This document proposes a collaborative learning to rank algorithm called SwarmRankCF that uses collaborative latent factors learned from user-item interactions as features for ranking items in recommender systems. It applies a particle swarm optimization algorithm to directly maximize mean average precision during training. The approach is evaluated on a dataset from an internet radio service with over 1,000 users and 35,000 unique artists, using a leave-one-out evaluation methodology to test its ability to rank hidden items highly for users.
1. Swarming to Rank for Recommender Systems
Ernesto Diaz-Aviles, Mihai Georgescu, and Wolfgang Nejdl
Overview
• Address the item recommendation task in the
context of recommender systems
• An approach to learning ranking functions
exploiting collaborative latent factors as features
• Instead of manually creating an item feature
vector, factorize a matrix of user-item interactions
•Use these collaborative latent factors as input to
the Swarm Intelligence(SI) ranking method
SwarmRank
SI for Recommender Systems
Swarm-RankCF Evaluation
• a collaborative learning to rank algorithm based on SI
• while learning to rank algorithms use hand-picked feature to Dataset: Real world data from internet radio:
represent items we learn such features based on user-item 5-core of the Last.fm Dataset – 1K Users
interactions, and apply a PSO-based optimization algorithm transactions 242,103
that directly maximizes Mean Average Precision.
Unique users 888
Items(artists) 35,315
Evaluation Methodology: All-but-one
protocol or leave-one-out holdout method
where hit(u) = 1, if the hidden item I is present in u’s
Top-N list of recommendations, and 0 otherwise.
Contact: Ernesto Diaz-Aviles, Mihai Georgescu
email: {diaz, georgescu}@L3S.de
L3S Research Center / Leibniz Universität Hannover
Appelstrasse 4, 30167 Hannover, Germany
phone: +49 511 762-19715
www.cubrikproject.eu