Abstract:
This paper presents a method for restaurant reviews mining based on semantic polarity analysis. It regards the taste, environment, service and price as the restaurant features. The user reviews are tagged with features sentence by sentence, and the complicated sentences with multi-features are divided into several simple feature sentences. It analyses the sentences’ semantic polarity and its intensity. As a result, the user can find out others’ opinions on certain restaurant about its certain feature conveniently. It provides useful guidance for the user.
Key words:
feature tag,
affective ontology,
semantic polarity
摘要: 提出一种基于语义极性分析的餐馆评论挖掘方法。将餐馆的食物口味、环境、服务、价格作为其特征,以句子为单位对用户评论进行特征标注。将具有多个特征的复杂特征句划分为简单特征句,分析评论句的语义极性和极性强度。使用户可方便地了解其他用户对某个餐馆某种特征的评价,为用户消费提供了有力指导。
关键词:
特征标识,
情感本体,
语义极性
CLC Number:
PAN Yu; LIN Hong-fei. Restaurant Reviews Mining Based on Semantic Polarity Analysis[J]. Computer Engineering, 2008, 34(17): 208-210.
潘 宇;林鸿飞. 基于语义极性分析的餐馆评论挖掘[J]. 计算机工程, 2008, 34(17): 208-210.