@inproceedings{abdul-hakim-etal-2026-m3ec,
title = "{M}${}^3${EC}: A Multimodal Multidocument Benchmark for Event Extraction and Event Coreference Resolution",
author = "Abdul Hakim, Zaber Ibn and
Sarker, Najibul Haque and
Hussain, Aafiya and
Alomari, Hani and
Ishmam, Alvi Md and
Asgarov, Ali and
Tang, Chia-Wei and
Thomas, Chris",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.2502/",
pages = "56843--56904",
ISBN = "979-8-89176-459-0",
abstract = "Understanding how the same real-world event is represented across text, video, and audio remains a fundamental challenge in information extraction. While prior datasets have explored events within individual modalities or limited modality pairs, a setting that combines all three modalities remains largely unexplored. We introduce M${}^3$EC (Multimodal Multidocument Event Coreference), a new benchmark for multimodal, multidocument event extraction and coreference resolution. It is built from tightly related news clusters that combine eventful videos, audio files, and news articles. We also introduce a semi-automated annotation pipeline, AutoEventAnnotator, to generate automated annotations at scale. Additionally, we construct a human-annotated test set for robust evaluation. To our knowledge, this is the first benchmark to support cluster-level multidocument event extraction and event coreference across text, video, and audio. Evaluation with strong multimodal LLMs shows that the majority of them fail to extract valid events in video and particularly in audio. We further show that models finetuned on the annotations generated by our automated pipeline achieve consistent improvements, including more than +30 F1 on video event extraction, +15 F1 on audio event extraction, and +40 CoNLL F1 on end-to-end coreference resolution, highlighting the usefulness of M${}^3$EC for training and evaluation in multimodal event understanding."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="abdul-hakim-etal-2026-m3ec">
<titleInfo>
<title>M³EC: A Multimodal Multidocument Benchmark for Event Extraction and Event Coreference Resolution</title>
</titleInfo>
<name type="personal">
<namePart type="given">Zaber</namePart>
<namePart type="given">Ibn</namePart>
<namePart type="family">Abdul Hakim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Najibul</namePart>
<namePart type="given">Haque</namePart>
<namePart type="family">Sarker</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aafiya</namePart>
<namePart type="family">Hussain</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hani</namePart>
<namePart type="family">Alomari</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alvi</namePart>
<namePart type="given">Md</namePart>
<namePart type="family">Ishmam</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ali</namePart>
<namePart type="family">Asgarov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chia-Wei</namePart>
<namePart type="family">Tang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chris</namePart>
<namePart type="family">Thomas</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</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</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">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Understanding how the same real-world event is represented across text, video, and audio remains a fundamental challenge in information extraction. While prior datasets have explored events within individual modalities or limited modality pairs, a setting that combines all three modalities remains largely unexplored. We introduce M³EC (Multimodal Multidocument Event Coreference), a new benchmark for multimodal, multidocument event extraction and coreference resolution. It is built from tightly related news clusters that combine eventful videos, audio files, and news articles. We also introduce a semi-automated annotation pipeline, AutoEventAnnotator, to generate automated annotations at scale. Additionally, we construct a human-annotated test set for robust evaluation. To our knowledge, this is the first benchmark to support cluster-level multidocument event extraction and event coreference across text, video, and audio. Evaluation with strong multimodal LLMs shows that the majority of them fail to extract valid events in video and particularly in audio. We further show that models finetuned on the annotations generated by our automated pipeline achieve consistent improvements, including more than +30 F1 on video event extraction, +15 F1 on audio event extraction, and +40 CoNLL F1 on end-to-end coreference resolution, highlighting the usefulness of M³EC for training and evaluation in multimodal event understanding.</abstract>
<identifier type="citekey">abdul-hakim-etal-2026-m3ec</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.2502/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>56843</start>
<end>56904</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T M³EC: A Multimodal Multidocument Benchmark for Event Extraction and Event Coreference Resolution
%A Abdul Hakim, Zaber Ibn
%A Sarker, Najibul Haque
%A Hussain, Aafiya
%A Alomari, Hani
%A Ishmam, Alvi Md
%A Asgarov, Ali
%A Tang, Chia-Wei
%A Thomas, Chris
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F abdul-hakim-etal-2026-m3ec
%X Understanding how the same real-world event is represented across text, video, and audio remains a fundamental challenge in information extraction. While prior datasets have explored events within individual modalities or limited modality pairs, a setting that combines all three modalities remains largely unexplored. We introduce M³EC (Multimodal Multidocument Event Coreference), a new benchmark for multimodal, multidocument event extraction and coreference resolution. It is built from tightly related news clusters that combine eventful videos, audio files, and news articles. We also introduce a semi-automated annotation pipeline, AutoEventAnnotator, to generate automated annotations at scale. Additionally, we construct a human-annotated test set for robust evaluation. To our knowledge, this is the first benchmark to support cluster-level multidocument event extraction and event coreference across text, video, and audio. Evaluation with strong multimodal LLMs shows that the majority of them fail to extract valid events in video and particularly in audio. We further show that models finetuned on the annotations generated by our automated pipeline achieve consistent improvements, including more than +30 F1 on video event extraction, +15 F1 on audio event extraction, and +40 CoNLL F1 on end-to-end coreference resolution, highlighting the usefulness of M³EC for training and evaluation in multimodal event understanding.
%U https://aclanthology.org/2026.emnlp-main.2502/
%P 56843-56904
Markdown (Informal)
[M3EC: A Multimodal Multidocument Benchmark for Event Extraction and Event Coreference Resolution](https://aclanthology.org/2026.emnlp-main.2502/) (Abdul Hakim et al., EMNLP 2026)
ACL
- Zaber Ibn Abdul Hakim, Najibul Haque Sarker, Aafiya Hussain, Hani Alomari, Alvi Md Ishmam, Ali Asgarov, Chia-Wei Tang, and Chris Thomas. 2026. M3EC: A Multimodal Multidocument Benchmark for Event Extraction and Event Coreference Resolution. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 56843–56904, Budapest, Hungary. Association for Computational Linguistics.