Examination and Categorization of Keywords Associated with Online Review Sorting

Abstract

In modern days, product sold online are reviewed online. For example, large online platforms like Amazon allows all users to leave reviews composed of a rating and a comment to any product, whether the purchased it or not. Reviews that are immediately visible influence the purchase decisions of potential customers, therefore the platforms need to sort the reviews so that the first few reflect the product’s quality accurately. However, current algorithms are still not yet mature enough for the task. In order to determine the credibility and reliability of online product reviews, we summarize the factors typically considered in review sorting studies into four categories: product-specific factors, general attitude factors, reviewer-specific factors, and time factors. We then examine the contents and implications of each category. Future designers of algorithms can use the factors to their advantages.