The Research And Implementation Of The Establishment Method Of Expert Databases Based On WEB

The construction and improvement of science and technology expert library is very important for scientific research workers and related enterprises.The establishment of a standard and efficient library of experts and put into use,can not only help enterprises and individuals to complete science and technology research and development,project development and other work,but also to promote and optimize the scientific and technological resources and scientific and technological personnel rational allocation.There are significant limitations in the current method of establishing an expert library.On the one hand,the input of expert information depends on the expert's own application or registration,which led to the lack of expert information and imcomplete imformation and other issues;On the other hand,the updating of most of the expert library rely on artificial maintenance of experts or administrators,which results in information accuracy,timeliness,difficult to maintain in such expert databases,thus affecting the reliability of the entire expert library.In order to solve the above problems,this paper designs and implements a set of WEBbased expert databases establishment method.The method uses the crawler to crawl the expert data from the Internet,and uses the expert information extraction module to transform the multisourced,unstructured data into the format-unified data.Finally,the expert data entity disambiguation can be used,to further reduce the duplication of information and error imformation in the databases,thereby improving the reliability of the entire expert library and query efficiency.The use of the method will facilitate the research works and related business expert data query and analysis,which can bring greater commercial value.The main work of the paper is as follows:(1)Proposed a design of expert library system based on hierarchical classification.Traditional design of databases uses a unified storage and classification method of data storage and processing,which can not meet the application requirements for complex and various data sources.The proposed hierarchical classification based on the expert database can effectively solve the data sources and data processing with high scalability.(2)Studied and designed a method to extract expert data from texts.This paper designs a method to extract the information of the characters from the text for the expert information extraction.This method is used to extract the natural language,unstructured text data into structured data with strict format.This method solves the shortcomings of the traditional word segmentation algorithm which can not extract the information of one person 's activities,and can more accurately identify and obtain the expert basic information,the activity information and the field information of experts from the text.(3)Studied and designed a method of entity disambiguation of expert data.Expert data entity disambiguation is used to remove the ambiguity in existing data in the expert library,to remove duplicate information and contradictory information,and to improve data reliability and query efficiency.The expert disambiguation method based on expert profiles can effectively identify two experts with the same name,identify information that describes the same expert,or two experts with the same name but different experts,and finally combines the information of the same expert.