@inproceedings{saha-etal-2026-reference,
title = "Reference-Free Schema Generation for Literature Review Tables via Multi-Faceted Rewards",
author = "Saha, Sinjoy and
Saha, Suman and
Farooque, Mahfuza and
Yin, Wenpeng",
editor = "T.Y.S.S., Santosh and
Rodriguez, Juan Diego and
de Gibert, Ona",
booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics ({ACL} 2026)",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.acl-srw.111/",
pages = "1253--1261",
ISBN = "979-8-89176-393-7",
abstract = "To accelerate scientific knowledge acquisition, LLMs are increasingly used to synthesize multiple papers into structured tables by inferring schemas and values. While value generation within a fixed schema can often be reduced to extractive question answering, the schema generation problem, determining which dimensions to compare a set of documents, lacks a formal mapping to standard NLP tasks. In this work, we formulate schema generation as a reinforcement learning problem and investigate whether these dimensions can be induced without access to gold-standard schemas. We design a multi-faceted reward framework capturing schema coverage, non-redundancy, relevance, and format, and train a small language model on a literature review dataset. Our approach yields consistent improvements over the untuned base model across intrinsic, reference-based, and LLM-judge metrics, and remains competitive with supervised fine-tuned models at 5$\times$ the parameter count on structural and diversity dimensions. All code, results and prompts are available in the GitHub repository: \url{https://github.com/sinjoysaha/rl-schema-generation}"
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="saha-etal-2026-reference">
<titleInfo>
<title>Reference-Free Schema Generation for Literature Review Tables via Multi-Faceted Rewards</title>
</titleInfo>
<name type="personal">
<namePart type="given">Sinjoy</namePart>
<namePart type="family">Saha</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Suman</namePart>
<namePart type="family">Saha</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mahfuza</namePart>
<namePart type="family">Farooque</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wenpeng</namePart>
<namePart type="family">Yin</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 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Santosh</namePart>
<namePart type="family">T.Y.S.S.</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Juan</namePart>
<namePart type="given">Diego</namePart>
<namePart type="family">Rodriguez</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ona</namePart>
<namePart type="family">de Gibert</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, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-393-7</identifier>
</relatedItem>
<abstract>To accelerate scientific knowledge acquisition, LLMs are increasingly used to synthesize multiple papers into structured tables by inferring schemas and values. While value generation within a fixed schema can often be reduced to extractive question answering, the schema generation problem, determining which dimensions to compare a set of documents, lacks a formal mapping to standard NLP tasks. In this work, we formulate schema generation as a reinforcement learning problem and investigate whether these dimensions can be induced without access to gold-standard schemas. We design a multi-faceted reward framework capturing schema coverage, non-redundancy, relevance, and format, and train a small language model on a literature review dataset. Our approach yields consistent improvements over the untuned base model across intrinsic, reference-based, and LLM-judge metrics, and remains competitive with supervised fine-tuned models at 5\times the parameter count on structural and diversity dimensions. All code, results and prompts are available in the GitHub repository: https://github.com/sinjoysaha/rl-schema-generation</abstract>
<identifier type="citekey">saha-etal-2026-reference</identifier>
<location>
<url>https://aclanthology.org/2026.acl-srw.111/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>1253</start>
<end>1261</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Reference-Free Schema Generation for Literature Review Tables via Multi-Faceted Rewards
%A Saha, Sinjoy
%A Saha, Suman
%A Farooque, Mahfuza
%A Yin, Wenpeng
%Y T.Y.S.S., Santosh
%Y Rodriguez, Juan Diego
%Y de Gibert, Ona
%S Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-393-7
%F saha-etal-2026-reference
%X To accelerate scientific knowledge acquisition, LLMs are increasingly used to synthesize multiple papers into structured tables by inferring schemas and values. While value generation within a fixed schema can often be reduced to extractive question answering, the schema generation problem, determining which dimensions to compare a set of documents, lacks a formal mapping to standard NLP tasks. In this work, we formulate schema generation as a reinforcement learning problem and investigate whether these dimensions can be induced without access to gold-standard schemas. We design a multi-faceted reward framework capturing schema coverage, non-redundancy, relevance, and format, and train a small language model on a literature review dataset. Our approach yields consistent improvements over the untuned base model across intrinsic, reference-based, and LLM-judge metrics, and remains competitive with supervised fine-tuned models at 5\times the parameter count on structural and diversity dimensions. All code, results and prompts are available in the GitHub repository: https://github.com/sinjoysaha/rl-schema-generation
%U https://aclanthology.org/2026.acl-srw.111/
%P 1253-1261
Markdown (Informal)
[Reference-Free Schema Generation for Literature Review Tables via Multi-Faceted Rewards](https://aclanthology.org/2026.acl-srw.111/) (Saha et al., ACL 2026)
ACL