Please use this identifier to cite or link to this item:
Title: Collaborative filtering with recency-based negative feedback
Authors: João Marques Silva
Alípio Jorge
João Gama
Issue Date: 2015
Abstract: Many online communities and services continuously generate data that can be used by recommender systems. When explicit ratings are not available, rating prediction algorithms are not directly applicable. Instead, data consists of positive-only user-item interactions, and the task is therefore not to predict ratings, but rather to predict good items to recommend - item prediction. One particular challenge of positive-only data is how to interpret absent user-item interactions. These can either be seen as negative or as unknown preferences. In this paper, we propose a recency-based scheme to perform negative preference imputation in an incremental matrix factorization algorithm designed for streaming data. Our results show that this approach substantially improves the accuracy of the baseline method, outperforming both classic and state-of-the-art algorithms.
metadata.dc.type: conferenceObject
Appears in Collections:LIAAD - Articles in International Conferences

Files in This Item:
File Description SizeFormat 
P-00G-PQP.pdf933.71 kBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.