CinemaScreen Recommender Agent: Combining Collaborative and Content-Based Filtering

  • Authors:
  • James Salter;Nick Antonopoulos

  • Affiliations:
  • University of Surrey;University of Surrey

  • Venue:
  • IEEE Intelligent Systems
  • Year:
  • 2006

Quantified Score

Hi-index 0.01

Visualization

Abstract

The CinemaScreen Film Recommender Agent system combines collaborative and content-based filtering techniques to enable recommendations for newly released films showing at users' local cinemas. The agent uses collaborative filtering to discover users with similar tastes and produce an initial set of predicted ratings. It then feeds this set through content-based filtering to expand and fine-tune it according to interfilm relationships. Testing against other schemes showed that the system maintains roughly equivalent precision and recall for normal recommendation tasks and achieves higher precision than competing techniques when specifically recommending new movies.This article is part of a special issue on AI, Agents, and the Web.