Abstract
To solve the problems in traditional search engine like large returning number, bad professionalism and low precision, we analyse the working principles of Nutch open source search engine. FMM algorithm is adopted based on word library, and key-word based vector space model is adopted to make topic relevance judgment. Then the key word is extended base on ontology to improve the sorting results. So Nutch-based agricultural vertical search engine is designed, with implementation of agricultural domain ontology on the information acquisition and filtering, retrieval and similar terms recommending of search engine. The experimental results show that our agricultural search engine can improve the precision of user retrieval and satisfies the professional demand of user.
