Font Size: a A A

Research On Intelligent Question Answering System For Enterprise Law

Posted on:2018-01-06Degree:MasterType:Thesis
Country:ChinaCandidate:X S ZhengFull Text:PDF
GTID:2416330596954793Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
With the rapid development of market economy,the legal system of enterprises is maturing.It's an urgent need for all enterprises to capture the legal knowledge closely related to their own interests quickly and effectively,since the diversity and complexity of the law.The traditional search engine which exist many shortcomings in the response information,like redundancy,error,unrelated,has been completely unable to meet the demands of legal counselors.Therefore the research on intelligent question answering system for enterprise law has become the focus for many research institutions.This thesis focuses on the design and implementation of the intelligent question answering system in the field of enterprise law,discussing a series of key technologies such as domain knowledge base construction,question comprehension analysis,information retrieval and answer extraction.The main works of the thesis are as following:(1)The thesis builds the set and the classification of enterprise law questions by analyzing the characteristics of law questions.In order to understand the intention of users' questions,the thesis extracts the word-bag,part of speech and word sense features of questions which are combined to a new feature and encoded to the form using by classifier.SVM is used to classify the enterprise law questions.The experimental results show that the classification has good accuracy.(2)The thesis builds the base of frequently asked questions(FAQ)library and paragraph keyword relation(PKR)library to solve the information retrieval problem of question answering system.The algorithm of information retrieval based on the cooperation of FAQ and PKR and the sentence similarity calculation is put forward to realize the information retrieval in enterprise law domain.(3)Aiming at the answer extraction of the intelligent question answering system,this thesis puts forward an answer extraction algorithm based on Multi Strategy fusion,which is used to extract the answers to the intelligent question answering system in the enterprise legal field,and realizes the answer extraction in the field of enterprise law.(4)The thesis designs an enterprise law intelligent question answering prototype system based on the above critical technology of question answering system.The modules of the prototype system are described in detail.
Keywords/Search Tags:Intelligent Question Answering, Enterprise Legal Field, Question Classification, Answer Extraction
PDF Full Text Request
Related items