Research Article
Multiple-Criteria Rating Recommendation with Ordered Weighted Averaging Aggregation Operators
@INPROCEEDINGS{10.1007/978-3-031-28816-6_4, author={Hiep Xuan Huynh and Loi Tan Nguyen and Hai Thanh Nguyen and Linh Thuy Thi Nguyen}, title={Multiple-Criteria Rating Recommendation with Ordered Weighted Averaging Aggregation Operators}, proceedings={Context-Aware Systems and Applications. 11th EAI International Conference, ICCASA 2022, Vinh Long, Vietnam, October 27-28, 2022, Proceedings}, proceedings_a={ICCASA}, year={2023}, month={3}, keywords={Multiple-criteria ratings Recommendation systems Reviews Context}, doi={10.1007/978-3-031-28816-6_4} }- Hiep Xuan Huynh
Loi Tan Nguyen
Hai Thanh Nguyen
Linh Thuy Thi Nguyen
Year: 2023
Multiple-Criteria Rating Recommendation with Ordered Weighted Averaging Aggregation Operators
ICCASA
Springer
DOI: 10.1007/978-3-031-28816-6_4
Abstract
In recent years, the Fourth Industrial Revolution in Industry 4.0 has exploded, along with the increasing development of websites, social networks, and other Internet services, leading to tremendous growth in collected data resources. Therefore, it is becoming more and more challenging to select useful information to make decisions. The recommendation systems are considered a great solution to assist humans in finding helpful information effectively and speedily. Such systems can automatically analyze, classify, select, and provide valuable information to users. Furthermore, they can explore reviews on products and services using artificial intelligence techniques to provide valuable recommendations. Users sometimes give reviews and ratings multiple times on the same products, but they differ depending on the user’s mood, context, behavior, etc. Thus, the problem is accurately determining the user’s rating when exploring such reviews. This work has proposed a solution for multiple-criteria rating analysis. This study has explored reviews on different criteria and integrated them into one aggregate rating by considering the similar relationship between the ratings, users, or products based on criteria in the collaborative filtering-based recommendation approach. The proposed method has performed better than traditional collaborative filtering-based methods on more than 5000 film reviews from the DePaulMovie dataset.
