Semantic Web Recommender System

A user interface for the semantic recommendation system “Universal Recommender” is to be programmed.

The Universal Recommender is a Java library developed at the DAI-Labor, which can be used to program recommendation systems for semantic data sets. A graphical user interface is to be programmed with which the following tasks can be performed:

  • Visualization of data sets (e.g. degree distributions, 2D embeddings, etc.)
  • Selection of algorithms and parameters
  • Automatic evaluation of recommendation systems
  • Persistence of recommendation configurations that can be embedded in a running recommendation system.

On the servers of the DAI-Laboratory there are several data sets that can be used. The Universal Recommender is available as a Java module. The graphical interface can be programmed as a desktop UI or as a web application.

Required knowledge

  • Java
  • A UI library or web programming (e.g. Swing, SWT, Richfaces, etc.)
  • Basic knowledge in the area of data mining/visualization of data