DISCOVR

Like its predecessor SPREADR, DISCOVR is a showcase of a shopping portal that aims at providing the visitor with better user- and context-aware recommendations for products and events than is possible with traditional recommendation algorithms such as collaborative filtering. In contrast to SPREADR, however, DISCOVR does not focus on spreading activation-based recommendation generation. While a first prototype of the shop was constructed from building blocks of an integrated architecture called “Context-aware Recommendations on Rails”, the current version now leverages the full power of the Hybreed Framework to create hybrid recommenders based upon a variety of different recommendation techniques. As in its predecessor, DISCOVR incorporates a number of internal and external context factors such as the user’s click history, the current location or the current weather to generate adequate recommendations.

Technology

  • Hybreed Framework
  • JSF
  • Java

Scope

CONTici

Context-adaptive interaction in cooperative knowledge processes

Contact

Jürgen Ziegler

Full Professor

Contributors

Tim Hussein

Former team member

Werner Gaulke

Former team member

Timm Linder

Former team member

Publications

Improving collaboration by using context views

Context-aware Recommendations

Automated Generation of a Faceted Navigation Interface Using Semantic Models

Context-aware Recommendations on Rails