Sentiment Analysis Of Medical Online Comments Based On Frame Semantics

With the rapid development of modern science,technology and the full popularity of the Internet,online medical and other Internet industries are booming,and a large number of doctor-patient dialogues and patient evaluation records have been accumulated,including a large number of evaluation information and emotional information.The acquisition and processing of this kind of information is helpful to evaluate comprehensively and accurately the patient's condition and the level of medical service,as well as to provide decision-making reference for the treatment and nursing work.However,the current medical knowledge engineering mainly focuses on the extraction and storage of objective information.The lexical semantic emotional resources for emotional analysis of medical comments are scarce,the emotional classification is shallow,the semantic information is incomplete,and the development of medical emotional analysis technology is seriously lagging behind.Therefore,it is necessary to extract and process the semantic information of medical reviews effectively,to realize the semantic understanding of emotional content,and to make the intelligent level of medical emotional knowledge mining higher.Firstly,this paper summarizes the current research status in the field of medical emotions at home and abroad,combs the framework semantic theory,and proposes a framework-based semantic representation model of medical emotions;secondly,through the analysis of medical online reviews,emotional words are extracted,words expressing the same cognitive scene are summarized into a framework,and its framework elements,framework-framework relations and words are defined.Affective tendency value,complete the construction of dictionary,provide the basis for emotional analysis;then propose the semantic annotation rules based on dependency syntax and the calculation method of emotional tendencies of phrase blocks,construct the emotional analysis method of medical onlinecomment based on dictionary and rules;collect medical online comment data for experiment,measure the performance of the emotional analysis method of medical online comment,and the experimental results before and after the revision of natural language processing results are compared and analyzed.The results show that the emotional analysis method of online medical comment presented in this paper has better accuracy,recall rate and F1 value.Finally,the work,innovations and shortcomings of this paper are summarized.The research results of this paper provide semantic resources for emotional semantic role analysis,polarity judgment and other related processing tasks,and provide basic technology for opinion mining and semantic query.