Dual-feedback credit assessment system and method based on emotion and credit

Abstract

The invention relates to a semantic recognition technology, a machine learning technology and a networked transaction credit assessment technology and discloses a dual-feedback credit assessment system and method based on emotion and credit, wherein the system and method overcome the defect that the existing credit assessment technology is influenced too much by subjective factors. The system comprises a comment semantic analysis system, a user credit system, a merchant credit system and a credit calculation system. The method includes the steps that input of comments of an evaluator is tracked; a dictionary segmentation unit segments feature words; an evaluator credit database receives the feature words; the evaluator credit database performs statistical analysis; the semantic analysis system analyzes emotional features and calculates the emotion weight; the evaluator credit database performs statistics on recent credit analysis; the evaluator credit database performs statistics on recent emotion analysis; the input of scores of the evaluator is tracked; the credit calculation system calculates the scores; an evaluated party credit database records the credit scores. With the system and method, the too much influence of the subjective factors and credit factors on credit assessment can be effectively adjusted, so that the credit evaluation tends to be credible and fair.

Description

technical field [0001] The invention relates to semantic recognition technology, machine learning technology and network transaction credit evaluation technology. It specifically relates to a dual-feedback credit evaluation system based on sentiment and historical credit, and a credit evaluation method. Background technique [0002] The credit system is one of the foundations for the existence of e-commerce. A good credit ecosystem can regulate and restrict the behavior of both parties to the transaction and promote the development of e-commerce. The existing e-commerce credit evaluation system on the Internet is simple and has the following shortcomings: [0003] Method 1: Use the star rating method to perform simple statistical evaluation on the star rating. In this way, the influence of subjective factors is too great, which is not conducive to obtaining fair and just evaluations, and it cannot effectively control the new malicious evaluations in the network, which affe...

AI Technical Summary

Problems solved by technology

In this way, the information display is simple and rude, the amount of information is too large to be easy to browse, and the content of the evaluation is not effectively used to help establish a credit system

Images