DEVELOPMENT OF A RAG-BASED CHATBOT TO SUPPORT MULTIPROFESSIONAL SERVICES IN PUBLIC EDUCATION
DOI:
https://doi.org/10.66104/14p87804Keywords:
Educational chatbot; artificial intelligence; RAG; multidisciplinary teams; educational governance.Abstract
The increasing complexity of social demands within the school environment has intensified the need for more structured institutional responses, particularly regarding the work of multidisciplinary teams in public education. In this context, persistent challenges are observed, including fragmented information, limited access to regulatory frameworks, and a lack of standardization in service provision, which undermines the effectiveness of actions and the reliability of institutional decision-making. In response to this problem, this study aims to develop a chatbot based on the Retrieval-Augmented Generation (RAG) approach to support multidisciplinary assistance through the systematization and retrieval of institutional protocols and normative documents. Methodologically, the research adopts the Design Science Research (DSR) approach, focused on the construction of a technological artifact capable of addressing real-world problems while integrating scientific rigor and practical relevance. The development process involved the organization of documentary databases, their structuring into computationally processable formats, and integration with language models, enabling semantic information retrieval and the generation of contextualized responses. The chatbot was designed as an informational mediator, capable of guiding professionals based on official guidelines, reducing reliance on individual interpretations and promoting greater consistency in procedures. The findings indicate that the use of RAG-based technologies contributes to the standardization of service workflows, enhances educational governance, and strengthens multidisciplinary practices by improving qualified access to information and supporting more reliable decision-making processes aligned with current regulations.
Downloads
References
ABEYSINGHE, Bhashithe; CIRCI, Ruhan. The Challenges of Evaluating LLM Applications: An Analysis of Automated, Human, and LLM-Based Approaches. arXiv:2406.03339v2, 2024.
ADAMSON, David; DYKE, Gregory; JANG, Hyeju; ROSÉ, Carolyn Penstein. Towards an agile approach to adapting dynamic collaboration support to student needs. International Journal of Artificial Intelligence in Education, [s. l.], v. 24, n. 1, p. 92-124, 2014. DOI: https://doi.org/10.1007/s40593-013-0012-6. DOI: https://doi.org/10.1007/s40593-013-0012-6
ATEIA, Samy; KRUSCHWITZ, Udo. Can Open-Source LLMs Compete with Commercial Models? Exploring the Few-Shot Performance of Current GPT Models in Biomedical Tasks. arXiv:2407.13511, 2024.
BENDER, Emily M.; GEBRU, Timnit; MCMILLAN-MAJOR, Angelina; SHMITCHELL, Shmargaret. On the dangers of stochastic parrots: can language models be too big? In: ACM CONFERENCE ON FAIRNESS, ACCOUNTABILITY, AND TRANSPARENCY, 2021. Proceedings [...]. New York: Association for Computing Machinery, 2021. p. 610-623. DOI: https://doi.org/10.1145/3442188.3445922. DOI: https://doi.org/10.1145/3442188.3445922
BRASIL. Lei nº 8.069, de 13 de julho de 1990. Dispõe sobre o Estatuto da Criança e do Adolescente e dá outras providências. Brasília, DF: Presidência da República, 1990. Disponível em: https://www.planalto.gov.br/ccivil_03/leis/l8069.htm. Acesso em: 23 abr. 2026.
BROWN, Tom B. et al. Language models are few-shot learners. Advances in Neural Information Processing Systems, [s. l.], v. 33, p. 1877-1901, 2020.
DEMING, William Edwards. Out of the crisis. Cambridge: MIT Center for Advanced Engineering Study, 1986.
GIL, Antonio Carlos. Métodos e técnicas de pesquisa social. 6. ed. São Paulo: Atlas, 2008.
HEVNER, Alan R.; MARCH, Salvatore T.; PARK, Jinsoo; RAM, Sudha. Design science in information systems research. MIS Quarterly, Minneapolis, v. 28, n. 1, p. 75-105, 2004. DOI: https://doi.org/10.2307/25148625. DOI: https://doi.org/10.2307/25148625
HOLMES, Wayne. Artificial intelligence in education. In: MIAO, Fengchun et al. (ed.). Artificial intelligence in education: challenges and opportunities for sustainable development. Paris: UNESCO, 2019.
JURAFSKY, Daniel; MARTIN, James H. Speech and language processing: an introduction to natural language processing, computational linguistics, and speech recognition. 3rd ed. draft. [S. l.]: [s. n.], 2020. Disponível em: https://web.stanford.edu/~jurafsky/slp3/. Acesso em: 23 abr. 2026.
LEWIS, Patrick et al. Retrieval-augmented generation for knowledge-intensive NLP tasks. Advances in Neural Information Processing Systems, [s. l.], v. 33, p. 9459-9474, 2020.
LIPSKY, Michael. Street-level bureaucracy: dilemmas of the individual in public services. New York: Russell Sage Foundation, 1980. DOI: https://doi.org/10.2307/1288305
MORAES, Silvia Maria Wanderley; DE SOUZA, Renato Rocha. Uma abordagem semiautomática para expansão e enriquecimento linguístico de bases de conhecimento de chatbots educacionais. In: CONFERÊNCIA INTERNACIONAL SOBRE INFORMÁTICA NA EDUCAÇÃO, 2015. Anais [...]. [S. l.]: TISE, 2015. p. 600-605.
PRESSMAN, Jeffrey L.; WILDAVSKY, Aaron. Implementation: how great expectations in Washington are dashed in Oakland. 3rd ed. Berkeley: University of California Press, 1984.
RUSSELL, Stuart J.; NORVIG, Peter. Artificial intelligence: a modern approach. 4th ed. Hoboken: Pearson, 2021.
SELWYN, Neil. Should robots replace teachers? AI and the future of education. Cambridge: Polity Press, 2019.
TOCANTINS. Secretaria de Estado da Educação. Manual PAFS: Plano de Acompanhamento Familiar e Socioemocional. Palmas: SEDUC-TO, [s. d.].
TOCANTINS. Secretaria de Estado da Educação. Programa Evasão Escolar: Nota Zero — PEENZ. Palmas: SEDUC-TO, [s. d.].
TOCANTINS. Secretaria de Estado da Educação. Proposta de trabalho da Unidade Técnica Executiva de Orientação Educacional/Núcleo Gestor Multiprofissional. Palmas: SEDUC-TO, [s. d.].
TOCANTINS. Secretaria de Estado da Educação. Protocolo de atendimento em situações específicas. Palmas: SEDUC-TO, [s. d.].
TOCANTINS. Secretaria de Estado da Educação. Regimento Escolar da Rede Estadual de Ensino 2026. Palmas: SEDUC-TO, [s. d.].
TOCANTINS. Secretaria de Estado da Educação. Resolução nº 105, de 22 de setembro de 2006. Dispõe sobre a normatização das equipes multiprofissionais na rede estadual de ensino. Palmas: SEDUC-TO, 2006.
TOCANTINS. Secretaria de Estado da Educação. PROPOSTA_POP_SEDUC_TO_2026: Procedimento Operacional Padrão proposto para a Secretaria de Estado da Educação do Tocantins. Palmas: SEDUC-TO, 2026.
WEIZENBAUM, Joseph. ELIZA: a computer program for the study of natural language communication between man and machine. Communications of the ACM, New York, v. 9, n. 1, p. 36-45, 1966. DOI: https://doi.org/10.1145/365153.365168. DOI: https://doi.org/10.1145/365153.365168
YAN, Lixiang et al. Practical and Ethical Challenges of Large Language Models in Education: A Systematic Scoping Review. arXiv:2303.13379, 2023. DOI: https://doi.org/10.1111/bjet.13370
YANG, Yuming et al. Benchmarking Multimodal RAG through a Chart-based Document Question-Answering Generation Framework. arXiv:2502.14864, 2025.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Bruno Marques de Lima, Fábio Emanuell Abreu Cardoso, Dr. Ramásio Ferreira de Melo, Dr. Leonardo de Andrade Carneiro
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish in this journal agree to the following terms:
Authors retain copyright and grant the journal the right of first publication, with the work simultaneously licensed under the Creative Commons Attribution License, which permits the sharing of the work with proper acknowledgment of authorship and initial publication in this journal;
Authors are authorized to enter into separate, additional agreements for the non-exclusive distribution of the version of the work published in this journal (e.g., posting in an institutional repository or publishing it as a book chapter), provided that authorship and initial publication in this journal are properly acknowledged, and that the work is adapted to the template of the respective repository;
Authors are permitted and encouraged to post and distribute their work online (e.g., in institutional repositories or on their personal websites) at any point before or during the editorial process, as this may lead to productive exchanges and increase the impact and citation of the published work (see The Effect of Open Access);
Authors are responsible for correctly providing their personal information, including name, keywords, abstracts, and other relevant data, thereby defining how they wish to be cited. The journal’s editorial board is not responsible for any errors or inconsistencies in these records.
PRIVACY POLICY
The names and email addresses provided to this journal will be used exclusively for the purposes of this publication and will not be made available for any other purpose or to third parties.
Note: All content of the work is the sole responsibility of the author and the advisor.
