Nowadays, recommender systems are an essential part of many digital platforms – among then, well-known sites such as Netflix or Amazon are constantly suggesting tailored, individual recommendations of movies or items to their users, based on similarity of movies and users. With the increasing amount of data being incorporated into the computation of these recommendations, it is necessary to reduce complexity and optimize for efficiency. Know-Center has now contributed to a paper dealing with this topic.

A new journal paper with contribution from Bernhard Geiger, a scientist from the Knowledge Discovery Team at Know-Center, who has co-authored the paper with Clemens Bloechl (Rohde & Schwarz, Muenchen) and Rana Ali Amjad (TU Muenchen), has now been accepted for publication in the renowned journal “IEEE Transactions on Knowledge and Data Engineering”, which is one of the top-ranked journals in knowledge engineering, is in Q1 since 2005, and has an impact factor of 3.4.

The authors’ paper is titled “Co-Clustering via Information-Theoretic Markov Aggregation” and  combines information theory and the theory of Markov chains to get a method for co-clustering that is mathematically justified and works well on real-world data sets. Looking at the example of Netflix, this method makes it possible to compute a matrix of similar users and similar movies at the same time, while also reducing the size of this matrix significantly.

For more details on the paper, you can read the abstract and download the full paper here (Open Access).

We congratulate all authors on their success!

Full bibliography:
Bloechl, R. A. Amjad and B. C. Geiger, “Co-Clustering via Information-Theoretic Markov Aggregation,” in IEEE Transactions on Knowledge and Data Engineering. (Volume, issue, page numbers, and year will be added soon.)