从产品评论中挖掘观点:原理与算法分析

Mining Opinions from Product Review:Principles&Algorithmic Analysis

本文首先对产品评论中的观点挖掘进行了形式化的描述,给出了基本的研究思路;接着分析了从产品评论中识别产品属性的原理,并对基于词汇共现的方法和基于机器学习的方法进行了比较;然后分析了对相关属性进行极性判断的原理,并对各种现行方法进行了比较;最后对观点挖掘在未来的发展热点进行了展望。

The paper first gives the formal definition of opinion mining in product review, describes the basic research ideas, analyzes the principles of identifying product attributes from product review, and compares the word based method and the machine learning-based method. Then, the paper analyzes the principles of determining the polarity of related attributes, and compares the existing methods. Finally, the paper predicts the development of opinion mining in future hotspot.