Conformity interest-aware weighted HyperGraph neural network for session-based recommendation

Dongjing Wang

doi:10.1016/j.asoc.2026.115059

Conformity interest-aware weighted HyperGraph neural network for session-based recommendation

2026-03-18

0

PRE

AI

DOI：10.1016/j.asoc.2026.115059

Original

Original request for help

Save

Abstract

En 中文

• We propose WHCN to capture personalized and conformity interests via leveraging item and category frequency distributions. • We devise CIFN for adaptive integration of users’ long- and short-term interests with conformity interests. • We introduce a DSSL module to enhance item and category representation learning via auxiliary supervision. • Extensive experiments on four real-world datasets verify the effectiveness of CWHNN.

Keywords:

Conformity interest

Weighted HyperGraph

Session-based recommendation

Interest integration

Representation learning

Conformity interest-aware weighted HyperGraph neural network for session-based recommendation

Abstract

Journal

Organization