Large Language Models for E‑Commerce

Large language models are changing how people search for, compare, and buy products online, and how sellers and platforms run their businesses. Product search, recommendation, advertising, conversational shopping, customer service, seller tools, and business analytics are all being rebuilt around LLMs.

E-commerce is also a hard research setting. Its data is large, multimodal, multilingual, and changes daily, and it ties directly to real user preferences and real transactions.

This is the third edition of LLM4ECommerce, after The Web Conference 2025 and KDD 2025. This year we put more weight on agentic and multi-agent e-commerce: shopping agents, personalized assistants, tool-using agents, and coordination among customers, sellers, platforms, and the AI agents acting for them. We bring together researchers and practitioners to discuss methods, applications, evaluation, safety, and deployment of LLM-powered e-commerce systems.

Submit papers in English, as a PDF, in the current ACM two-column conference format. For LaTeX, use \documentclass[sigconf,anonymous,review]{acmart}. Review is double-blind, so remove author names and affiliations.

• Paper types Original research, work in progress with preliminary results, and position papers are all welcome.

Accepted papers will be presented as posters, and selected papers will also be given a 10-minute oral slot. Following WSDM policy, at least one author of each accepted paper must present in person. Remote presentations are not supported.

By default, accepted papers will appear in the WSDM 2027 Companion Proceedings. A paper that is not presented in person will be excluded from the proceedings. If you prefer not to have your accepted paper included, let the organizers know and we will opt it out.