NSTL国家科技图书文献中心

1. Classifying Potentially Non-compliant Portuguese Language Sentences Concerning Privacy Policies NSTL国家科技图书文献中心

Matheus Tocchini | Igor M. Rocha ... - 《Intelligent Systems,Part IV》 - Brazilian Conference on Intelligent Systems - 2025, - 107～121 - 共15页

摘要： compliance. We also introduced a corpus in Portuguese, with | Privacy policies (PPol) are extensive and | contain complex sentences that are difficult to | understand. They detail what happens with a person's | personal data and must comply with specific legislation

关键词： Machine learning | Privacy policy | Natural language processing | Data protection law | LGPD

2. GemBode and PhiBode: Adapting Small Language Models to Brazilian Portuguese NSTL国家科技图书文献中心

Gabriel Lino Garcia | Pedro Henrique Paiol...... - 《Progress in Pattern Recognition,Image Analysis,Computer Vision,and Applications,Part I》 - Iberoamerican Congress on Pattern Recognition - 2025, - 228～243 - 共16页

摘要： significantly to the field of Portuguese* language processing* | provided by large language* models have reshaped* | Language Models (SLMs). We have developed a unique | . This paper significantly contributes to language | language.

关键词： Small language models | Portuguese | Bode | Generative artificial intelligence | Natural language processing

3. Gender-Neutral English to Portuguese Machine Translator: Promoting Inclusive Language NSTL国家科技图书文献中心

Ricardo Trainotti Ra... | Evangelos Milios ... - 《Intelligent Systems,Part IV》 - Brazilian Conference on Intelligent Systems - 2025, - 180～195 - 共16页

摘要： bias in the Portuguese* language. However, it can also* | languages, like Portuguese. One of the most complex | as Portuguese* is the precise preservation of* | English to Portuguese* MT through post-processing* | dataset of English-Portuguese* sentences that was*

关键词： NLP | Machine translator | Fairness | Portuguese

4. BERTweet.BR: a pre-trained language model for tweets in Portuguese NSTL国家科技图书文献中心

Fernando,Carneiro | Daniela,Vianna ... - 《Neural computing & applications》 - 2025,37(6) - 4363～4385 - 共23页

摘要： monolingual general-domain Brazilian Portuguese language | Recent advancements in neural language* models* | the sixth most spoken language* globally, has* | users; however, no pre-trained language* model in* | . Besides the language, tweets-based pre-trained models

关键词： Social media | Short noisy texts | Monolingual language model | Sentiment analysis

5. Assessing European and Brazilian Portuguese LLMs for NER in Specialised Domains NSTL国家科技图书文献中心

Rafael Oleques Nunes | Joaquim Santos ... - 《Intelligent Systems,Part I》 - Brazilian Conference on Intelligent Systems - 2025, - 215～230 - 共16页

摘要： variants in Large Language* Models for the task of named* | previous works that did not consider the same language | This paper discusses the impact of Portuguese | tests were made on a Brazilian Portuguese* legal and a* | European Portuguese* historical corpora. The models taken*

关键词： Named entity recognition | Large language models | Portuguese language variants

6. Hate Speech Detection in Portuguese Using BERTimbau NSTL国家科技图书文献中心

Joao Otavio Rodrigue... | Gabriel Lino Garcia ... - 《Progress in Pattern Recognition,Image Analysis,Computer Vision,and Applications,Part I》 - Iberoamerican Congress on Pattern Recognition - 2025, - 244～255 - 共12页

摘要：Hate speech refers to language* expressions* | natural language* processing tasks, the efforts mainly* | focused on English, and languages like Portuguese* remain* | Portuguese pre-trained BERT model, BERTimbau. Experimental | that attack individuals or groups based on specific

关键词： Hate speech | Natural language processing | Machine learning | Undersampling | Portuguese language

7. Large Language Model for Querying Databases in Portuguese NSTL国家科技图书文献中心

Lourenco Figueiredo | Paulo Pinheiro ... - 《Progress in Artificial Intelligence,Part III》 - EPIA Conference on Artificial Intelligence - 2025, - 3～14 - 共12页

摘要： Portuguese Natural Language queries. | Language queries. Using a Large Language Model (LLM), SQL | provide accessibility through Portuguese* Natural* | This study introduces a system that helps non | -expert users find information easily without knowing

关键词： Natural language processing | Natural language to SQL | Large language model | GPT-4 turbo | Sports facility management | Databases

8. Impact of Quantization on Large Language Models for Portuguese Classification Tasks NSTL国家科技图书文献中心

Danilo Samuel Jodas | Gabriel Lino Garcia ... - 《Progress in Pattern Recognition,Image Analysis,Computer Vision,and Applications,Part I》 - Iberoamerican Congress on Pattern Recognition - 2025, - 213～227 - 共15页

摘要：Large Language* Models have emerged as* | capabilities in the context of natural language* processing* | in language* model architectures, their exponential* | methodologies on an open-source large language* model tailored* | surrounding the optimization of large language* models*

关键词： Quantization | Large language models | Natural language processing | Generative artificial intelligence | Bode

9. A Galician-Portuguese Generative Model NSTL国家科技图书文献中心

Pablo Gamallo | Pablo Rodriguez ... - 《Progress in Artificial Intelligence,Part III》 - EPIA Conference on Artificial Intelligence - 2025, - 292～304 - 共13页

摘要：Large language* models (LLMs) have* | revolutionized natural language* processing, but their* | , where English continues to be the predominant language | -pt-gl, focused precisely on these two language | Portuguese is marginal and that of the Galician variety is

关键词： Large language models | Generative models | Galician | Portuguese | Continual pretraining

10. Open Sentence Embeddings for Portuguese with the Serafim PT* Encoders Family NSTL国家科技图书文献中心

Luis Gomes | Antonio Branco ... - 《Progress in Artificial Intelligence,Part III》 - EPIA Conference on Artificial Intelligence - 2025, - 267～279 - 共13页

摘要： encoders for Portuguese* with various sizes, suited to* | Sentence encoder encode the semantics of their | input, enabling key downstream applications such as | classification, clustering, or retrieval. In this paper, we | present Serafim, a family of open-source sentence

摘要： compliance. We also introduced a corpus in Portuguese, with | Privacy policies (PPol) are extensive and | contain complex sentences that are difficult to | understand. They detail what happens with a person's | personal data and must comply with specific legislation

关键词： Machine learning | Privacy policy | Natural language processing | Data protection law | LGPD

摘要： significantly to the field of Portuguese* language processing* | provided by large language* models have reshaped* | Language Models (SLMs). We have developed a unique | . This paper significantly contributes to language | language.

关键词： Small language models | Portuguese | Bode | Generative artificial intelligence | Natural language processing

摘要： bias in the Portuguese* language. However, it can also* | languages, like Portuguese. One of the most complex | as Portuguese* is the precise preservation of* | English to Portuguese* MT through post-processing* | dataset of English-Portuguese* sentences that was*

关键词： NLP | Machine translator | Fairness | Portuguese

摘要： monolingual general-domain Brazilian Portuguese language | Recent advancements in neural language* models* | the sixth most spoken language* globally, has* | users; however, no pre-trained language* model in* | . Besides the language, tweets-based pre-trained models

关键词： Social media | Short noisy texts | Monolingual language model | Sentiment analysis

摘要： variants in Large Language* Models for the task of named* | previous works that did not consider the same language | This paper discusses the impact of Portuguese | tests were made on a Brazilian Portuguese* legal and a* | European Portuguese* historical corpora. The models taken*

关键词： Named entity recognition | Large language models | Portuguese language variants

摘要：Hate speech refers to language* expressions* | natural language* processing tasks, the efforts mainly* | focused on English, and languages like Portuguese* remain* | Portuguese pre-trained BERT model, BERTimbau. Experimental | that attack individuals or groups based on specific

关键词： Hate speech | Natural language processing | Machine learning | Undersampling | Portuguese language

摘要： Portuguese Natural Language queries. | Language queries. Using a Large Language Model (LLM), SQL | provide accessibility through Portuguese* Natural* | This study introduces a system that helps non | -expert users find information easily without knowing

关键词： Natural language processing | Natural language to SQL | Large language model | GPT-4 turbo | Sports facility management | Databases

摘要：Large Language* Models have emerged as* | capabilities in the context of natural language* processing* | in language* model architectures, their exponential* | methodologies on an open-source large language* model tailored* | surrounding the optimization of large language* models*

关键词： Quantization | Large language models | Natural language processing | Generative artificial intelligence | Bode

摘要：Large language* models (LLMs) have* | revolutionized natural language* processing, but their* | , where English continues to be the predominant language | -pt-gl, focused precisely on these two language | Portuguese is marginal and that of the Galician variety is

关键词： Large language models | Generative models | Galician | Portuguese | Continual pretraining

摘要： encoders for Portuguese* with various sizes, suited to* | Sentence encoder encode the semantics of their | input, enabling key downstream applications such as | classification, clustering, or retrieval. In this paper, we | present Serafim, a family of open-source sentence

关键词： Sentence encoder | Large language model | Portuguese

4008-161-200 800-990-8900

国家科技图书文献中心

摘要： compliance. We also introduced a corpus in Portuguese, with | Privacy policies (PPol) are extensive and | contain complex sentences that are difficult to | understand. They detail what happens with a person's | personal data and must comply with specific legislation

关键词： Machine learning | Privacy policy | Natural language processing | Data protection law | LGPD

摘要： significantly to the field of Portuguese language processing | provided by large language models have reshaped | Language Models (SLMs). We have developed a unique | . This paper significantly contributes to language | language.

关键词： Small language models | Portuguese | Bode | Generative artificial intelligence | Natural language processing

摘要： bias in the Portuguese language. However, it can also | languages, like Portuguese. One of the most complex | as Portuguese is the precise preservation of | English to Portuguese MT through post-processing | dataset of English-Portuguese sentences that was

关键词： NLP | Machine translator | Fairness | Portuguese

摘要： monolingual general-domain Brazilian Portuguese language | Recent advancements in neural language models | the sixth most spoken language globally, has | users; however, no pre-trained language model in | . Besides the language, tweets-based pre-trained models

关键词： Social media | Short noisy texts | Monolingual language model | Sentiment analysis

摘要： variants in Large Language Models for the task of named | previous works that did not consider the same language | This paper discusses the impact of Portuguese | tests were made on a Brazilian Portuguese legal and a | European Portuguese historical corpora. The models taken

关键词： Named entity recognition | Large language models | Portuguese language variants

摘要：Hate speech refers to language expressions | natural language processing tasks, the efforts mainly | focused on English, and languages like Portuguese remain | Portuguese pre-trained BERT model, BERTimbau. Experimental | that attack individuals or groups based on specific

关键词： Hate speech | Natural language processing | Machine learning | Undersampling | Portuguese language

摘要： Portuguese Natural Language queries. | Language queries. Using a Large Language Model (LLM), SQL | provide accessibility through Portuguese Natural | This study introduces a system that helps non | -expert users find information easily without knowing

关键词： Natural language processing | Natural language to SQL | Large language model | GPT-4 turbo | Sports facility management | Databases

摘要：Large Language Models have emerged as | capabilities in the context of natural language processing | in language model architectures, their exponential | methodologies on an open-source large language model tailored | surrounding the optimization of large language models

关键词： Quantization | Large language models | Natural language processing | Generative artificial intelligence | Bode

摘要：Large language models (LLMs) have | revolutionized natural language processing, but their | , where English continues to be the predominant language | -pt-gl, focused precisely on these two language | Portuguese is marginal and that of the Galician variety is

关键词： Large language models | Generative models | Galician | Portuguese | Continual pretraining

摘要： encoders for Portuguese with various sizes, suited to | Sentence encoder encode the semantics of their | input, enabling key downstream applications such as | classification, clustering, or retrieval. In this paper, we | present Serafim, a family of open-source sentence

关键词： Sentence encoder | Large language model | Portuguese

4008-161-200 800-990-8900

国家科技图书文献中心

摘要： significantly to the field of Portuguese* language processing* | provided by large language* models have reshaped* | Language Models (SLMs). We have developed a unique | . This paper significantly contributes to language | language.

摘要： bias in the Portuguese* language. However, it can also* | languages, like Portuguese. One of the most complex | as Portuguese* is the precise preservation of* | English to Portuguese* MT through post-processing* | dataset of English-Portuguese* sentences that was*

摘要： monolingual general-domain Brazilian Portuguese language | Recent advancements in neural language* models* | the sixth most spoken language* globally, has* | users; however, no pre-trained language* model in* | . Besides the language, tweets-based pre-trained models

摘要： variants in Large Language* Models for the task of named* | previous works that did not consider the same language | This paper discusses the impact of Portuguese | tests were made on a Brazilian Portuguese* legal and a* | European Portuguese* historical corpora. The models taken*

摘要：Hate speech refers to language* expressions* | natural language* processing tasks, the efforts mainly* | focused on English, and languages like Portuguese* remain* | Portuguese pre-trained BERT model, BERTimbau. Experimental | that attack individuals or groups based on specific

摘要： Portuguese Natural Language queries. | Language queries. Using a Large Language Model (LLM), SQL | provide accessibility through Portuguese* Natural* | This study introduces a system that helps non | -expert users find information easily without knowing

摘要：Large Language* Models have emerged as* | capabilities in the context of natural language* processing* | in language* model architectures, their exponential* | methodologies on an open-source large language* model tailored* | surrounding the optimization of large language* models*

摘要：Large language* models (LLMs) have* | revolutionized natural language* processing, but their* | , where English continues to be the predominant language | -pt-gl, focused precisely on these two language | Portuguese is marginal and that of the Galician variety is

摘要： encoders for Portuguese* with various sizes, suited to* | Sentence encoder encode the semantics of their | input, enabling key downstream applications such as | classification, clustering, or retrieval. In this paper, we | present Serafim, a family of open-source sentence