Abstract: At present, consumers are used to judging the quality of goods by online reviews, however, e-commerce sites are always filled with lots of less useful reviews. A method for assessing the helpfulness of Chinese online reviews based on writing behavior and semantic features is proposed in this paper. It recognizes the writing behavior such as tail-insertion, non-tail-insertion, selected-modification by real-time monitoring the comment text box value change in Web client, and then according to the linear weighted model established on writing behavior and semantic features of reviews, it assesses the helpfulness in server. The experimental results show that the model can accurately and efficiently recognize the useful reviews.

Key words: assessing the helpfulness of reviews, writing behavior, testing the quality of reviews, semantic features of reviews

摘要： 针对电子商务网站充斥着大量有用性较低的评论，提出一种基于用户书写行为与语义特征的中文评论有用性评估方法。方法通过在Web客户端实时监听评论文本框值的变化，识别出句尾插入、非句尾插入、句尾删除、非句尾删除等书写行为，在服务器端根据书写行为、评论的语义特征建立的线性评估模型计算用户评论的有用性。实验结果表明该方法能够较为准确地识别有用性较高的评论。

关键词: 评论有用性评估, 书写行为, 评论质量检测, 评论语义特征