有用性
可读性
惊喜
主观性
计算机科学
心理学
认识论
社会心理学
哲学
程序设计语言
作者
Alton Y.K. Chua,Xiaoyu Chen
标识
DOI:10.1007/978-3-030-96957-8_10
摘要
Some fake online reviews may have overlapping textual features with authentic reviews. This paper explores the nuanced differences between helpful and unhelpful reviews, as well as fake and authentic reviews. Four textual features—polarity, subjectivity, readability, and depth—are used for the investigation. Results suggest that subjectivity, readability and depth help to distinguish between helpful and unhelpful reviews but not between fake and authentic ones. However, polarity offers a clue to differentiate between helpful fake and helpful authentic reviews. Specifically, for positive entries, helpful fake reviews contain more contents indicative of surprise while helpful authentic had more expectation-confirmed words like satisfaction. For negative entries, helpful fake reviews contained more contents indicative of anger while authentic helpful ones carried more anxiety.
科研通智能强力驱动
Strongly Powered by AbleSci AI