Original scientific paper
https://doi.org/10.1080/00051144.2023.2284026
An approach to improve the accuracy of rating prediction for recommender systems
Thon-Da Nguyen
; University of Economics and Law, Ho Chi Minh City, Vietnam; Vietnam National University, Ho Chi Minh City, Vietnam
*
* Corresponding author.
Abstract
Sentiment analysis is critical for classifying users on social media and reviewing products throughcomments and reviews. At the same time, rating prediction is a popular and valuable topic inresearch on recommendation systems. This study improves the accuracy of ratings in recom-mendation systems through the combination of rating prediction and sentiment analysis fromcustomer reviews. New ratings have been generated based on original ratings and sentimentanalysis. Experimental results show that in almost all cases, revised ratings using a deep learning-based algorithm called LightGCN on 7 various real-life datasets improve rating prediction. Inparticular, rating prediction metrics (RMSE and MAE, R2, and explained variance) of the proposedapproach (with revised ratings) are better than those of the typical approach (with unrevisedratings). Furthermore, evaluating ranking metrics (also top-k item recommendation metrics) forthis model also shows that our proposed approach (with revised ratings) is much more effectivethan the original approach (with unrevised ratings). Our significant contribution to this researchis to propose a better rating prediction model that uses a supplement factor sentiment score toenhance the accuracy of rating prediction.
Keywords
VADER; sentiment analysis; recommender systems; rating prediction; recommendation systems
Hrčak ID:
322946
URI
Publication date:
21.11.2023.
Visits: 0 *