How to deal with negative preferences in recommender systems: a theoretical framework

Area of Science:

• Computer Science
• Artificial Intelligence
• Recommender Systems

Background:

• Negative feedback is crucial for expressing preferences but is underutilized in current recommender systems.
• Existing recommendation approaches primarily focus on positive user feedback, neglecting the value of negative information.

Purpose of the Study:

• To explore and formalize methods for exploiting negative user preferences in generating recommendations.
• To provide a unified framework for analyzing and comparing different approaches to incorporating negative feedback.

Main Methods:

• Utilizing a logical semantics framework for the recommendation process.
• Identifying and categorizing three core conceptual approaches for handling negative user preferences.
• Analyzing variations within these conceptual approaches.

Main Results:

• Demonstrated three distinct conceptual strategies for utilizing negative user preferences.
• Showcased how existing recommendation techniques align with specific alternatives within the proposed framework.
• Established a formal basis for comparing different methods of negative preference incorporation.

Conclusions:

• Negative user preferences offer a valuable, yet often overlooked, data source for enhancing recommendation quality.
• The developed formal framework enables a systematic understanding and comparison of methods for leveraging negative feedback.
• This research opens avenues for more sophisticated and personalized recommender systems by effectively integrating user dislikes.