Design And Implementation Of A Hidden Markov Model Based Model For Legal Named Entity Recognition

Using Natural Language Processing(NLP)to analyze legal text,can provide valid legislative evidence for legal worker,thus assisting the legal decision-making and legislation process.So,it has become an important research topic to apply NLP technology on dealing with legal text effectively.Named Entity Recognition(NER)is one of the most important tasks of Natural Language Processing,while named entities in the field of law usually have more word nesting than normal named entities.Therefore,recognition for the named entity of legal text is relatively more difficult.Against this background,we propose a Legal Named Entity Recognition model based on Hidden Markov Model(HMM)in this paper,to do NER and entity relation extraction work for Chinese legal text.Firstly,this paper investigates the development of Chinese NER technology at home and abroad,research status of Chinese word segmentation technology,information extraction(information extraction,IE)technology,and legal text analysis model.After that,this paper introduces hypothesis and construction process of HMM model,and describes HMM's training and solving process.Secondly,this paper puts forward a Legal Named Entity Recognition model based on HMM.In this model,NERs from shallow to deep level are carried out by using a series of HMM models,the N-Gram model is used to segment the sentence and the output of the lower level Hidden Markov Model is used as input of the high-level model,and then the search engine is used to reduce synonymous entities,thus,it completes the whole NER process for the legal text.After that,the proposed model is tested by using the open data set and the legal text data set,compared with three other remarkable NER models.And in the experiment,our proposed model behaves 90% F-1 value for person name entity recognition,both location and organization name F-1 value are also higher than other three models,which shows that our model's performance is better.Thirdly,on the basis of the model,we designed and developed a legal text analysis platform.The platform accomplishes the function of information retrieval,entity relation mapping and so on,which provides a convenient tool for text analysis researchers and legal workers.Lastly,the platform of the legal text analysis platform is tested,and the test results show that the search engine relied Chinese legal text analysis model based on HMM can effectively accomplish the naming entity recognition task.