DESIRE Recommender System

This project investigates the relationship between online behavior and emotional reaction items featured on the webpage. After observing a positive relationship between viewing time of, and preference for, items, the DESIRE Recommender System was created to exploit this relationship. DESIRE is unlike more common “Collaborative Filtering” recommender systems in that it does not compare a user’s preferences or behaviors to those of other users. With DESIRE, your recommendations are based exclusively on you.

Key Publications

  1. Parsons, J. & Ralph, P. (2007). Generating effective recommendations by exploiting viewing time and item attributes. In Proceedings of The Design Science Research in Information Systems and Technology Conference, DESRIST ’07, Pasadena, CA, USA, May.
  2. Ralph, P. & Parsons, J. (2006). A Framework for automatic online personalization. In Proceedings of The Hawaii International Conference on Systems Sciences, HICSS-39, Poipu, Kauai, Hawaii, January.
  3. Parsons, J., Ralph, P., & Gallagher K. (2004). Using viewing time to infer user preference in recommender systems. In Proceedings of The AAAI Workshop in Semantic Web Personalization, San Jose, California, July.

Back to Research