RubE: Rule-based methods for extracting product features from online consumer reviews

详细信息    查看全文

• 作者：Yin Kang ; ^ky6@umbc.eduAuthor Vitae ; Lina Zhou ^{zhoul@umbc.edu}Author Vitae
• 关键词：Product feature extraction ; Rule-based method ; Objective feature ; Indirect dependency relation
• 刊名：Information & Management
• 出版年：2017
• 出版时间：March 2017
• 年：2017
• 卷：54
• 期：2
• 页码：166-176
• 全文大小：1315 K
• 卷排序：54

文摘

The rule-based method is useful for analyzing online consumer reviews. The performance can be improved by recognizing different types of features and designing methods accordingly. For product feature extraction, subjective feature-oriented methods perform better than objective feature-oriented methods. Direct dependency relation and review-specific patterns resulted in best recall in each category. The generality of rule-based methods can be approached by developing domain-independent rules and pruning strategies.