Lune

ACL Findings2026

SessionIntentBench: A Multi-task Inter-session Intention-shift Modeling Benchmark for E-commerce Customer Behavior Understanding

Yuqi Yang, Weiqi Wang, Baixuan Xu, Wei Fan, Qing Zong, Chunkit Chan, Zheye Deng, Xin Liu, Yifan Gao, Changlong Yu, Chen Luo, Yang Li

Abstract

Session history is a common way of recording user interaction behaviors throughout a browsing activity involving multiple products. For example, if a user clicks on a product webpage and then leaves, it might be because certain features do not satisfy the user, which serves as an important indicator of on-the-spot user preferences. However, prior works fail to capture and model customer intention effectively because of insufficient information exploitation, relying only on apparent information such as descriptions and titles. There is also a lack of data and corresponding benchmarks for explicitly modeling intention in E-commerce product purchase sessions. To address these issues, we introduce the concept of an intention tree and propose a dataset curation pipeline. Together, we construct a sibling multimodal benchmark, SESSIONINTENTBENCH, that evaluates L(V)LMs' capability to understand intersession intention shifts through four subtasks. With 1,952,177 intention entries, 1,132,145 session intention trajectories, and 13,003,664 available tasks mined from 10,905 sessions, we provide a scalable way to exploit existing session data for customer intention understanding. We conduct human annotations to collect ground-truth labels for a subset of the collected data to form an evaluation gold set. Extensive experiments on the annotated data further confirm that current L(V)LMs fail to capture and utilize intention across complex session settings. Further analysis shows that injecting intention enhances LLM performance.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

Builds onCited byLeft to right by year; the more cited, the higher.

Builds on13

  1. Chain-of-Thought Prompting Elicits Reasoning in Large Language ModelsJason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma et al.NeurIPS 2022 · 22,562 citations
  2. Visual Instruction TuningHaotian Liu, Chunyuan Li, Qingyang Wu, Yong Jae LeeNeurIPS 2023 · 11,349 citations
  3. Bridging Language and Items for Retrieval and Recommendation: Benchmarking LLMs as Semantic EncodersYupeng Hou, Jiacheng Li, Xiangjun Fu, Zhankui He et al.ACL 2026 · 346 citations
  4. Dual Intent Enhanced Graph Neural Network for Session-based New Item RecommendationDi Jin, Luzhi Wang, Yizhen Zheng, Guojie Song et al.WWW 2023 · 47 citations
  5. Enhancing User Intent Capture in Session-Based Recommendation with Attribute PatternsXin Liu, Zheng Li, Yifan Gao, Jingfeng Yang et al.NeurIPS 2023 · 30 citations

Related papers

  1. IntentionQA: A Benchmark for Evaluating Purchase Intention Comprehension Abilities of Language Models in E-commerceWenxuan Ding, Weiqi Wang, Sze Heng Douglas Kwok, Minghao Liu et al.EMNLP Findings 2024 · 1 citation
  2. EcomScriptBench: A Multi-task Benchmark for E-commerce Script Planning via Step-wise Intention-Driven Product AssociationWeiqi Wang, Limeng Cui, Xin Liu, Sreyashi Nag et al.ACL 2025 · 15 citations
  3. MIND: Multimodal Shopping Intention Distillation from Large Vision-language Models for E-commerce Purchase UnderstandingBaixuan Xu, Weiqi Wang, Haochen Shi, Wenxuan Ding et al.EMNLP 2024 · 4 citations
  4. eCeLLM: Generalizing Large Language Models for E-commerce from Large-scale, High-quality Instruction DataBo Peng, Xinyi Ling, Ziru Chen, Huan Sun et al.ICML 2024 · 53 citations
  5. ShoppingBench: A Real-World Intent-Grounded Shopping Benchmark for LLM-based AgentsJiangyuan Wang, Kejun Xiao, Qi Sun, Huaipeng Zhao et al.AAAI 2026 · 14 citations
Dusk over the sea between two cliffs drawn in fine vertical lines