Abstract:

Early reviews play a crucial role in shaping consumer decisions and influencing product sales. This study examines the behavior of early reviewers by analyzing their reviews on major e-commerce platforms like Amazon and Yelp. Products are categorized into three lifecycle stages—early, majority, and laggard—where users who post reviews in the early stage are identified as early reviewers. The findings reveal that early reviewers tend to give higher ratings and write more helpful reviews compared to others. Additionally, their ratings and the helpfulness of their reviews significantly impact product popularity. By modeling the review posting process as a competitive game, the study introduces a novel margin-based embedding model to predict early reviewers. Experimental results across multiple datasets demonstrate that this approach outperforms existing methods in accurately identifying early reviewers.

Description:

.