Research Of Vertical Search Engine For Apple Pests And Diseases Based On Ontology

With the rapid development of the Internet,the expanded network information provides wealthy pests information to the farmers.In the field of effective use of information,new challenges need to be solved.In view of the low precision of the existing search engines,keyword-based information retrieval technologies cannot meet the user's semantic needs.This paper studies the vertical search engine of apple pests on the basis of the domain ontology.The main contents and conclusions of this paper are as follows:(1)The ontology construction of apple pests.In order to solve the problem of low efficiency of ontology construction and too strong dependence on domain experts,this paper constructed the ontology of apple pests based on the literature domain ontology construction method.First of all,we used the field corpus to segment Chinese words and clean the data.The domain concept was obtained by the relevance judgment.Then,the correlation between the domain concepts was obtained by the correlation analysis and clustering.Finally,the relational database was used to store the domain ontology.This method reduced the dependence on domain experts and improved the automation of domain ontology construction.(2)Construction of the apple 's pests vertical search engine.Aiming at the problem of low accuracy of information retrieval of apple pests,the domain ontology was introduced into the construction of vertical search engine.Firstly,the relevant information needed to be collected and filtered.After selecting the high quality URL as the initial seeds and crawling the relevant pages with the reptile,we analyzed the pages,extracted the links in the page,and used the regular expression to filter the link.Besides,the filtered URL was placed in the URL library.The relevance of the subject was extracted from the web page text and highly relevant pages were stored in the computer.Furthermore,the index was established by the inverted indexing strategy.Besides we achieved the information of the extended query and related words of the visual recommendation.Finally,the optimization of the theme of reptiles,information extraction performance and apple pest of the vertical search engine were tested.Experiments showed that the optimized crawler speed was increased by 12.7 times.The precision of the information extraction based on the wrapper was kept above 94.6% and theF-index performance was above 87.5%.In this paper,the search engine returned the appropriate number of results,and the correlation was better,higher than the average search engine 22%,higher than the "agricultural search network" 5.2%.(3)Design and implementation of prototype system.In order to facilitate the use and management of the information retrieval platforms,the prototype system of apple pests based on the ontology were designed and realized.The construction of domain ontology,dynamic crawling of web pages,dynamic indexing,extended query,visual recommendations of the related words and other functions were achieved.