Analisando Fluxogramas Gerados por LLMs para Problemas Introdutórios de Programação
Resumo
Fluxogramas continuam sendo utilizados na aprendizagem de programação introdutória, auxiliando na construção visual de algoritmos e na compreensão do controle de fluxo. Com a popularização dos Grandes Modelos de Linguagem (LLMs) entre estudantes, torna-se relevante investigar suas capacidades de gerar fluxogramas. Este estudo avalia os modelos ChatGPT 5.5 e Gemini 3.5 em dez problemas cobrindo os principais conteúdos de programação introdutória. Os resultados mostram que, mesmo em tarefas básicas, os modelos ainda apresentam limitações na produção de fluxogramas corretos, reforçando o potencial dessa atividade como ferramenta de aprendizagem de programação com menor interferência da IA generativa.
Palavras-chave:
Fluxogramas, LLMs, Programação Introdutória
Referências
Darda, A. and Jain, R. (2024) "Code Generation from Flowchart using Optical Character Recognition & Large Language Model", Authorea Preprints.
Dingle, A. and Krulis, M. (2024) "Tackling Students' Coding Assignments with LLMs", Proceedings of the 1st International Workshop on Large Language Models for Code, Lisbon, Portugal, p. 94-101.
He, M., Zeng, J., Jiang, Y., Zhang, W., Liu, Z., Shi, X. and Zhou, A. (2025) "Flow2Code: Evaluating Large Language Models for Flowchart-based Code Generation Capability", Findings of the Association for Computational Linguistics: ACL 2025, p. 8124-8146.
Hong, A. and Hong, G. (2024) "The Effectiveness of Coding LLMs and the Challenges in Teaching CS1/2: A Case Study", Journal of Computing Sciences in Colleges, Vol. 40, No. 1, p. 122-131.
ISO (1985) "Information Processing - Documentation Symbols and Conventions for Data, Program and System Flowcharts, Program Network Charts and System Resources Charts", ISO 5807:1985, Geneva, Switzerland.
Manzano, J. A. N. G. (2004) "Revisão e Discussão da Norma ISO 5807-1985 (E) Proposta para Padronização Formal da Representação Gráfica da Linha de Raciocínio Lógico Utilizada no Desenvolvimento da Programação de Computadores a Ser Definida no Brasil", Revista Eletrônica Thesis, Vol. 1, p. 1-31.
Minu, M. S., Babu, A., Senthil, R. and Ahmed, T. A. S. S. (2025) "Prompt2Diagram: Transforming Natural Language into Visual Diagrams Using Advanced NLP and LLM", Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies (ICRDICCT 2025), v. 4, p. 161-169.
Muepu, D. M., Watanobe, Y., Ibn, A. M. F. and Shahajada, M. M. (2025) "Comparative Evaluation of ChatGPT, Gemini, and DeepSeek in Educational Problem Solving", IEEE 18th International Symposium on Embedded Multicore/Many-core Systems-on-Chip (MCSoC), p. 530-537.
Ouh, E. L., Gan, B. K. S., Jin Shim, K. and Wlodkowski, S. (2023) "ChatGPT, Can You Generate Solutions for my Coding Exercises? An Evaluation on its Effectiveness in an Undergraduate Java Programming Course", Proceedings of the 2023 Conference on Innovation and Technology in Computer Science Education, Turku, Finland, p. 54-60.
Rahe, C. and Maalej, W. (2025) "How Do Programming Students Use Generative AI?", Proceedings of the ACM on Software Engineering, Vol. 2, p. 978-1000.
SBC (2022) "Computação: Complemento à BNCC", Sociedade Brasileira de Computação, Brasil.
Shuvo, U. A., Dip, S. A., Vaskar, N. R. and Al Islam, A. B. M. A. (2025) "Assessing ChatGPT's Code Generation Capabilities with Short vs Long Context Programming Problems", Proceedings of the 11th International Conference on Networking, Systems, and Security, NSysS '24, Association for Computing Machinery, New York, NY, USA, p. 32-40.
Xue, Y., Chen, H., Bai, G. R., Tairas, R. and Huang, Y. (2024) "Does ChatGPT Help With Introductory Programming? An Experiment of Students Using ChatGPT in CS1", Proceedings of the 46th International Conference on Software Engineering: Software Engineering Education and Training, Lisbon, Portugal, p. 331-341.
Zheng, K. and Shi, Y. (2026) "CodeFlow: LLM-Generated Flowchart Feedback for Programming Students", Proceedings of the 57th ACM Technical Symposium on Computer Science Education V.2, p. 1609-1610.
Dingle, A. and Krulis, M. (2024) "Tackling Students' Coding Assignments with LLMs", Proceedings of the 1st International Workshop on Large Language Models for Code, Lisbon, Portugal, p. 94-101.
He, M., Zeng, J., Jiang, Y., Zhang, W., Liu, Z., Shi, X. and Zhou, A. (2025) "Flow2Code: Evaluating Large Language Models for Flowchart-based Code Generation Capability", Findings of the Association for Computational Linguistics: ACL 2025, p. 8124-8146.
Hong, A. and Hong, G. (2024) "The Effectiveness of Coding LLMs and the Challenges in Teaching CS1/2: A Case Study", Journal of Computing Sciences in Colleges, Vol. 40, No. 1, p. 122-131.
ISO (1985) "Information Processing - Documentation Symbols and Conventions for Data, Program and System Flowcharts, Program Network Charts and System Resources Charts", ISO 5807:1985, Geneva, Switzerland.
Manzano, J. A. N. G. (2004) "Revisão e Discussão da Norma ISO 5807-1985 (E) Proposta para Padronização Formal da Representação Gráfica da Linha de Raciocínio Lógico Utilizada no Desenvolvimento da Programação de Computadores a Ser Definida no Brasil", Revista Eletrônica Thesis, Vol. 1, p. 1-31.
Minu, M. S., Babu, A., Senthil, R. and Ahmed, T. A. S. S. (2025) "Prompt2Diagram: Transforming Natural Language into Visual Diagrams Using Advanced NLP and LLM", Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies (ICRDICCT 2025), v. 4, p. 161-169.
Muepu, D. M., Watanobe, Y., Ibn, A. M. F. and Shahajada, M. M. (2025) "Comparative Evaluation of ChatGPT, Gemini, and DeepSeek in Educational Problem Solving", IEEE 18th International Symposium on Embedded Multicore/Many-core Systems-on-Chip (MCSoC), p. 530-537.
Ouh, E. L., Gan, B. K. S., Jin Shim, K. and Wlodkowski, S. (2023) "ChatGPT, Can You Generate Solutions for my Coding Exercises? An Evaluation on its Effectiveness in an Undergraduate Java Programming Course", Proceedings of the 2023 Conference on Innovation and Technology in Computer Science Education, Turku, Finland, p. 54-60.
Rahe, C. and Maalej, W. (2025) "How Do Programming Students Use Generative AI?", Proceedings of the ACM on Software Engineering, Vol. 2, p. 978-1000.
SBC (2022) "Computação: Complemento à BNCC", Sociedade Brasileira de Computação, Brasil.
Shuvo, U. A., Dip, S. A., Vaskar, N. R. and Al Islam, A. B. M. A. (2025) "Assessing ChatGPT's Code Generation Capabilities with Short vs Long Context Programming Problems", Proceedings of the 11th International Conference on Networking, Systems, and Security, NSysS '24, Association for Computing Machinery, New York, NY, USA, p. 32-40.
Xue, Y., Chen, H., Bai, G. R., Tairas, R. and Huang, Y. (2024) "Does ChatGPT Help With Introductory Programming? An Experiment of Students Using ChatGPT in CS1", Proceedings of the 46th International Conference on Software Engineering: Software Engineering Education and Training, Lisbon, Portugal, p. 331-341.
Zheng, K. and Shi, Y. (2026) "CodeFlow: LLM-Generated Flowchart Feedback for Programming Students", Proceedings of the 57th ACM Technical Symposium on Computer Science Education V.2, p. 1609-1610.
Publicado
05/10/2026
Como Citar
ACIOLI, Maria H. L. S.; TEDESCO, Patricia C. de A. R.; COSTA, Evandro de B..
Analisando Fluxogramas Gerados por LLMs para Problemas Introdutórios de Programação. In: SIMPÓSIO BRASILEIRO DE INFORMÁTICA NA EDUCAÇÃO (SBIE), 37. , 2026, Goiânia/GO.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2026
.
p. 2133-2146.
DOI: https://doi.org/10.5753/sbie.2026.28227.
