Research On Collaborative Filtering Algorithm And Its Application In Tax Risk Monitoring System

In response to the tax risk,organize relevant personnel in 2015 in GuiZhou province local tax bureau "risk monitoring and control system" is designed and developed,the system through analysis and deal with the tax risk,effectively reduced the loss of national tax.However,the following problems occurred during the operation of the system:1.The number of professional staff of the tax bureau for risk response is insufficient,and tax personnel cannot easily learn risk response knowledge in the system.2.The system relies on leadership decision-making in many links,and the system fails to provide more scientific and objective reference data for relevant decisions.3.Tax personnel have no reference cases similar to the tax risks they are currently dealing with in the process of risk response.In this paper,the main work done:(1)Through the use of collaborative filtering recommendation algorithm,in the risk monitoring and control system development and design the personalized recommendation module,risk cases so as to solve problems 1 and 2.(2)By three attributes matching method to solve the problem.In algorithm in the process of practical application,the text will be combined with the characteristics "tax risk monitoring and control system",through the introduction of a social relation network,the system to use existing in the structure of data,the change on the process of collaborative filtering algorithm,the same personnel merge in large group in a similar user group,and reduce the existing algorithm when looking for a similar user base data sparse,solved because the system of evaluation of tax officials have insufficient data and unable to find a similar user base.This paper is to design and implement the process of problem solving,and design the corresponding flow chart is given and the relevant page rendering,and through the analysis of system operation data,the use effect of the proposed algorithm.