A dynamic multi-algorithm collaborative-filtering system

Nowadays users have access to an immense number of media content. They are able to consume thousands of TV channels and millions of video clips from online portals like YouTube. Due to the immense number of available content, users can have the problem to find content of interest. Recommendation systems are able to filter the immense number of recommendations and they are able to recommend content which fits to the interests of users. However, this research work presents a newly developed recommendation system which is able to increase the accuracy of predictions for recommendations. The newly developed recommendation system uses several algorithms and dynamically selects the most accurate algorithm. The system takes state-of-the-art algorithms and newly developed collaborative-filtering algorithms into account. The research work of this thesis proves that a dynamic selection of the most accurate filtering algorithm by considering more algorithms is able to increase the accuracy of the predictions significantly.

Autorentext

Christian Überall graduated in Media Informatics at the University of Applied Sciences Gießen-Friedberg, Germany, in 2007. In 2012 he optained a PhD in Information Engineering at City University London. His PhD research focused on recommendation systems. His research interests include recommendation systems, multimedia applications and usability.