A Study On The Normalization Of University Institution Names

Name normalization is to gather all forms of names and attribute information in order,and a method of combing organization structure,which can effectively solve the bottleneck problem of scientific research institutions in the name of the record confusion and hierarchical fuzzy on information retrieval and evaluation.The emergence of new institutions,elimination of traditional institutions,renamed,split,restructuring and merger,made from the same institution has an even more former name,similar to the name,full name,institution referred to as the combined with institutions and non-standard name written form used interchangeably,Results in a reduction of existing organization name recognition,fuzzy organization affiliation and associated mechanism,Causing serious problems for the name given to the connecting point of the information retrieval,statistical analysis,measurement and evaluation activities,which in turn affect the retrieval efficiency,and the credibility of statistical analysis and quantitative evaluation.Based on the scientific research institutions in our country as the research object,the name of the organization were analyzed and summarized from two aspects of speech and word formation statistics,According to the rules,building institutional identity vocabularies,and author’s institutional data based on academic papers database,exploration Agency name recognition,organization name normalization.On the basis of this will be the same entity belongs to a different evolutionary name.Finally,by analyzing seized recall and precision of the results,Provide the basis for institution-building specification name-alias mapping table.In order to solve agency name description chaos,fuzzy relationship and other issues to seek an effective solution.