Hybrid recommendation by incorporating the sentiment of product reviews

• DOI: 10.1016/j.ins.2023.01.051
• Publication Date: 2023-01-10T16:55:24Z
• Authors: Mehdi Elahi; Danial Khosh Kholgh; Mohammad Sina Kia...; Mourad Oussalah; Sorush Saghari
• Keywords: Sentiment Analysis

Hybrid recommender systems utilize advanced algorithms capable of learning heterogeneous sources data and generating personalized recommendations for users. The can range from user preferences (e.g., ratings or reviews) to item content description category). Prior studies in the field have primarily relied on "ratings" as feedback, when building profiles evaluating quality recommendation. While are informative, they may still fail represent a comprehensive picture actual preferences. In contrast, there other types feedback that differently complementarily users their preferences, including reviews sentiments encapsulated within them. Such reveal important parts user's profile not necessarily correlated with ratings, hence, potentially reflect different side profile. this paper, we propose novel form hybrid system, analyzing extracting incorporated into recommendation process. We used generate incorporating additional data, such review sentiment. conducted analyses showed always highly music domain). This might mean sentiment be indicative aspect an alternative signal feedback. Hence, both our proposed system. selected two common datasets evaluation, Amazon Digital Music Video Games, superior performance system compared baselines. comparison were made evaluation scenarios, namely, considered