next generation internet.

In the modern Internet there is an avalanche of information. The purpose of this research area is to develop methods of filtering information in such a way that the computer only offers the user relevant information, instead of the user needed to search through irrelevant data before finding what he needs.

The area involves recommendation engines, search engines, e-commerce and social network analysis.

Barend Lutz Measuring public sentiment on democracy from Twitter
Jan Buys Probabilistic Tree Automata for Language Models and Grammar Correction
Marcel Dunaiski Mining Ontological Knowledge from Wikipedia
Simon de la Rouviere Effectiveness of online social-network information stream filtering methods.
Niel de la Rouviere Learning Chinese Vocabulary: a SRS implementation utilizing the unique properties of Chinese Orthography
Dirko Coetsee Normalisation of noisy web text by using Machine Learning techniques
Peter Hayward Topic-based unsupervised semantic parsing
Dirk Badenhorst Predicting trends using internet activity
Francois de Villiers Filtering Information Streams
Stephan Gouws Deep Unsupervised Feature Learning for Natural Language Processing