EURO-Online login
- New to EURO? Create an account
- I forgot my username and/or my password.
- Help with cookies
(important for IE8 users)
3792. Multi-objective optimal recommender systems
Invited abstract in session MB-3: Optimization in Online Environments, stream Data Science Meets Optimization.
Monday, 10:30-12:00Room: 1005 (building: 202)
Authors (first author is the speaker)
| 1. | Elaheh Lotfian
|
| Statistics, Tarbiat modares university | |
| 2. | Alireza Kabgani
|
| Department of Mathematics, University of Antwerp |
Abstract
In this talk, we introduce a multi-objective recommender system aimed at
enhancing both the accuracy and diversity of top-n recommendation lists.
To address this optimization challenge, we present a customized hybrid
AMOSA_NSGA-II (HAN) algorithm, facilitating the creation of a Pareto set
for top-n lists. Additionally, we provide a methodology for selecting the
optimal list for each user within this Pareto set. Initially, we generate
preliminary top-n lists through item-based collaborative filtering.
Subsequently, the second stage addresses a bi-objective optimization
problem related to recommendation lists, utilizing the customized HAN
algorithm. Finally, the third stage focuses on producing optimal
personalized top-n lists for individual users. To assess the effectiveness of
our method, we implement it on real-world datasets, conduct a thorough
performance evaluation, and compare its results with existing approaches
in the literature.
Keywords
- Multi-Objective Decision Making
- OR and the Internet
Status: accepted
Back to the list of papers