@inproceedings{k-r-2026-tvks,
title = "{TVKS}@{MML}ow 2026: Factorized Stance $\times$ Target Classification of {T}amil Political Memes with an Axis-Wise Multimodal Ensemble",
author = "K, Kevin selva and
R, Thilagavathy",
editor = "Ponnusamy, Rahul and
Kumaresan, Prasanna Kumar and
Rajiakodi, Saranya and
B, Premjith and
MC, Shunmuga Priya and
Murugappan, Abirami and
G, Jyothish Lal and
V, Sowmya and
Chakravarthi, Bharathi Raja",
booktitle = "Proceedings of the Fifth Workshop on Multimodal Machine Learning in Low-resource Languages",
month = nov,
year = "2026",
address = "Hengqin, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.mmlow-main.7/",
pages = "70--75",
ISBN = "979-8-89176-445-3",
abstract = "We describe the TVKS submission to the MMLow 2026 shared task on multilevel classification of Tamil political memes, which asks systems to predict a stance label (Level 1: Support/Praise vs. Troll/Oppose) and a four-way target label (Level 2) that combines stance with a Person/Party distinction. We factorize the Level 2 label into two binary axes{---}stance and person/party target{---}reconstructing Level 2 as their product rather than predicting four classes directly. A flat four-way head drives the rarest class (Troll/Oppose Against Party) to an F1 of 0.000; factorization recovers a non-degenerate prediction over all four classes. We also fix a calibration failure caused by stacking a class-balanced sampler with a class-weighted loss, and combine two multimodal models (MuRIL + ViT and MuRIL + ResNet-50) in an axis-wise ensemble that routes each binary axis to its stronger model. Our system reaches a macro-F1 of 0.8594 on Level 1 and 0.5926 on Level 2 (mean 0.7260), ranking second of seven teams. An error analysis attributes 77{\%} of Level 2 errors to the person/party axis, where architecturally distinct visual encoders converge to the same ceiling{---}evidence that the residual gap is dataset ambiguity rather than a modelling deficiency."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="k-r-2026-tvks">
<titleInfo>
<title>TVKS@MMLow 2026: Factorized Stance \times Target Classification of Tamil Political Memes with an Axis-Wise Multimodal Ensemble</title>
</titleInfo>
<name type="personal">
<namePart type="given">Kevin</namePart>
<namePart type="given">selva</namePart>
<namePart type="family">K</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Thilagavathy</namePart>
<namePart type="family">R</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Fifth Workshop on Multimodal Machine Learning in Low-resource Languages</title>
</titleInfo>
<name type="personal">
<namePart type="given">Rahul</namePart>
<namePart type="family">Ponnusamy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Prasanna</namePart>
<namePart type="given">Kumar</namePart>
<namePart type="family">Kumaresan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Saranya</namePart>
<namePart type="family">Rajiakodi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Premjith</namePart>
<namePart type="family">B</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shunmuga</namePart>
<namePart type="given">Priya</namePart>
<namePart type="family">MC</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Abirami</namePart>
<namePart type="family">Murugappan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jyothish</namePart>
<namePart type="given">Lal</namePart>
<namePart type="family">G</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sowmya</namePart>
<namePart type="family">V</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bharathi</namePart>
<namePart type="given">Raja</namePart>
<namePart type="family">Chakravarthi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Hengqin, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-445-3</identifier>
</relatedItem>
<abstract>We describe the TVKS submission to the MMLow 2026 shared task on multilevel classification of Tamil political memes, which asks systems to predict a stance label (Level 1: Support/Praise vs. Troll/Oppose) and a four-way target label (Level 2) that combines stance with a Person/Party distinction. We factorize the Level 2 label into two binary axes—stance and person/party target—reconstructing Level 2 as their product rather than predicting four classes directly. A flat four-way head drives the rarest class (Troll/Oppose Against Party) to an F1 of 0.000; factorization recovers a non-degenerate prediction over all four classes. We also fix a calibration failure caused by stacking a class-balanced sampler with a class-weighted loss, and combine two multimodal models (MuRIL + ViT and MuRIL + ResNet-50) in an axis-wise ensemble that routes each binary axis to its stronger model. Our system reaches a macro-F1 of 0.8594 on Level 1 and 0.5926 on Level 2 (mean 0.7260), ranking second of seven teams. An error analysis attributes 77% of Level 2 errors to the person/party axis, where architecturally distinct visual encoders converge to the same ceiling—evidence that the residual gap is dataset ambiguity rather than a modelling deficiency.</abstract>
<identifier type="citekey">k-r-2026-tvks</identifier>
<location>
<url>https://aclanthology.org/2026.mmlow-main.7/</url>
</location>
<part>
<date>2026-11</date>
<extent unit="page">
<start>70</start>
<end>75</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T TVKS@MMLow 2026: Factorized Stance \times Target Classification of Tamil Political Memes with an Axis-Wise Multimodal Ensemble
%A K, Kevin selva
%A R, Thilagavathy
%Y Ponnusamy, Rahul
%Y Kumaresan, Prasanna Kumar
%Y Rajiakodi, Saranya
%Y B, Premjith
%Y MC, Shunmuga Priya
%Y Murugappan, Abirami
%Y G, Jyothish Lal
%Y V, Sowmya
%Y Chakravarthi, Bharathi Raja
%S Proceedings of the Fifth Workshop on Multimodal Machine Learning in Low-resource Languages
%D 2026
%8 November
%I Association for Computational Linguistics
%C Hengqin, China
%@ 979-8-89176-445-3
%F k-r-2026-tvks
%X We describe the TVKS submission to the MMLow 2026 shared task on multilevel classification of Tamil political memes, which asks systems to predict a stance label (Level 1: Support/Praise vs. Troll/Oppose) and a four-way target label (Level 2) that combines stance with a Person/Party distinction. We factorize the Level 2 label into two binary axes—stance and person/party target—reconstructing Level 2 as their product rather than predicting four classes directly. A flat four-way head drives the rarest class (Troll/Oppose Against Party) to an F1 of 0.000; factorization recovers a non-degenerate prediction over all four classes. We also fix a calibration failure caused by stacking a class-balanced sampler with a class-weighted loss, and combine two multimodal models (MuRIL + ViT and MuRIL + ResNet-50) in an axis-wise ensemble that routes each binary axis to its stronger model. Our system reaches a macro-F1 of 0.8594 on Level 1 and 0.5926 on Level 2 (mean 0.7260), ranking second of seven teams. An error analysis attributes 77% of Level 2 errors to the person/party axis, where architecturally distinct visual encoders converge to the same ceiling—evidence that the residual gap is dataset ambiguity rather than a modelling deficiency.
%U https://aclanthology.org/2026.mmlow-main.7/
%P 70-75
Markdown (Informal)
[TVKS@MMLow 2026: Factorized Stance × Target Classification of Tamil Political Memes with an Axis-Wise Multimodal Ensemble](https://aclanthology.org/2026.mmlow-main.7/) (K & R, MMLow 2026)
ACL