Research On The Method Of Information Extraction Based On GATE

Along with the advance of the judicial data in our country,more and more judicial information is published in the form of web pages for public inquiries.As of March 2015,the national courts at all levels of public information has reached more than 7000000 cases.If the use of these public cases in the commercial application of the information on the case of the referee,it is necessary to carry out the case of the core field of the structured processing,which is usually done by manual operation.In the face of such a large number of cases of information,manual processing is obviously very low efficiency.How to use the information technology to simplify the process becomes a problem to be solved at present.In view of this research background,this paper presents a method of judicial case information extraction based on GATE,which is based on the JAPE rule based information extraction technology,the GATE system as the technical basis for the research and development platform.Firstly,this paper studies the status quo of the research topic related to the research,and then summarizes the basic theory,including the introduction of GATE system,the basic grammar of JAPE rules,the implementation of the principle and application of the introduction.Then,based on the platform of GATE system,this paper mainly studies the method of judicial information extraction based on GATE,including the end of the information processing,the JAPE rule extraction method in the case of information field,the result of the organization's output,etc..Finally,the information extraction method is applied to the actual batch of data.The results show that the method can achieve more than 94%,the recall rate can reach more than 86%,the overall performance of F-measure can reach more than 92%,prove that the method is scientific and effective.