Themen:

• Statistics and numerics with Perl 6
• comparison of matrix and vector data types in different programming languages (Java, C#, C++, Python)
• efficient automated categorization of Wikipedia articles
• mash-ups with machine learning features
• comparison of freely availabl software for classification problems (SVM, LR, ...)
• Recommender systems:
  • implementation of a webservice interface for recommender systems
  • porting a recommender system library from C# to Java
  • implementation of additional engines for MyMediaLite
  • tag recommendation
  • parallel algorithms (multicore, Hadoop)
  • diversification
  • generating music playlists
  • a lunch recommender system for the University of Hildesheim