2026 SIGIR Workshop on eCommerce


Melbourne, Australia


All previous editions: SIGIReCom'25 | SIGIReCom'24 | SIGIReCom'23 | SIGIReCom'22 | SIGIReCom'21 | SIGIReCom'20 | SIGIReCom'19 | SIGIReCom'18 | SIGIReCom'17

The special theme of this year's workshop is User Interaction and Experience: Agentic-driven Trends

Overview

The SIGIR Workshop on eCommerce will serve as a platform for publication and discussion of Information Retrieval, NLP, and Computer Vision research relative to their applications in the domain of eCommerce. This workshop will bring together practitioners and researchers from academia and industry to discuss the challenges and approaches to product search and recommendation in eCommerce.

The rapid evolution of AI agents, large language models (LLMs), and conversational interfaces is fundamentally reshaping how users interact with eCommerce platforms. As agentic systems begin to mediate search, browsing, and discovery experiences, new challenges emerge around user trust, UX design, evaluation, and system architecture. The goal of our workshop is to bridge these gaps and foster collaboration for this crucial exchange of information between IR academia and industry.

To support this goal, the special theme of ECOM26 is User Interaction and Experience: Agentic-driven Trends.

In addition to submitting papers, there are two additional ways to participate:

  • Data Challenge: in collaboration with the TREC Product Search Track organizers, focused on query reformulation for task-oriented eCommerce queries.
  • Agents Anonymous: an interactive, "group therapy"-style activity. Submit a 1-minute problem pitch of a real world challenge you are actively grappling with, and receive feedback from an expert panel and peers at the workshop.

SIGIR eCom is a full day workshop taking place on July 24, 2026 in conjunction with SIGIR 2026 in Melbourne, Australia. SIGIR eCom'26 will be an in-person workshop. If you have extenuating circumstances that prevent travel, please contact us (contact information at bottom of page).

SIGIR eCom'26 Workshop Schedule

July 24, 2026
9:00 am Workshop Opening
Surya Kallumadi
9:05 am Invited Talk: Agentic eCommerce: Emerging Patterns and Enduring Challenges
Owen Elliott     Marqo
9:50 am Oral Presentations (12 min each)

Scaling Dense Retrieval with LLM-Annotated Training Data: Structured Mining and Progressive Curriculum for E-Commerce Sponsored Search
Md Omar Faruk Rokon, Shasvat Desai, Jhalak Acharya, Isha Shah, Kumar Priyam, Brahanyaa Somasundaram, Vamsee Tangirala, Minuteresa Thomas, Vivek Arora, Vijay Manchi, Hong Yao and Kuang-Chih Lee     Walmart

Pekka: MLLM-based Product Representation Learning with User Engagement Signals for E-Commerce
Rui Yan, Kevin Huang, Miao Li, Zhenyuan Liu, Mina Ghashami, Michael Schlichtkrull, Miguel Arduengo, Wenliang Gao, Bella Shi, Michael Du, Fan Xia, Zhaoheng Zheng, I-Ta Lee, Minghai Chen, Yash Desai, Rana Alkhoury Maroun, Yugandhar Nanda, Weijian Han, Dongxin Liu, Santiago Villa Fernandez, Jack Wang, Guang Yang, Saurabh Gupta, Ben Schulte, Deepak Chandra and Damien Lefortier     Meta

A Bag-of-Documents Model for Query Specificity
Aritra Mandal, Daniel Tunkelang and Zhe Wu     eBay
10:30 am Break
11:00 am Panel: User Interaction and Experience: Agentic-driven Trends
11:45 am Oral Presentations (12 min each)

AutoRelAnnotator: Calibrated Model Cascades for Cost-Efficient Relevance Evaluation in Sponsored Search
Md Omar Faruk Rokon, Shasvat Desai, Hong Yao and Kuang-Chih Lee     Walmart

Do We Need Retrieval for Query Correction? A Study of RAG Approaches in Quick Commerce Search
Naman Jain     Uber
12:10 pm Poster Teasers (1 min each)

INSPIRE: Intent-aware Neural Sponsored Product Retrieval for E-commerce
Shasvat Desai, Hong Yao, Utkarsh Porwal and Kuang-Chih Lee     Walmart

Improving Pointwise Predictions with a Pairwise Ranker
Eric Hallman     eBay

Scalable Visual Attribute Recognition in E-Commerce Products via Automated Synthetic Label Generation
Victor Martinez, Virginia Negri, Brayan Impata and Sergio Alvarez     Amazon

Problem Formulation over Feature Engineering: Evidence from a Production Reranking System in a C2C Marketplace
Ryo Watanabe, Yuki Yada, Yusuke Shido, Shinya Yaginuma and Ryo Tanaka     Mercari Inc.

Robust HNSW Vector Retrieval under Dynamic Product Updates via Local Rewiring
Anurag Sengupta     Independent

Product-Aware Architectures for Conversion Rate Prediction in E-Commerce Advertising
Aditya Chichani, Ameya Raul, Amey Dharwadker and Saurabh Gupta     Meta

Unified Multi-Task Relevance Modeling for E-Commerce: Comparing Task Routing Architectures Across LLMs and Cross-Encoders
Md Omar Faruk Rokon, Jhalak Acharya, Shasvat Desai, Hong Yao and Kuang-Chih Lee     Walmart

Agentic Prompt Optimization for E-Commerce Catalog Enrichment at Scale
Peng Gao, Athanasios N. Nikolakopoulos, Zhu Cheng, Andrea Scarinci, Umit Batur and Suleiman Khan     Amazon Catalog AI

Structured & Tiered Agentic Response Evaluation (STARe): LLM-Judge Framework for Conversational Shopping Agents
Akash Singh, Vishal Shivhare, Digvijay Bhadouria, Flint Luu, Pinchu Sabu, Vikram Singh, Sanket Thapliyal, Pierre Vandenbussche and Wenjia Xie     eBay

Understanding and Improving Decision-Making UX in Conversational E-Commerce Agents via Evidence Selection
Tianqi Liu     University of Delaware

CaptionFuse: Training-Free Composed Product Retrieval Via VLM Intent Captioning and Weighted Embedding Fusion
Zhao Gao, Mina Metias, Mohamed Noordeen Alaudeen and Bassel Saleh     AWS GenAI Innovation Center

Bottoms, Towers, and Tops: Evaluating Component Interactions in Multi-Task Learning for E-Commerce
Piotr Bajger, Aleksander Wawer and Wojciech Kutak     Allegro

Marginal Expected Revenue for Jointly Ranking Auction and Fixed-Price Listings in E-Commerce Sponsored Search
Greg Kocher and Sanjana Arun     eBay

Route, Don’t Guess: Adaptive Interaction Strategy Selection for Agentic Product Search
Vinesh Gudla, Xiao Xiao, Ji Xin and Tejaswi Tenneti     Instacart, Ambience Healthcare
12:30 pm Lunch
1:30 pm Invited Talk: From Product Search to Shopping Agents at Scale
Chen Luo     Amazon
2:15 pm Poster Session
3:00 pm Break
3:30 pm Oral Presentations

LoSparse: Low-dimensional Sparse Retrieval Model
Kohei Uno, Shogo Kisa and Yohei Iseki     LY Corporation

ECOM26 Data Challenge
Dean Alvarez

3:55 pm Agents Anonymous
4:45 pm Closing
 

Important Dates

Paper submission deadline May 2, 2026 (11:59 P.M. AoE)
Notification of acceptance May 21, 2026
Camera ready version of papers due June 15, 2026
SIGIR eCom full day workshop July 24, 2026

Call For Papers

We invite quality research contributions, position and opinion papers addressing relevant challenges in the domain of eCommerce. We invite submission of papers and posters representing original research, preliminary research results, proposals for new work, position and opinion papers. All submitted papers and posters will be single-blind and will be peer reviewed by an international program committee of researchers of high repute. Accepted submissions will be presented at the workshop.

Topics of interest include, but are not limited to:

  • User Interaction and Experience: Agentic-driven Trends (2026 special theme)
    • Agentic search, browsing, and discovery experiences in eCommerce
    • Trade-offs between accuracy, cost, latency, and scalability in agentic pipelines
    • User interaction patterns and UX considerations for agentic and conversational systems
    • Offline and online evaluation of agentic systems
    • Mobile vs. desktop experiences for agentic and conversational interfaces
    • Adapting AB testing methodologies for new UI paradigms
    • Personalized experiences while maintaining user trust and fair marketplaces
    • Transitioning users from current eCommerce UX to agentic experiences
  • IR Fundamentals for eCommerce
    • Personalization and contextualization
    • Unified and universal search and recommendations
    • Indexing and search in rapidly changing environments (e.g., auction sites)
    • Experimentation techniques
    • Small Language Models (SLMs) for efficient and scalable eCommerce IR
    • Cross-lingual search and machine translation
  • Ads and Sponsored Products in eCommerce
    • Sponsored search and recommendations
    • Joint optimization of organic and sponsored results
    • Auctions, bidding, and ranking for sponsored products
    • Ads impact on relevance and user experience
  • Ranking and Whole Page Relevance (WPR)
    • Optimization for IR and business metrics
    • Diversity in product search and recommendations
    • Relevance models for multi-faceted entities
    • Relevance vs. revenue trade-offs
    • Deterministic sorts (e.g., price low to high)
    • Temporal dynamics and seasonality
  • Query and Document Understanding
    • Query intent, query suggestions, and auto-completion
    • Strategies for resolving low or zero recall queries
    • Cross-modal search (e.g., text, structured data, images)
    • Categorization and facets
    • Reviews and sentiment analysis
  • Recommendation and Personalization
    • Personalization & contextualization, including the use of personal facets such as age, gender, location
    • Privacy, bias and ethics in eCommerce IR
    • Blending recommendations and search results
  • Representations and Data
    • Semantic representation of products, queries, and customers
    • Construction and use of knowledge graphs for eCommerce
    • Large foundational recommendation models
  • Visual Search in eCommerce
    • Large-scale visual search challenges and solutions
    • Multimodal search and combining visual and textual information
    • Combining vision and language models
    • Explainable AI for visual search
  • Other Challenges
    • Trust, transparency, and fairness in eCommerce
    • UX for eCommerce
    • The role of search in trust and security for marketplaces
    • Question answering and chatbots for eCommerce
    • Fairness and transparency at the system and ecosystem level

Data/Resource Track

In order to promote academic research in the eCommerce domain, we plan to accept a small number of high quality dataset contributions. These submissions should be accompanied by a clear and detailed description of the dataset, some potential questions and applications that arise from it. Preliminary empirical investigations conveying any insight about the data will increase the quality of the submission.

Submission Instructions

All papers will be peer reviewed (single-blind) by the program committee and judged by their relevance to the workshop, especially to the main themes identified above, and their potential to generate discussion. All submissions must be in PDF formatted according to the latest CEUR single column format available at https://www.overleaf.com/latex/templates/template-for-submissions-to-ceur-workshop-proceedings-ceur-ws-dot-org/wqyfdgftmcfw. The short (8-page) and long (15-page) limits are extended to account for this.

For instructions and LaTeX/Overleaf/docx templates, see: https://ceur-ws.org/HOWTOSUBMIT.html#CEURART

  • Read up to and including the "License footnote in paper PDFs" section.
  • Please Use Emphasizing Capitalized Style for Paper Titles.
  • Double check \copyrightyear{2026} for the correct year
  • Use the following conference name styling: \conference{ECOM'26: SIGIR Workshop on eCommerce, Jul 24, 2026, Melbourne, Australia}
Submissions must describe work that is not previously published, not accepted for publication elsewhere, and not currently under review elsewhere. All submissions must be in English. The workshop follows a single-blind reviewing process. We do not accept anonymized submissions. Please note that at least one of the authors of each accepted paper must register for the workshop and present the paper in person.

Long paper limit: 15 pages. References are not counted in the page limit.
Short paper limit: 8 pages. References are not counted in the page limit.

Submissions to SIGIR eCom will be made through the SIGIR EasyChair instance. Select "2026 SIGIR Workshop on eCommerce" as the track: https://easychair.org/conferences/?conf=sigir2026

Data Challenge

ECOM26 will host a data challenge by building on the Search task from the TREC Product Search and Recommendations Track on query reformulation for task-oriented eCommerce queries. The data challenge will consist of two phases: An open phase and a closed phase. The open phase will begin in mid-April and run through early June. The close phase will begin just after the end of the open phase and run for two weeks. Precise dates and a link to the competition portal will be posted soon.

Agents Anonymous

Agents Anonymous is an interactive, "group therapy"-style activity designed to facilitate candid discussion of hard-earned lessons, open questions, and frustrations of building real-world systems. You will submit a brief problem pitch describing a challenge you are currently grappling with.

The organizers will select 4–5 cases and the selected cases will be discussed during the workshop in a structured, time-boxed format that includes a problem pitch, clarifying questions, expert panel feedback, open group brainstorming, and facilitated synthesis. The emphasis is on sharing patterns, heuristics, and battle-tested experience to create space for insights that are rarely discussed in formal academic settings and create a sense of shared community.

Submission link is now open!