@inproceedings{pinto-etal-2026-evaluating,
title = "Evaluating Generative Large Language Models for {P}ortuguese Scientific Information Extraction",
author = "Pinto, Tom{\'a}s and
Silva, Catarina and
Goncalo Oliveira, Hugo",
editor = "Rehm, Georg and
Dietze, Stefan and
Dessi, Danilo and
Maynard, Diana and
Schimmler, Sonja",
booktitle = "Proceedings of Natural Scientific Language Processing ({NSLP}) @ {LREC} 2026",
month = may,
year = "2026",
address = "Palma, Mallorca (Spain)",
publisher = "ELRA Language Resources Association (ELRA)",
url = "https://aclanthology.org/2026.nslp-1.15/",
doi = "10.63317/5inv25yddbgv",
pages = "155--167",
abstract = "Scientific Information Extraction (IE), which identifies entities and their relations from scientific texts, is essential for building Scientific Knowledge Graphs (SciKGs) that encode structured knowledge and enable applications such as semantic search, question answering, and literature reasoning. Large Language Models (LLMs) have shown strong capabilities in processing unstructured text, yet most advances focus on English, with limited exploration for less-resourced languages like Portuguese. The reliability of generative LLMs, including Portuguese-targeted models like the sovereign AMALIA, for structured extraction of scientific knowledge from literature text remains underexplored. We evaluate low- to mid-scale generative LLMs (8{--}12B parameters) on scientific Named Entity Recognition (NER) and Relation Extraction (RE), using a Portuguese-translated dataset of computer science article abstracts. Overall, our results show moderate performance and indicate that the adaptation strategy has a greater impact than model choice: prompting yields unstable performance and poor RE scores, while fine-tuning consistently improves both NER and RE and reduces cross-model variability. These findings suggest that, at this scale, prompting alone is insufficient for SciKG construction and underscore the need for supervised adaptation. We provide a detailed error analysis and outline directions for advancing Portuguese scientific IE."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="pinto-etal-2026-evaluating">
<titleInfo>
<title>Evaluating Generative Large Language Models for Portuguese Scientific Information Extraction</title>
</titleInfo>
<name type="personal">
<namePart type="given">Tomás</namePart>
<namePart type="family">Pinto</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Catarina</namePart>
<namePart type="family">Silva</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hugo</namePart>
<namePart type="family">Goncalo Oliveira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of Natural Scientific Language Processing (NSLP) @ LREC 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Georg</namePart>
<namePart type="family">Rehm</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Stefan</namePart>
<namePart type="family">Dietze</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Danilo</namePart>
<namePart type="family">Dessi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Diana</namePart>
<namePart type="family">Maynard</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sonja</namePart>
<namePart type="family">Schimmler</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA Language Resources Association (ELRA)</publisher>
<place>
<placeTerm type="text">Palma, Mallorca (Spain)</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Scientific Information Extraction (IE), which identifies entities and their relations from scientific texts, is essential for building Scientific Knowledge Graphs (SciKGs) that encode structured knowledge and enable applications such as semantic search, question answering, and literature reasoning. Large Language Models (LLMs) have shown strong capabilities in processing unstructured text, yet most advances focus on English, with limited exploration for less-resourced languages like Portuguese. The reliability of generative LLMs, including Portuguese-targeted models like the sovereign AMALIA, for structured extraction of scientific knowledge from literature text remains underexplored. We evaluate low- to mid-scale generative LLMs (8–12B parameters) on scientific Named Entity Recognition (NER) and Relation Extraction (RE), using a Portuguese-translated dataset of computer science article abstracts. Overall, our results show moderate performance and indicate that the adaptation strategy has a greater impact than model choice: prompting yields unstable performance and poor RE scores, while fine-tuning consistently improves both NER and RE and reduces cross-model variability. These findings suggest that, at this scale, prompting alone is insufficient for SciKG construction and underscore the need for supervised adaptation. We provide a detailed error analysis and outline directions for advancing Portuguese scientific IE.</abstract>
<identifier type="citekey">pinto-etal-2026-evaluating</identifier>
<identifier type="doi">10.63317/5inv25yddbgv</identifier>
<location>
<url>https://aclanthology.org/2026.nslp-1.15/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>155</start>
<end>167</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Evaluating Generative Large Language Models for Portuguese Scientific Information Extraction
%A Pinto, Tomás
%A Silva, Catarina
%A Goncalo Oliveira, Hugo
%Y Rehm, Georg
%Y Dietze, Stefan
%Y Dessi, Danilo
%Y Maynard, Diana
%Y Schimmler, Sonja
%S Proceedings of Natural Scientific Language Processing (NSLP) @ LREC 2026
%D 2026
%8 May
%I ELRA Language Resources Association (ELRA)
%C Palma, Mallorca (Spain)
%F pinto-etal-2026-evaluating
%X Scientific Information Extraction (IE), which identifies entities and their relations from scientific texts, is essential for building Scientific Knowledge Graphs (SciKGs) that encode structured knowledge and enable applications such as semantic search, question answering, and literature reasoning. Large Language Models (LLMs) have shown strong capabilities in processing unstructured text, yet most advances focus on English, with limited exploration for less-resourced languages like Portuguese. The reliability of generative LLMs, including Portuguese-targeted models like the sovereign AMALIA, for structured extraction of scientific knowledge from literature text remains underexplored. We evaluate low- to mid-scale generative LLMs (8–12B parameters) on scientific Named Entity Recognition (NER) and Relation Extraction (RE), using a Portuguese-translated dataset of computer science article abstracts. Overall, our results show moderate performance and indicate that the adaptation strategy has a greater impact than model choice: prompting yields unstable performance and poor RE scores, while fine-tuning consistently improves both NER and RE and reduces cross-model variability. These findings suggest that, at this scale, prompting alone is insufficient for SciKG construction and underscore the need for supervised adaptation. We provide a detailed error analysis and outline directions for advancing Portuguese scientific IE.
%R 10.63317/5inv25yddbgv
%U https://aclanthology.org/2026.nslp-1.15/
%U https://doi.org/10.63317/5inv25yddbgv
%P 155-167
Markdown (Informal)
[Evaluating Generative Large Language Models for Portuguese Scientific Information Extraction](https://aclanthology.org/2026.nslp-1.15/) (Pinto et al., NSLP 2026)
ACL