Public Sentiment toward Personalized Recommendation System: Evidence in Chinese Social Media
(A2026-131869)

AUTHORS

Lewis Bingwen Huang, University of Macau; Joseph Sy-Changco, University of Macau

ABSTRACT

Personalized Recommendation System (PRS), a prevalent technology that can be easily found in cyberspace, is a hot topic for research in recent years. It is an algorithm technique to provide tailor-made content that fits users’ preferences according to their digital footprints. This study aims to explore and examine the Chinese public attitude toward the personalized recommendation system by mining user-generated contents (UGCs) on Weibo. Through sentiment analysis and topic modeling, the study discovers a neutral to positive sentiment of the general public, and also demonstrates a down trending satisfaction toward PRS over the past few years. Furthermore, increased fanbase, strengthened social engagement, and better user experience are found to be the main advantages brought by PRS, while the echo filter effect, bubble filter effect, annoying Ads, and privacy issues are the main concerns caused by the technology. For managerial implication, it is suggested that moderately diversifying the content pushed and increasing the transparency of the algorithm can be some feasible directions for social platform operators and developers so as to reduce the negative sentiments of the users toward algorithms and to increase the marketing power. As for the online retailer, it should be aware of the gradually decreasing sentiment toward PRS, which might lead to a lower marketing effectiveness, and thus reviewing their marketing strategies and reevaluating the resource allocation on PRS are suggested.