iRecommend: An Accurate, Scalable, User-Powered Distributed Recommendation Architecture
Computing & Wireless : Computing Methods
Available for non-exclusive licensing
- Yin Zhang, Ph.D. , Computer Science
- Tae Won Cho , Computer Science
- K.K. Ramakrishnan , AT&T Corporation
- Divesh Srivastava , AT&T Corporation
Recommendation systems are widely used in Internet applications. Users rely on recommendation systems to help filter some of the vast content available online, and businesses can use them to improve service to their users and increase profits by leading users to products that are really relevant. In current recommendation systems, users only play a passive role and have limited control over the recommendation generation process. As a result, there is often considerable mismatch between the recommendations made by these systems and the actual user interests, which are fine-grained and constantly evolving.
This invention is a user-powered distributed recommendation architecture, where a user will define a recommendation request paired with a community declaration. Community declarations can flexibly define fine-grained communities of interest in a declarative fashion by individual users. The user can then obtain recommendations accurately tailored to their interests by aggregating opinions of users in such communities.
- Combines a progressive sampling technique with data perturbation methods
- Communities are defined by the user by imposing constraints upon the pool of respondents
Content providers, for content recommendation and prefetching
Proof of concept
- 3 U.S. patents application filed