RouteRec: Behavior-Guided Sparse Routing for Sequential Recommendation
Open Access
- 6 November 2026
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 4452-4463
Abstract
Sessionized interaction histories contain behavioral patterns that can improve sequential recommendation. However, existing models process all sessions through the same parameterized blocks, regardless of their behavioral differences. Mixture of Experts (MoE) enables conditional computation, but it leaves open what should guide expert allocation. We propose RouteRec, a sequential recommender that uses observed session behavior as the routing criterion. RouteRec summarizes four types of behavioral evidence from sessionized histories: interaction tempo, item-group focus, repetition and carryover, and popularity tendency. It uses these cues to route computation at macro, mid, and micro scopes. Cue-derived scores first select expert groups; within each selected group, the current backbone state then refines expert selection. Across six public datasets and 18 dataset--metric combinations, RouteRec ranks first in 12 and second in three, yielding the best overall average rank of 1.61 compared with 4.11 for the next-best baseline. Additional analyses suggest that the behavioral cues guide expert allocation beyond added capacity and produce routing patterns aligned with observed behavior. Our code is available at https://github.com/jy1559/RouteRec.Keywords
Funding Information
- Institute of Information & Communications Technology Planning & Evaluation (RS-2026-25526850)
This publication has 35 references indexed in Scilit:
- A survey on popularity bias in recommender systemsUser Modeling and User-Adapted Interaction, 2024
- Session-aware recommendation: A surprising quest for the state-of-the-artInformation Sciences, 2021
- Self-Attentive Sequential Recommendation2018
- Personalized Top-N Sequential Recommendation via Convolutional Sequence Embedding2018
- Neural Attentive Session-based Recommendation2017
- The MovieLens DatasetsACM Transactions on Interactive Intelligent Systems, 2015
- Inferring Networks of Substitutable and Complementary Products2015
- Modeling User Activity Preference by Leveraging User Spatial Temporal Characteristics in LBSNsIEEE Transactions on Systems, Man, and Cybernetics: Systems, 2014
- Factorizing personalized Markov chains for next-basket recommendation2010
- Music Recommendation and Discovery2010