Skip to content. | Skip to navigation

Informatik 5
Information Systems
Prof. Dr. M. Jarke
Sections
Personal tools
You are here: Home Theses A Matrix Factorisation based Algorithm for Decentralised Personalisation

Contact

Prof. Dr. M. Jarke
RWTH Aachen
Informatik 5
Ahornstr. 55
D-52056 Aachen
Tel +49/241/8021501
Fax +49/241/8022321

How to find us

Annual Reports

Disclaimer

Webmaster

 

 

A Matrix Factorisation based Algorithm for Decentralised Personalisation

Thesis type
  • Bachelor
Student Clemens Frank
Status Finished
Submitted in 2017
File download
Supervisor(s)
Advisor(s)

Nowadays Internet services are personalised by so-called recommender systems, which filter relevant content for a user. When user profiles and content are decentralised in order to enhance privacy, the recommender system has to work distributed as well. In this thesis, we evaluate a distributed recommender algorithm for decentralised Online Social Networks, where preferences of users are only shared between two parties at a time and without a central storage. Our results show that the quality of the recommendations is only 5% worse than recommendations by centralised recommender algorithms, thus the trade-off between accuracy and privacy through distribution is acceptable.

Document Actions