Reducing Offline Evaluation Bias in Recommendation Systems
release_vjrof7qe4jaufa5bml4rrfl5jq
by
Arnaud De Myttenaere, Boris
Golden
2014
Abstract
Recommendation systems have been integrated into the majority of large online
systems. They tailor those systems to individual users by filtering and ranking
information according to user profiles. This adaptation process influences the
way users interact with the system and, as a consequence, increases the
difficulty of evaluating a recommendation algorithm with historical data (via
offline evaluation). This paper analyses this evaluation bias and proposes a
simple item weighting solution that reduces its impact. The efficiency of the
proposed solution is evaluated on real world data extracted from Viadeo
professional social network.
In text/plain
format
Archived Files and Locations
application/pdf 360.7 kB
file_uzk5rnunmvf3pkb4ns7iekzgxm
|
arxiv.org (repository) web.archive.org (webarchive) |
1407.0822v1
access all versions, variants, and formats of this works (eg, pre-prints)