Research On Refusal Of Automobile Insurance Claim Based On Statistical Syntax Analysis Model

Against the backdrop of the COVID-19,the reform of commercial vehicle fees,the global economic slowdown and local turbulence,the growth of property insurance companies slowed down,and the profitability of property insurance companies at the underwriting end continued to decline,especially for small and medium-sized property insurance companies,who often failed to form economies of scale for super large insurance institutions such as PICC,Ping An Insurance,Pacific Insurance,etc.Therefore,refined cost control has become the way out for such companies.How to scientifically control costs and scientifically settle claims has become the key.On the one hand,insurance companies must perform the contract for reasonable claims and legal claims,and on the other hand,unreasonable claims should be regulated.When an unreasonable claim arises,the insurance company’s refusal to pay is called refusal of compensation.This thesis demonstrates the legal language as a rule system can be calculated.On this basis,this thesis collects the laws and regulations that can support the refusal of claims,as well as the terms used by companies in the industry when refusing claims.Then these texts are cleaned,segmented,and a corpus is built.Then,based on the corpus,the language is trained,and the language model is built using tools.This thesis validates the established language model,and finally obtains the analysis results.The main conclusions of this thesis can be summarized as follows:First,the segmentation results of the reasons for refusal of compensation are usually two-character words;Secondly,on the basis of manually marking the claim refusal French,the probabilistic context-free model based on statistical syntax analysis,combined with the jieba word breaker and Standford parser,can be used to effectively analyze the syntactic structure of legal language.