A System for Analyzing the Frequency of Product Feature Mentions in Review Videos
Fumiya Yamaguchi, Aiko Kobayashi, Mayumi Ueda, Da Li, Shinsuke Nakajima · 2024
In online shopping, there exists a risk that purchased items may differ from expectations. Consequently, more users are turning to product review videos for guidance. However, due to the abundance of reviews on the internet, efficiently accessing desired information can be challenging for users. Therefore, we are developing an analytical system aimed at supporting online shopping. Specifically, we provide users with a quantification of the features mentioned in the videos, facilitating efficient video consumption. This paper reports on the development of a method using BERT to analyze product features from subtitles and comments in review videos, aiming to identify the mentioned aspects.