Abstract:
In recent years, recommender systems have gained significant attention due to their ability to assist users in finding relevant and personalized recommendations in various domains. Multi-criteria recommender systems, in particular, aim to consider multiple dimensions or criteria in order to generate more accurate and diverse recommendations. One of the key challenges in these systems is capturing and modeling the preferences of users, which are often subjective and imprecise.
This paper proposes a fuzzy-based approach for modelling user preferences in multi-criteria recommender systems. Fuzzy logic provides a flexible and intuitive framework for dealing with uncertainty and imprecision in user preferences. The proposed approach leverages fuzzy sets and linguistic variables to represent user preferences across multiple criteria. By using fuzzy membership functions, the approach quantifies the degree of satisfaction or preference for each criterion, allowing for a more nuanced representation of user preferences.
To incorporate fuzzy-based user preferences into the recommendation process, a fuzzy inference mechanism is employed. The fuzzy inference mechanism combines the fuzzy preferences of users with the item characteristics and generates a recommendation score for each item. The items with higher recommendation scores are then presented to the users as personalized recommendations.
To evaluate the effectiveness of the proposed approach, experiments are conducted on a real-world dataset. The results demonstrate that the fuzzy-based approach outperforms traditional methods in terms of recommendation accuracy and diversity. Furthermore, the approach provides a transparent and interpretable representation of user preferences, enabling users to understand and modify their preferences if desired.
In conclusion, this paper presents a fuzzy-based approach for modelling user preferences in multi-criteria recommender systems. The proposed approach addresses the challenges of subjective and imprecise user preferences by using fuzzy logic techniques. The experimental results validate the effectiveness of the approach in generating accurate and diverse recommendations. This research contributes to the advancement of recommender systems by providing a robust and user-centric approach for modelling preferences in multi-criteria scenarios.
A FUZZY-BASED APPROACH FOR MODELLING PREFERENCES OF USERS IN MULTI-CRITERIA RECOMMENDER SYSTEMS, GET MORE COMPUTER SCIENCE PROJECT TOPICS AND MATERIALS
