Skip to content
Book Open access

Beyond Independent Predictions: Learning the Browsing Journey for Sequence-Level Re-ranking

Nov 2026 · Proceedings of the 35th ACM International Conference on Information and Knowledge Management · pp. 8752-8759 · 0 citations · 22 references
Computer Science

TL;DR

The Impression-Click Evaluator (ICE), a plug-and-play framework that predicts a browsing-depth distribution to estimate per-position impression probabilities, and leverages predicted impressions to suppress spurious context from unexposed items during click modeling, and introduces a click distribution matching objective to capture intra-sequence dependencies.

Abstract

Re-ranking aims to present users with an ordered list that maximizes sequence-level utility. Industrial deployments widely adopt the Generator--Evaluator framework, where the generator proposes candidate sequences, and the evaluator scores them to select the best one for display. The fidelity of the evaluator's utility estimates thus directly governs recommendation quality. However, existing evaluators predominantly follow the Independent Behavior Modeling (IBM) paradigm, which simplifies the continuous user browsing journey into independent per-item predictions (e.g., CTR). This oversimplification ignores two physical realities: (i) a click can occur only after the item has been reached and impressed, and (ii) items compete for finite user attention. As a result, IBM produces systematically biased evaluations and suboptimal sequence selection. To address this, we propose a Joint Behavior Modeling (JBM) perspective that explicitly models the display ? impression ? click process to capture causal and competitive dependencies within a sequence. Building on JBM, we develop the Impression-Click Evaluator (ICE), a plug-and-play framework that (i) predicts a browsing-depth distribution to estimate per-position impression probabilities, (ii) leverages predicted impressions to suppress spurious context from unexposed items during click modeling, and (iii) introduces a click distribution matching objective to capture intra-sequence dependencies. Extensive offline experiments across diverse backbones, together with a 14-day online A/B test in the Taobao App (+0.70% clicks, +3.67% revenue) and full-traffic deployment, validate the effectiveness and robustness of our approach.

Read PDF

Similar papers

Book Open access Nov 2026

Contextual Ranker in Production: Session-Aware Similar Item Ranking for Lower-Funnel E-Commerce

Contextual Ranker is presented, a production ranking system for eBay's ''Similar Items'' module on product detail pages that encodes recently viewed items with a transformer, computes both session-aware and session-independent candidate relevance, and fuses these heterogeneous signals through a multi-head attention mod...

Yu-Chen Guo, Prudhvi Raj Dachapally · 0 citations
Book Open access Nov 2026

Temporal Moment Generation: A New Recommendation Task and Cost-Efficient Distillation for Outbound Marketing at Scale

Outbound marketing at scale relies on content-first programs and item-ranking pipelines that match a fixed catalog to recent activity, missing the long-horizon interests and seasonal context (e.g., back-to-school, Earth Day) that make a message relevant. Missing these signals has played a key role in engagement decline...

Badari Udupi Bindumadhav · 0 citations
Sep 2026

When LLM-Inferred User Context Adds Value in Production Streaming Recommendation

Contextual information in recommender systems is shifting from static, predefined variables toward latent representations inferred from behavior. Large language models support this shift by rendering an unstructured interaction history as a natural-language summary, which yields a thematic user context that can be enco...

Milad Sabouri, Neeraj Sharma, Sardar Hamidian et al. · 0 citations
Open access 2026

Hodge-Guided Active Preference Elicitation for Efficient Pairwise Ranking

Findings show that Hodge-theoretic structure is informative not only for aggregating collected preferences, but also for deciding which preferences to collect.

Eskander Bejaoui, M. O. Aoueileyine, R. Bouallègue · 0 citations
Preprint Nov 2026

SurvEval: Listwise Re-Ranking Evaluator via Survival-Based Exposure and Multi-Objective Value Modeling

Re-ranking is a crucial stage in recommender systems, yet existing evaluators oversimplify user browsing assumptions and collapse multi-objective signals into a single scalar, limiting their ability to reflect true listwise utility. We propose \textbf{SurvEval}, a listwise evaluator that jointly models user exposure an...

Ya-Chen Yan, Liu-Bo Li · 0 citations
Book Open access Nov 2026

PRISM: Pruning-Enhanced Intent Sampling for Recommendation Multi-List Generation

List Composer is introduced, a decoupled module that transforms pointwise candidate pools into deliverable exposure lists within generative recommender systems within generative recommender systems, and achieves consistent improvements in recommendation quality and serving efficiency.

Zhe-Li Zhou, Si-Thanh Dong, Wei-Wen Liu et al. · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.