A Comparative Study on the Integration of Eye-Tracking in Recommender Systems

Area of Science:

• Human-Computer Interaction
• Information Science
• Artificial Intelligence

Background:

• Recommender systems traditionally rely on explicit user feedback or static profiles.
• Capturing implicit user preferences in real-time is crucial for enhancing personalization.
• Eye tracking offers a non-intrusive method to gather detailed user interaction data.

Purpose of the Study:

• To investigate the integration of eye tracking technologies into recommender systems.
• To evaluate the impact of eye tracking on recommendation personalization, accuracy, and user engagement.
• To propose a framework for real-time recommendation generation using gaze-based data.

Main Methods:

• A comprehensive literature review synthesizing current studies on eye tracking in recommender systems.
• Development of a structured framework comprising an Eye Tracking Module, Preferences Module, and Recommender Module.
• Utilizing eye tracking metrics like fixation duration and gaze patterns for preference inference.

Main Results:

• Gaze-based feedback significantly improves recommendation accuracy.
• Eye tracking provides a valuable source of implicit, real-time user preference data.
• Current limitations include controlled environments, sample diversity, and equipment costs.

Conclusions:

• Eye tracking holds substantial promise for advancing recommender systems beyond static profiles.
• A proposed framework enables adaptive recommendations through continuous refinement of user preferences.
• Future work should focus on accessible tools and integrating diverse behavioral indicators for wider applicability.