AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Multi-granularity Intent Modeling with Adversarial Robustness for Sequential Recommendation

Yangyi Fang, Haolin Shi

AAAI 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Recommendation representation learning · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use MIMAR-SRec instead of strongest baseline ICSRecsequential recommendation · SportsNDCG@20: 0.0381 -> 0.0482Improved
use MIMAR-SRec instead of strongest baseline on dense datasetsequential recommendation · ML-1MHR@10: 0.3222 -> 0.3392Improved
add multi-granularity intent alignment (MIA) (ablation)MIMAR-SRec · SportsNDCG@20: 0.0454 -> 0.0482Improved
add cross-user target-aware contrast (CUTAC) (ablation)MIMAR-SRec · SportsNDCG@20: 0.0423 -> 0.0482Improved
add intent-similarity adversarial robustness (ISAR) (ablation)MIMAR-SRec · SportsNDCG@20: 0.0437 -> 0.0482Improved
add cross-user target-aware contrast (CUTAC) (ablation)MIMAR-SRec · ML-1MNDCG@20: 0.2264 -> 0.2313Improved
add MIA and CUTAC together (ablation)MIMAR-SRec · SportsNDCG@20: 0.0401 -> 0.0482Improved
add CUTAC and ISAR together (ablation)MIMAR-SRec · SportsNDCG@20: 0.0395 -> 0.0482Improved
inject 10% to 50% noisy interactions into test sequencesMIMAR-SRec vs CL4SRec and ICSRec · SportsNDCG@20Improved
increase number of intent prototypes K beyond a moderate value (ablation)MIMAR-SRec · all datasetsNDCG@20Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy