Loading...
Please wait, while we are loading the content...
Trust and Helpfulness in Amazon Reviews : Final Report
| Content Provider | Semantic Scholar |
|---|---|
| Author | Shinzaki, Dylan Stuckman, Kate Yates, Robert |
| Copyright Year | 2013 |
| Abstract | On Amazon, many purchase reviews are dishonest spam entries written to skew product ratings [1]. Though users have the opportunity to rate reviews as helpful or unhelpful, sociological factors and prior ratings influence users to rate these reviews for reasons other than the truth of their content [2, 3]. Many studies have evaluated the content of these user reviews to detect spam entries by mining and classifying the text entry. However, [4] proposes a graph based algorithm to determine the honesty of reviews and trustworthiness of the reviewer for general product review data. In this project, we propose to apply this algorithm to Amazon review data and compare helpfulness data to the resulting quantitative assessments of honesty of reviews and trustworthiness of reviewers. Finally, we analyze the trust of the reviewers and their correlation with helpfulness in order to classify spam. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://snap.stanford.edu/class/cs224w-2013/projects2013/cs224w-060-final.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
