用户行为可靠性评价综合框架

Abstract

用户行为可靠性评价对于网络服务的发展具有重要作用，已有研究分别从概率统计、用户意图、用户行为模式以及数据挖掘等角度对其进行广泛研究。本文在定义网络用户行为的可靠性评价基础上，梳理现有算法和模型，针对现有评价模型存在的问题，提出一种包括用户行为数据收集层、用户行为划分层、用户行为模式训练层、不可靠用户行为鉴别层以及用户行为管理与控制层的用户行为可靠性评价综合模型框架，能够在一定程度上解决用户行为可靠性评价中的恶意机器人问题。

The reliability evaluation of user behavior is playing an important role on the development of network services.The current researches about reliability evaluation of user behavior include the probability statistics,user behavior intention,user behavior model and data mining.On the basis of the reliability evaluation of network user behavior,the definition of network user behavior’s trust evaluation and the summary of the existing research,the paper aims at the existing problems in the current evaluation models,and tries to propose an integrated framework of reliability evaluation method of network user behavior.There are five layers in this framework,which are data collection layer,user behavior division layer,user behavior training layer,unreliability behavior identification layer and user behavior management and control layer.This framework makes a positive effect in improving the solution to the problem of bad machine behaviors in the reliability evaluation of user behavior.