@inproceedings{mohammad-2026-polafusion,
title = "{P}ola{F}usion at {S}em{E}val-2026 Task 9: Ensemble Transformers with Targeted Augmentation for Multilingual Polarization Detection",
author = "Mohammad, Abdullah",
editor = "Kochmar, Ekaterina and
Ghosh, Debanjan and
North, Kai and
Komachi, Mamoru",
booktitle = "Proceedings of the 20th {I}nternational {W}orkshop on {S}emantic {E}valuation (2026)",
month = jul,
year = "2026",
address = "San Diego, California, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.semeval-1.361/",
pages = "2877--2885",
ISBN = "979-8-89176-414-9",
abstract = "We present PolaFusion, our system for SemEval-2026 Task 9, which requires detecting polarization in social media posts across 22 languages, classifying its type (Subtask 2), and identifying its rhetorical manifestation (Subtask 3). The task is characterized by severe and pervasive class imbalance across all three subtasks and all 22 languages. We address this through a combination of three strategies: a hierarchical gating architecture where a binary gatekeeper model gates two specialist classifiers trained exclusively on polarized content; an eight-model mega-ensemble combining fivefold mDeBERTa-v3-base and three-fold XLM-RoBERTa-large with soft-vote probability aggregation; and a Macro-F1-aware augmentation strategy using Qwen3-235B that generates synthetic minority-class examples only for language-label pairs that are both scarce and poorly learned. Throughout training, inverse-frequency class weighting within BCEWithLogitsLoss forces the model to attend proportionally to rare labels. Our system achieves official Macro-F1 scores of 0.800, 0.576, and 0.502 on Subtasks 1{--}3 respectively, outperforming the POLAR baseline by +0.040, +0.089, and +0.082 average Macro-F1 across languages. Our code is publicly available at https://github.com/Abdullah4152/PolaFuse."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="mohammad-2026-polafusion">
<titleInfo>
<title>PolaFusion at SemEval-2026 Task 9: Ensemble Transformers with Targeted Augmentation for Multilingual Polarization Detection</title>
</titleInfo>
<name type="personal">
<namePart type="given">Abdullah</namePart>
<namePart type="family">Mohammad</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 20th International Workshop on Semantic Evaluation (2026)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Kochmar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Debanjan</namePart>
<namePart type="family">Ghosh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kai</namePart>
<namePart type="family">North</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mamoru</namePart>
<namePart type="family">Komachi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-414-9</identifier>
</relatedItem>
<abstract>We present PolaFusion, our system for SemEval-2026 Task 9, which requires detecting polarization in social media posts across 22 languages, classifying its type (Subtask 2), and identifying its rhetorical manifestation (Subtask 3). The task is characterized by severe and pervasive class imbalance across all three subtasks and all 22 languages. We address this through a combination of three strategies: a hierarchical gating architecture where a binary gatekeeper model gates two specialist classifiers trained exclusively on polarized content; an eight-model mega-ensemble combining fivefold mDeBERTa-v3-base and three-fold XLM-RoBERTa-large with soft-vote probability aggregation; and a Macro-F1-aware augmentation strategy using Qwen3-235B that generates synthetic minority-class examples only for language-label pairs that are both scarce and poorly learned. Throughout training, inverse-frequency class weighting within BCEWithLogitsLoss forces the model to attend proportionally to rare labels. Our system achieves official Macro-F1 scores of 0.800, 0.576, and 0.502 on Subtasks 1–3 respectively, outperforming the POLAR baseline by +0.040, +0.089, and +0.082 average Macro-F1 across languages. Our code is publicly available at https://github.com/Abdullah4152/PolaFuse.</abstract>
<identifier type="citekey">mohammad-2026-polafusion</identifier>
<location>
<url>https://aclanthology.org/2026.semeval-1.361/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>2877</start>
<end>2885</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T PolaFusion at SemEval-2026 Task 9: Ensemble Transformers with Targeted Augmentation for Multilingual Polarization Detection
%A Mohammad, Abdullah
%Y Kochmar, Ekaterina
%Y Ghosh, Debanjan
%Y North, Kai
%Y Komachi, Mamoru
%S Proceedings of the 20th International Workshop on Semantic Evaluation (2026)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, USA
%@ 979-8-89176-414-9
%F mohammad-2026-polafusion
%X We present PolaFusion, our system for SemEval-2026 Task 9, which requires detecting polarization in social media posts across 22 languages, classifying its type (Subtask 2), and identifying its rhetorical manifestation (Subtask 3). The task is characterized by severe and pervasive class imbalance across all three subtasks and all 22 languages. We address this through a combination of three strategies: a hierarchical gating architecture where a binary gatekeeper model gates two specialist classifiers trained exclusively on polarized content; an eight-model mega-ensemble combining fivefold mDeBERTa-v3-base and three-fold XLM-RoBERTa-large with soft-vote probability aggregation; and a Macro-F1-aware augmentation strategy using Qwen3-235B that generates synthetic minority-class examples only for language-label pairs that are both scarce and poorly learned. Throughout training, inverse-frequency class weighting within BCEWithLogitsLoss forces the model to attend proportionally to rare labels. Our system achieves official Macro-F1 scores of 0.800, 0.576, and 0.502 on Subtasks 1–3 respectively, outperforming the POLAR baseline by +0.040, +0.089, and +0.082 average Macro-F1 across languages. Our code is publicly available at https://github.com/Abdullah4152/PolaFuse.
%U https://aclanthology.org/2026.semeval-1.361/
%P 2877-2885
Markdown (Informal)
[PolaFusion at SemEval-2026 Task 9: Ensemble Transformers with Targeted Augmentation for Multilingual Polarization Detection](https://aclanthology.org/2026.semeval-1.361/) (Mohammad, SemEval 2026)
ACL