Temporal recommendation on graphs via long- and short-term preference fusion

TLDR

Recommender systems struggle to capture user preferences over time because simple temporal correlations are often meaningless and user behavior is driven by both long‑term and short‑term preferences. This work aims to represent and exploit users’ long‑term and short‑term preferences for improved temporal recommendation. The authors introduce a Session‑based Temporal Graph that jointly models long‑ and short‑term preferences and propose an Injected Preference Fusion algorithm that extends personalized random walks for recommendation. Experiments on citation and social‑bookmarking datasets show that the IPF method improves recommendation accuracy by 15–34 % over the prior state‑of‑the‑art.

Abstract

Accurately capturing user preferences over time is a great practical challenge in recommender systems. Simple correlation over time is typically not meaningful, since users change their preferences due to different external events. User behavior can often be determined by individual's long-term and short-term preferences. How to represent users' long-term and short-term preferences? How to leverage them for temporal recommendation? To address these challenges, we propose Session-based Temporal Graph (STG) which simultaneously models users' long-term and short-term preferences over time. Based on the STG model framework, we propose a novel recommendation algorithm Injected Preference Fusion (IPF) and extend the personalized Random Walk for temporal recommendation. Finally, we evaluate the effectiveness of our method using two real datasets on citations and social bookmarking, in which our proposed method IPF gives 15%-34% improvement over the previous state-of-the-art.