Long-tail Session-based Recommendation

TL;DR

This paper introduces TailNet, a new network architecture designed to enhance long-tail session-based recommendations by balancing popularity and niche preferences, thereby improving diversity without sacrificing accuracy.

Contribution

The paper presents a novel architecture, TailNet, that explicitly models long-tail item preferences in session-based recommendation systems, addressing a gap in existing methods.

Findings

TailNet outperforms state-of-the-art methods in real-world datasets.

It improves recommendation diversity by effectively modeling long-tail items.

The approach maintains competitive accuracy while enhancing long-tail recommendation quality.

Abstract

Session-based recommendation focuses on the prediction of user actions based on anonymous sessions and is a necessary method in the lack of user historical data. However, none of the existing session-based recommendation methods explicitly takes the long-tail recommendation into consideration, which plays an important role in improving the diversity of recommendation and producing the serendipity. As the distribution of items with long-tail is prevalent in session-based recommendation scenarios (e.g., e-commerce, music, and TV program recommendations), more attention should be put on the long-tail session-based recommendation. In this paper, we propose a novel network architecture, namely TailNet, to improve long-tail recommendation performance, while maintaining competitive accuracy performance compared with other methods. We start by classifying items into short-head (popular) and long-tail (niche) items based on click frequency. Then a novel is proposed and applied in TailNet to determine user preference between two types of items, so as to softly adjust and personalize recommendations. Extensive experiments on two real-world datasets verify the superiority of our method compared with state-of-the-art works.