Design And Implementation Of Search Engine In Expert Discovery Platform

With the slowdown of regional economic growth,regional innovation and transformation have become an important direction to break through the bottleneck of regional economy development.While Experts who have the latest academic knowledg can guide and advise the government and enterprises,and transport talents,which is an important role in regional innovation.In the actual process of promoting regional innovation,the main problem faced by government and enterprises is how to locate expert information that meets the needs in a large amount of academic data.Due to the lack of reliable and accurate information source support,governments and enterprises cannot accurately find the required experts,which hinders the regional innovation and development.The main manifestation of expert information is academic achievements.By obtaining the academic papers published by experts,analyzing the domain characteristics of experts,and constructing an expert search engine,it helps government enterprises locate experts and promote regional innovation and development.In order to design and implement the expert search engine,this thesis first obtains the expert paper data from the network as the text corpus data,and carries out the research work,including: according to the data characteristics of the expert papers,research the expert retrieval method,introduce the author theme model,and establish the author theme relationship.And combined with the traditional language query model to determine the hybrid query model of expert relevance;use the improved PageRank web page sorting method to sort the importance of experts;comprehensive relevance query model and importance sorting model,as an expert search engine search Model;processing expert paper data,and according to the expert query model,respectively generated the inverted index of the language model and the inverted index of the author topic model,storing the index data,providing the search data for the expert search engine;detailing the workflow of the expert search engine With each sub-module,the search engine can quickly and accurately obtain the search results.The various modules of search engine was completed by python,and combines with query model and index data to complete the overall implementation of search engine.Among them,the expert search results are composed of two parts.The expert relevance score represents the relevance of experts and query words.The expert importance score represents the professional reliability of the experts.Meanwhile,the experiment is designed to verify the query accuracy and query speed of the search engine,and use the relevant query indicators to verify the query effect and performance of the expert search engine.Itcombines text clustering method and traditional retrieval model,and applies it to character search engine.At the same time,it introduces two retrieval indexes of relevance and importance,which avoids the singularity of traditional character retrieval,so that expert search engine provides retrieval results,which can be more realistic...