@inproceedings{zhou-etal-2025-texts,
title = "Texts or Images? A Fine-grained Analysis on the Effectiveness of Input Representations and Models for Table Question Answering",
author = "Zhou, Wei and
Mesgar, Mohsen and
Adel, Heike and
Friedrich, Annemarie",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2025",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-acl.117/",
doi = "10.18653/v1/2025.findings-acl.117",
pages = "2307--2318",
ISBN = "979-8-89176-256-5",
abstract = "In table question answering (TQA), tables are encoded as either texts or images. Prior work suggests that passing images of tables to multi-modal large language models (MLLMs) performs comparably to using textual input with large language models (LLMs). However, the lack of controlled setups limits fine-grained distinctions between these approaches. In this paper, we conduct the first controlled study on the effectiveness of several combinations of table representations and model types from two perspectives: question complexity and table size. We build a new benchmark based on existing TQA datasets. In a systematic analysis of seven pairs of MLLMs and LLMs, we find that the best combination of table representation and model varies across setups. We propose FRES, a method selecting table representations dynamically, and observe a 10{\%} average performance improvement compared to using both representations indiscriminately."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhou-etal-2025-texts">
<titleInfo>
<title>Texts or Images? A Fine-grained Analysis on the Effectiveness of Input Representations and Models for Table Question Answering</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wei</namePart>
<namePart type="family">Zhou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohsen</namePart>
<namePart type="family">Mesgar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Heike</namePart>
<namePart type="family">Adel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Annemarie</namePart>
<namePart type="family">Friedrich</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wanxiang</namePart>
<namePart type="family">Che</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joyce</namePart>
<namePart type="family">Nabende</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Shutova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammad</namePart>
<namePart type="given">Taher</namePart>
<namePart type="family">Pilehvar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-256-5</identifier>
</relatedItem>
<abstract>In table question answering (TQA), tables are encoded as either texts or images. Prior work suggests that passing images of tables to multi-modal large language models (MLLMs) performs comparably to using textual input with large language models (LLMs). However, the lack of controlled setups limits fine-grained distinctions between these approaches. In this paper, we conduct the first controlled study on the effectiveness of several combinations of table representations and model types from two perspectives: question complexity and table size. We build a new benchmark based on existing TQA datasets. In a systematic analysis of seven pairs of MLLMs and LLMs, we find that the best combination of table representation and model varies across setups. We propose FRES, a method selecting table representations dynamically, and observe a 10% average performance improvement compared to using both representations indiscriminately.</abstract>
<identifier type="citekey">zhou-etal-2025-texts</identifier>
<identifier type="doi">10.18653/v1/2025.findings-acl.117</identifier>
<location>
<url>https://aclanthology.org/2025.findings-acl.117/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>2307</start>
<end>2318</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Texts or Images? A Fine-grained Analysis on the Effectiveness of Input Representations and Models for Table Question Answering
%A Zhou, Wei
%A Mesgar, Mohsen
%A Adel, Heike
%A Friedrich, Annemarie
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Findings of the Association for Computational Linguistics: ACL 2025
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-256-5
%F zhou-etal-2025-texts
%X In table question answering (TQA), tables are encoded as either texts or images. Prior work suggests that passing images of tables to multi-modal large language models (MLLMs) performs comparably to using textual input with large language models (LLMs). However, the lack of controlled setups limits fine-grained distinctions between these approaches. In this paper, we conduct the first controlled study on the effectiveness of several combinations of table representations and model types from two perspectives: question complexity and table size. We build a new benchmark based on existing TQA datasets. In a systematic analysis of seven pairs of MLLMs and LLMs, we find that the best combination of table representation and model varies across setups. We propose FRES, a method selecting table representations dynamically, and observe a 10% average performance improvement compared to using both representations indiscriminately.
%R 10.18653/v1/2025.findings-acl.117
%U https://aclanthology.org/2025.findings-acl.117/
%U https://doi.org/10.18653/v1/2025.findings-acl.117
%P 2307-2318
Markdown (Informal)
[Texts or Images? A Fine-grained Analysis on the Effectiveness of Input Representations and Models for Table Question Answering](https://aclanthology.org/2025.findings-acl.117/) (Zhou et al., Findings 2025)
ACL