@inproceedings{yan-etal-2026-sharechatinsights,
title = "{S}hare{C}hat{I}nsights: An Automated Pipeline for Analyzing Large-Scale Human{--}{LLM} Conversations",
author = "Yan, Yueru and
Gong, Yilin and
Wu, Siqi and
Le, Thai",
editor = "Agrawal, Sweta and
Abad, Alberto",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-demos.54/",
pages = "558--569",
ISBN = "979-8-89176-461-3",
abstract = "Large corpora of conversations between humans and large language models (LLMs) provide a valuable record of how AI systems are used in the wild, but extracting empirical insights from them requires substantial engineering to build custom pipelines that parse multi-platform chat logs and process platform-native artifacts. We present ShareChatInsights, an interactive system that enables natural-language querying and analysis of ShareChat, a large-scale corpus of human-LLM conversations spanning five commercial assistants. An AI agent refines each request through four stages: identify, scope, model, and analyze, turning an open-ended question into a fully specified study that the system executes and exports as a runnable script. Users can configure filters to search, inspect, or export individual conversations on a separate search interface. Users can also opt to publish completed analyses to a shared community gallery. We evaluate the system on 46 requests drawn from prior literature, validate the results against independent LLM and human judgments, and present two empirical case studies. By integrating natural-language querying with reusable analytical infrastructure, ShareChatInsights lowers the barrier to large-scale empirical research on human-LLM conversations for researchers without extensive technical skills."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="yan-etal-2026-sharechatinsights">
<titleInfo>
<title>ShareChatInsights: An Automated Pipeline for Analyzing Large-Scale Human–LLM Conversations</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yueru</namePart>
<namePart type="family">Yan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yilin</namePart>
<namePart type="family">Gong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Siqi</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Thai</namePart>
<namePart type="family">Le</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations</title>
</titleInfo>
<name type="personal">
<namePart type="given">Sweta</namePart>
<namePart type="family">Agrawal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alberto</namePart>
<namePart type="family">Abad</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-461-3</identifier>
</relatedItem>
<abstract>Large corpora of conversations between humans and large language models (LLMs) provide a valuable record of how AI systems are used in the wild, but extracting empirical insights from them requires substantial engineering to build custom pipelines that parse multi-platform chat logs and process platform-native artifacts. We present ShareChatInsights, an interactive system that enables natural-language querying and analysis of ShareChat, a large-scale corpus of human-LLM conversations spanning five commercial assistants. An AI agent refines each request through four stages: identify, scope, model, and analyze, turning an open-ended question into a fully specified study that the system executes and exports as a runnable script. Users can configure filters to search, inspect, or export individual conversations on a separate search interface. Users can also opt to publish completed analyses to a shared community gallery. We evaluate the system on 46 requests drawn from prior literature, validate the results against independent LLM and human judgments, and present two empirical case studies. By integrating natural-language querying with reusable analytical infrastructure, ShareChatInsights lowers the barrier to large-scale empirical research on human-LLM conversations for researchers without extensive technical skills.</abstract>
<identifier type="citekey">yan-etal-2026-sharechatinsights</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-demos.54/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>558</start>
<end>569</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T ShareChatInsights: An Automated Pipeline for Analyzing Large-Scale Human–LLM Conversations
%A Yan, Yueru
%A Gong, Yilin
%A Wu, Siqi
%A Le, Thai
%Y Agrawal, Sweta
%Y Abad, Alberto
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-461-3
%F yan-etal-2026-sharechatinsights
%X Large corpora of conversations between humans and large language models (LLMs) provide a valuable record of how AI systems are used in the wild, but extracting empirical insights from them requires substantial engineering to build custom pipelines that parse multi-platform chat logs and process platform-native artifacts. We present ShareChatInsights, an interactive system that enables natural-language querying and analysis of ShareChat, a large-scale corpus of human-LLM conversations spanning five commercial assistants. An AI agent refines each request through four stages: identify, scope, model, and analyze, turning an open-ended question into a fully specified study that the system executes and exports as a runnable script. Users can configure filters to search, inspect, or export individual conversations on a separate search interface. Users can also opt to publish completed analyses to a shared community gallery. We evaluate the system on 46 requests drawn from prior literature, validate the results against independent LLM and human judgments, and present two empirical case studies. By integrating natural-language querying with reusable analytical infrastructure, ShareChatInsights lowers the barrier to large-scale empirical research on human-LLM conversations for researchers without extensive technical skills.
%U https://aclanthology.org/2026.emnlp-demos.54/
%P 558-569
Markdown (Informal)
[ShareChatInsights: An Automated Pipeline for Analyzing Large-Scale Human–LLM Conversations](https://aclanthology.org/2026.emnlp-demos.54/) (Yan et al., EMNLP 2026)
ACL