一种情感极性感知的信任模型

Sentiment polarity-awareness trust model

针对在线评论中文字评论和等级评价不一致的问题,提出基于情感极性分析的综合评价生成算法,并依据隐马尔可夫模型(Hidden Markov Model,HMM)构建了信任计算模型,将按时间排序的综合评价作为观测序列用于寻找最优的信任状态序列,把实体处于最可信状态的概率作为其可信度.该模型不仅提高了信任计算的准确性,还体现了信任的动态性.仿真实验验证了所提模型的有效性.

Focusing on the discordance between text comments and ratings, a synthetical evaluation generating algorithm is proposed by using sentiment polarity analysis,and a trust computation model is constructed based on the HMM.The optimal hidden trust state sequence is found according to the observation sequence which is a synthetical evaluation sequence.The probability of an entity in the most credible state is the trust value.In this model,the precision of trust computation is improved,and trust dynamics is reflected.Finally,satisfactory results are obtained with simulated experiments.