Social Navigation through the Spoken Web: Improving Audio Access through Collaborative Filtering in Gujarat, India

The Spoken Web project is building a telecom Web that provides a platform for information access and collaborative problem-solving for the developing world. Phone-based audio navigation to relevant content in this new kind of Web has become a major challenge. In this paper we delineate the opportunities for applying collaborative filtering to improve audio navigation on the telecom Web and describe our initial experiences with similarity-based and latent variable methods of collaborative filtering.