The Key Technologies Of Agriculture Search Engine Research

With the development of the digital agriculture,more than a billion Yuan investment has been funded in agricultural information application for resources' collection,analyzing and processing.Accumulated agricultural knowledge and data resources have been close to 100 TB magnitudes.How to get personalized agricultural knowledge and information in massive resources quickly and efficiently has become an urgently problem.The emergence of search engine is effective solution to the information "lost" problem.Existed general search engines lack customer-oriented function for demand in special application domains,which characterized with concentrated information and complexes classification.And search result for specialized subjects and areas are usually inaccurate or not overall.Recently,vertical search engine is appeared which focuses on overall information and timely updating for special domains' search.Development of vertical search engine and agriculture search engine are introduced in the paper.Then principles of search engine and two vertical search technologies,topic distillation and information extraction,are analyzed.Focused on how to filter the unrelated topic pages with URL and content during web pages crawling,and extract information from filtered pages,agricultural information filtering and extraction methods module is presented.Based on the Ultraseek search engine architecture,an agricultural search engine (AgSo-so) is presented.Research areas,such as the artificial intelligence,information extraction and data mining,are integrated in AgSo-so.With adaptation and optimization of relative algorithm,a topic distillation and information extraction module is developed,in which J2EE technology is used for secondary development.