words matching strategy is proposed. The main work is the followings:(1) According to the characteristics of agricultural data, a thesaurus is established. The synonymous are stored in a row in the processing, and the first word of each row is theme, so the other is its synonymous of the word.(2) In this paper, we employee a query expansion plan based on the query sentence, and propose a query expansion method based on thesaurus. The method matches the key word with the thesaurus. If the key word matches a word in the thesaurus, and the words in the same row are used as a synonym for the key word.(3) In the preprocessing of query semantics, the query syntax is analyzed, and the key words are divided into two categories such as the single keyword and the combination of attributes and key words. This improves the rates of precision and recall for keyword queries in the system.(4) In this paper, we use the keyword search method based on the classification of Steiner tree. The problem is transformed to a Steiner tree which is classified according to the number of key words in the tuple.We use classification steiner tree to sort the query results.Experimental results show that the system designed by this paper the rates of recall and precision can meet the application requirements in agricultural information retrieval field.Also, it can effectively avoid the redundant data. |
PDF Full Text Request