Daniela Barreiro Claro
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Daniela Barreiro Claro has a total of 109 co-authors. Here are the top co-authors by number of shared publications:
- Babacar Mane
UFBA
6 shared publications Cleiton Fernando Lima Sena
Federal University of Bahia
5 shared publications- Marlo Souza
UFBA
5 shared publications - Rita Suzana Pitangueira Maciel
Federal University of Bahia, Brazil
5 shared publications Denivaldo Lopes
4 shared publicationsMarlo Souza
FORMAS Research Group, Federal University of Bahia (UFBA)–Institute of Computing (IC), Salvador, Brazil
4 shared publications
Latest Publications
- 1 October 2026
- Published by Springer Nature
- p. 595-610
Grammatical Representations in Transformer Attention: A Multidomain Study of Portuguese via UD
- 1 October 2026
- Published by Springer Nature
- p. 595-610
Grammatical Representations in Transformer Attention: A Multidomain Study of Portuguese via UD
- 25 May 2026
- Published by Sociedade Brasileira de Computação
- p. 344-351
Do Dado à Ação: Modelando Pragmatismo no contexto da Internet das Coisas
Este artigo propõe uma modelagem matemática para a interoperabilidade pragmática na Internet das Coisas (IdC). Utilizando Aprendizado por Reforço (Q-Learning), o modelo infere ações orientadas a objetivos sob incerteza de contexto. Em simulações de chão de fábrica, o algoritmo superou a operação humana na estabilidade do sistema e na otimização de recursos. Argumentamos sobre a viabilidade da interoperabilidade pragmática entre máquina e humanos, evidenciando que uma modelagem formal bem estruturada pode ser o caminho para avanços neste nível de comunicação.
- 25 May 2026
- Published by Sociedade Brasileira de Computação
- p. 344-351
Do Dado à Ação: Modelando Pragmatismo no contexto da Internet das Coisas
Este artigo propõe uma modelagem matemática para a interoperabilidade pragmática na Internet das Coisas (IdC). Utilizando Aprendizado por Reforço (Q-Learning), o modelo infere ações orientadas a objetivos sob incerteza de contexto. Em simulações de chão de fábrica, o algoritmo superou a operação humana na estabilidade do sistema e na otimização de recursos. Argumentamos sobre a viabilidade da interoperabilidade pragmática entre máquina e humanos, evidenciando que uma modelagem formal bem estruturada pode ser o caminho para avanços neste nível de comunicação.
- Open Access
- 1 December 2025
- Published by FapUNIFESP (SciELO) in SciELO Preprints
Os caminhos do Projeto ALiB na intersecção entre a Linguística e a Computação: Vitalidade, Geolocalização e Mudança Semântico-Lexical
O Projeto Atlas Linguístico do Brasil (Projeto ALiB) inaugurou novas direções para a Dialetologia brasileira ao adotar uma metodologia pluridimensional e ao estabelecer uma colaboração inédita entre linguistas e cientistas da Computação. Essa parceria resultou na concepção do ALiBWeb, um sistema de terceira geração voltado para a disponibilização integrada de áudios, cartas linguísticas e informações georreferenciadas. A criação dessa infraestrutura tecnológica fortaleceu a interface entre Linguística e Computação, impulsionando a formação de novos pesquisadores e a produção de novos projetos. Apesar desses avanços, muitos dos desenvolvimentos computacionais associados ao Projeto ALiB permanecem pouco conhecidos pela comunidade linguística, evidenciando a necessidade de maior divulgação das contribuições provenientes dessa intersecção interdisciplinar. Assim, o presente artigo visa apresentar um conjunto de trabalhos desenvolvidos sob a ótica da Computação, sendo analisado sob a perspectiva do Projeto ALiB, descrevendo os caminhos percorridos e os casos de interface entre a Linguística e a Computação. Um total de três trabalhos que versam sobre vitalidade, geolocalização e mudança semântico lexical foi analisado e os seus resultados foram apresentados.
- 1 December 2025
- Published by FapUNIFESP (SciELO) in SciELO Preprints
Open AccessOs caminhos do Projeto ALiB na intersecção entre a Linguística e a Computação: Vitalidade, Geolocalização e Mudança Semântico-Lexical
O Projeto Atlas Linguístico do Brasil (Projeto ALiB) inaugurou novas direções para a Dialetologia brasileira ao adotar uma metodologia pluridimensional e ao estabelecer uma colaboração inédita entre linguistas e cientistas da Computação. Essa parceria resultou na concepção do ALiBWeb, um sistema de terceira geração voltado para a disponibilização integrada de áudios, cartas linguísticas e informações georreferenciadas. A criação dessa infraestrutura tecnológica fortaleceu a interface entre Linguística e Computação, impulsionando a formação de novos pesquisadores e a produção de novos projetos. Apesar desses avanços, muitos dos desenvolvimentos computacionais associados ao Projeto ALiB permanecem pouco conhecidos pela comunidade linguística, evidenciando a necessidade de maior divulgação das contribuições provenientes dessa intersecção interdisciplinar. Assim, o presente artigo visa apresentar um conjunto de trabalhos desenvolvidos sob a ótica da Computação, sendo analisado sob a perspectiva do Projeto ALiB, descrevendo os caminhos percorridos e os casos de interface entre a Linguística e a Computação. Um total de três trabalhos que versam sobre vitalidade, geolocalização e mudança semântico lexical foi analisado e os seus resultados foram apresentados.
- Open Access
- 29 September 2025
- Published by Sociedade Brasileira de Computação
- p. 295-306
Syntactic Analysis in Transformers through Attention Heads
The advances in Natural Language Processing (NLP) have led to the development of Transformer architectures, such as BERT. One of the most prominent features of this architecture is the attention mechanism. However, the application of attention mechanisms to languages other than English, like Brazilian Portuguese, remains underexplored. This work analyzes the attention heads in a Transformer architecture, considering the syntactic relations in a sentence. We analyze how attention patterns align with syntactic dependencies involving phenomena such as transitive verbs, reflexive pronouns, and subordinate clauses, as realized in Brazilian Portuguese. Results described the existence of specialized heads within arcs, such as subject–verb and verb–object. These findings open up new research opportunities to evaluate syntactic sensitivity in Transformer models and to contribute to the development of more linguistically informed models for Brazilian Portuguese.
- 29 September 2025
- Published by Sociedade Brasileira de Computação
- p. 295-306
Open AccessSyntactic Analysis in Transformers through Attention Heads
The advances in Natural Language Processing (NLP) have led to the development of Transformer architectures, such as BERT. One of the most prominent features of this architecture is the attention mechanism. However, the application of attention mechanisms to languages other than English, like Brazilian Portuguese, remains underexplored. This work analyzes the attention heads in a Transformer architecture, considering the syntactic relations in a sentence. We analyze how attention patterns align with syntactic dependencies involving phenomena such as transitive verbs, reflexive pronouns, and subordinate clauses, as realized in Brazilian Portuguese. Results described the existence of specialized heads within arcs, such as subject–verb and verb–object. These findings open up new research opportunities to evaluate syntactic sensitivity in Transformer models and to contribute to the development of more linguistically informed models for Brazilian Portuguese.
- Open Access
- 19 May 2025
- Published by Sociedade Brasileira de Computação
- p. 280-289
Organizational Interoperability: A Systematic Mapping Study
Context: Organizational interoperability enables collaboration amomg organizations with different structures and processes, promoting efficiency and integration in sectors such as public administration, health, and education. Problem: Implementing organizational interoperability presents significant challenges, including the lack of commonly accepted processes, difficulties interpreting administrative procedures and legislation, and the absence of conceptual models to define inter-organizational relationships. There are few studies that report organizational interoperability, and there is no consensus on several concepts and structures. Solution: The study maps existing approaches, including models for assessing organizational interoperability maturity, and a conceptual model based on Systems of Information Systems (SoIS) to help create interoperability connections between organizations. Information systems theory: Organizational interoperability research is associated with organizational information processing theory. Method: A systematic literature mapping was conducted in databases such as IEEE Xplore, ACM Digital Library, ScienceDirect, and Scopus. Summary of Results: Ten main studies were identified. In general, organizational interoperability has been treated as the ability of organizations to align business processes and expectations for effective information exchange. Challenges include the need for standardized processes, difficulties with regulations, and the absence of conceptual models. Solutions involve process-oriented approaches, but adoption still needs to be improved. Contributions and Impact in the area of information systems: The study contributes by providing a comprehensive view of organizational interoperability, identifying gaps, challenges, and possible solutions. It reinforces the importance of interoperability for inter-organizational collaboration to promote a more efficient and collaborative environment.
- 19 May 2025
- Published by Sociedade Brasileira de Computação
- p. 280-289
Open AccessOrganizational Interoperability: A Systematic Mapping Study
Context: Organizational interoperability enables collaboration amomg organizations with different structures and processes, promoting efficiency and integration in sectors such as public administration, health, and education. Problem: Implementing organizational interoperability presents significant challenges, including the lack of commonly accepted processes, difficulties interpreting administrative procedures and legislation, and the absence of conceptual models to define inter-organizational relationships. There are few studies that report organizational interoperability, and there is no consensus on several concepts and structures. Solution: The study maps existing approaches, including models for assessing organizational interoperability maturity, and a conceptual model based on Systems of Information Systems (SoIS) to help create interoperability connections between organizations. Information systems theory: Organizational interoperability research is associated with organizational information processing theory. Method: A systematic literature mapping was conducted in databases such as IEEE Xplore, ACM Digital Library, ScienceDirect, and Scopus. Summary of Results: Ten main studies were identified. In general, organizational interoperability has been treated as the ability of organizations to align business processes and expectations for effective information exchange. Challenges include the need for standardized processes, difficulties with regulations, and the absence of conceptual models. Solutions involve process-oriented approaches, but adoption still needs to be improved. Contributions and Impact in the area of information systems: The study contributes by providing a comprehensive view of organizational interoperability, identifying gaps, challenges, and possible solutions. It reinforces the importance of interoperability for inter-organizational collaboration to promote a more efficient and collaborative environment.
- 30 January 2025
- Published by Springer Nature
- p. 427-441
Scaling and Adapting Large Language Models for Portuguese Open Information Extraction: A Comparative Study of Fine-Tuning and LoRA
- 30 January 2025
- Published by Springer Nature
- p. 427-441
Scaling and Adapting Large Language Models for Portuguese Open Information Extraction: A Comparative Study of Fine-Tuning and LoRA
- Open Access
- 17 January 2025
- Published by Sociedade Brasileira de Computação in Journal of the Brazilian Computer Society
- Vol. 31 (1)
A bilingual analysis of multi-head attention mechanism for image captioning based on morphosyntactic information
Image Captioning is the task of describing the information conveyed by an image, i.e., its visual content in a natural language. Most of the current researches make use of the encoder-decoder architecture to create relations between images (the inputs) and text (the output). These relations are originated from the attention mechanisms, present on the Transformer model, and can be leveraged to understand how the image-text relationship is encoded during training and inference times. This work investigates the hypothesis that the attention mechanism behaves analogously for words that share morphosyntactic labels within texts. To this matter, the attention weights for each predicted word --- posed as the "focus" given in the image at each step --- are gathered, averaged and inspected; also, the analysis are performed taking into account one model trained with English captions and another trained with Portuguese captions, therefore comparing two languages with different morphological organization. Our results show that words with the same functioning in the sentence,e.g., being prone to similar inflections, usually have the same focal point in the image. Our work sheds light to the importance of linguistic studies for the vision-language area, reinforcing the benefits of including language-aware knowledge during training.
- 17 January 2025
- Published by Sociedade Brasileira de Computação in Journal of the Brazilian Computer Society
- Vol. 31 (1)
Open AccessA bilingual analysis of multi-head attention mechanism for image captioning based on morphosyntactic information
Image Captioning is the task of describing the information conveyed by an image, i.e., its visual content in a natural language. Most of the current researches make use of the encoder-decoder architecture to create relations between images (the inputs) and text (the output). These relations are originated from the attention mechanisms, present on the Transformer model, and can be leveraged to understand how the image-text relationship is encoded during training and inference times. This work investigates the hypothesis that the attention mechanism behaves analogously for words that share morphosyntactic labels within texts. To this matter, the attention weights for each predicted word --- posed as the "focus" given in the image at each step --- are gathered, averaged and inspected; also, the analysis are performed taking into account one model trained with English captions and another trained with Portuguese captions, therefore comparing two languages with different morphological organization. Our results show that words with the same functioning in the sentence,e.g., being prone to similar inflections, usually have the same focal point in the image. Our work sheds light to the importance of linguistic studies for the vision-language area, reinforcing the benefits of including language-aware knowledge during training.
- Open Access
- 28 November 2024
- Published by Sociedade Brasileira de Computação
- p. 493-500
Brazilian Consumer Protection Code: a methodology for a dataset to Question-Answer (QA) Models
This work introduces the methodology for building a new dataset based onthe Brazilian Consumer Protection Code (CDC), focusing on question-answer (QA) models. The dataset collected legal data, including CDC articles, legal summaries, and court rulings from the Superior Court of Justice (STJ). Automated data extraction techniques using Python were employed, and advanced language models such as Llama3-8b-8192, Gemma2-9b-it, and GPT-4o-mini were used to generate question-answer (QA) structures. This work presents our methodology for creating such a dataset to be used by language models for training in the legal domain, particularly in the CDC domain.
- 28 November 2024
- Published by Sociedade Brasileira de Computação
- p. 493-500
Open AccessBrazilian Consumer Protection Code: a methodology for a dataset to Question-Answer (QA) Models
This work introduces the methodology for building a new dataset based onthe Brazilian Consumer Protection Code (CDC), focusing on question-answer (QA) models. The dataset collected legal data, including CDC articles, legal summaries, and court rulings from the Superior Court of Justice (STJ). Automated data extraction techniques using Python were employed, and advanced language models such as Llama3-8b-8192, Gemma2-9b-it, and GPT-4o-mini were used to generate question-answer (QA) structures. This work presents our methodology for creating such a dataset to be used by language models for training in the legal domain, particularly in the CDC domain.
- Open Access
- 18 July 2024
- Published by Fundacao Getulio Vargas in Revista Brasileira de Casos de Ensino em Administração
- Vol. 14 (especial1) , a7
Hackathon on-line na aprendizagem baseada em projetos: um relato de experiência no desenvolvimento de protótipo de inteligência artificial
Objetivo: Apresentar os resultados da aplicação do modelo Hackathon online associado a um método de aprendizagem baseada em projeto em um componente curricular de extensão universitária. Escopo: O relato identifica e discute as estratégias necessárias para garantir a colaboração interdisciplinar, o engajamento discente e a aplicação prática de saberes transversais em um ambiente virtual. Originalidade: O caso aborda a conjunção de metodologias ativas no campo de ensino, e extensão integrando campos multidisciplinares no desenvolvimento de habilidades e competências interpessoais. Relevância: A necessidade de repensar os métodos de ensino, pesquisa e extensão, destacando a importância da incorporação de abordagens inovadoras, como o modelo Hackathon online e aprendizagem baseada em projetos, para preparar os estudantes para os desafios e oportunidades da era digital.
- 18 July 2024
- Published by Fundacao Getulio Vargas in Revista Brasileira de Casos de Ensino em Administração
- Vol. 14 (especial1) , a7
Open AccessHackathon on-line na aprendizagem baseada em projetos: um relato de experiência no desenvolvimento de protótipo de inteligência artificial
Objetivo: Apresentar os resultados da aplicação do modelo Hackathon online associado a um método de aprendizagem baseada em projeto em um componente curricular de extensão universitária. Escopo: O relato identifica e discute as estratégias necessárias para garantir a colaboração interdisciplinar, o engajamento discente e a aplicação prática de saberes transversais em um ambiente virtual. Originalidade: O caso aborda a conjunção de metodologias ativas no campo de ensino, e extensão integrando campos multidisciplinares no desenvolvimento de habilidades e competências interpessoais. Relevância: A necessidade de repensar os métodos de ensino, pesquisa e extensão, destacando a importância da incorporação de abordagens inovadoras, como o modelo Hackathon online e aprendizagem baseada em projetos, para preparar os estudantes para os desafios e oportunidades da era digital.
- Open Access
- 20 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-11
A Temporal Network Visualization and Data Analysis on Two Decades of SBSI
Context: The Brazilian Information Systems (BIS) community has gained importance in national and international scenarios. Understanding its evolution aids in defining contributions to future research. Problem: Gaining knowledge of large communities is not trivial, involving collecting adequate data, analyzing networks’ temporal behavior, and exploring different techniques of textual, graphical, and classification of structural evolution. Solution: Using two decades of SBSI publications provides an overview of collaboration networks’ temporal behavior to understand community evolution. IS Theory: This work is under social network theory to understand relationships between networks, links, or connections to identify interaction processes among authors and networks. Method: We conducted an exploratory study with a quantitative and qualitative approach to identify perspectives of the SBSI community phenomena as patterns, trends, and authors’ relationships in the networks. Summary of Results: The findings with 20 years include a network with 3,511 nodes, 3,610 relationships, and 771 detected communities, with an average Louvain modularity of 0.9559 indicating a solid community structure within the network but few connections where only 12 communities publish frequently over the two decades. Also, dividing 20 years by four timeslices, there are 104 community births, 20 grows, 31 contractions, and 331 deaths. Contributions and Impact in IS area: This work provides temporal BIS community behavior analysis using novel visualization techniques with structural, temporal, and evolution taxonomies to gain knowledge for decision-making for sustainable community growth.
- 20 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Open AccessA Temporal Network Visualization and Data Analysis on Two Decades of SBSI
Context: The Brazilian Information Systems (BIS) community has gained importance in national and international scenarios. Understanding its evolution aids in defining contributions to future research. Problem: Gaining knowledge of large communities is not trivial, involving collecting adequate data, analyzing networks’ temporal behavior, and exploring different techniques of textual, graphical, and classification of structural evolution. Solution: Using two decades of SBSI publications provides an overview of collaboration networks’ temporal behavior to understand community evolution. IS Theory: This work is under social network theory to understand relationships between networks, links, or connections to identify interaction processes among authors and networks. Method: We conducted an exploratory study with a quantitative and qualitative approach to identify perspectives of the SBSI community phenomena as patterns, trends, and authors’ relationships in the networks. Summary of Results: The findings with 20 years include a network with 3,511 nodes, 3,610 relationships, and 771 detected communities, with an average Louvain modularity of 0.9559 indicating a solid community structure within the network but few connections where only 12 communities publish frequently over the two decades. Also, dividing 20 years by four timeslices, there are 104 community births, 20 grows, 31 contractions, and 331 deaths. Contributions and Impact in IS area: This work provides temporal BIS community behavior analysis using novel visualization techniques with structural, temporal, and evolution taxonomies to gain knowledge for decision-making for sustainable community growth.