Engenharia de Prompts em Assistentes Conversacionais para Promoção de Autocuidado baseados em Modelos Amplos de Linguagem
Resumo
A implementação de assistentes conversacionais com Modelos Amplos de Linguagem traz desafios, incluindo assegurar interações seguras e informativas. Este artigo explora o uso da engenharia de prompts na criação da MarIA, uma assistente virtual baseada no GPT-3.5 para pacientes com diabetes, visando promover o autocuidado. MarIA foi desenvolvida em versões distintas variando o nível de empatia, estilo de diálogo e personalização. Um experimento com 35 pacientes foi realizado e as interações dos usuários ao usarem as versões da MarIA foram analisadas. As informações trocadas evidenciaram diferenças nas taxas de engajamento, demonstrando a eficácia da customização, além de evidenciarem diálogos eficientes e seguros, evitando frustrações ou riscos.
Referências
Balsa, J., Félix, I., Cláudio, A. P., Carmo, M. B., Silva, I. C. E., Guerreiro, A., Guedes, M., Henriques, A., & Guerreiro, M. P. (2020). Usability of an Intelligent Virtual Assistant for Promoting Behavior Change and Self-Care in Older People with Type 2 Diabetes. Journal of medical systems, 44(7), 130.
Bodenheimer, T., Davis, C., and Holman, H. (2007). Helping Patients Adopt Healthier Behaviors. Clinical Diabetes, 25(2), 66-70.
Dhuliawala, S., Komeili, M., Xu, J., Raileanu, R., Li, X., Celikyilmaz, A., and Weston, J. (2023). Chain-of-verification reduces hallucination in large language models.
Furtado, E. S., Furtado, L. S. (2023). Como ensinar Aspectos de IHC de forma desplugada? O uso de um espaço cultural como metáfora de interação. Revista Brasileira de Informática na Educação, [S. l.], v. 31, p. 488–510, 2023.
Goh, E., Bunning, B., Khoong, E., Gallo, R., Milstein, A., Centola, D., & Chen, J. H. (2023). ChatGPT Influence on Medical Decision-Making, Bias, and Equity: A Randomized Study of Clinicians Evaluating Clinical Vignettes. medRxiv : the preprint server for health sciences, 2023.11.24.23298844.
Heisler, M. and Resnicow, K. (2008). Helping Patients Make and Sustain Healthy Changes: A Brief Introduction to Motivational Interviewing in Clinical Diabetes Care. Clinical Diabetes, 26(4), 161–165.
Huang, M., Zhu, X., and Gao, J. (2020). Challenges in Building Intelligent Open-Domain Dialog Systems. ACM Transactions on Information Systems (TOIS), 38(3), 1-32. arXiv:1905.05709v3.
Liu, P., Yuan, W., Fu, J., Jiang, Z., Hayashi, H., & Neubig, G. (2023). Pre-train, prompt, and predict: A systematic survey of prompting methods in natural language processing. ACM Computing Surveys, 55(9), 1-35. arXiv:2107.13586
Muñoz, D., Pedell, S., and Sterling, L. (2022). Understanding Confidence of Older Adults for Embracing Mobile Technologies. In Proceedings of the 34th Australian Conference on Human-Computer Interaction (pp. 38-50).
Nikitina, S., Callaioli, A., and Baez, M. (2018). Smart conversational agents for reminiscence. In 1st Workshop on Software Engineering for Cognitive Services (pp. 52–57). ACM, NY, USA.
Nigh, M. (2023). ChatGPT3 Prompt Engineering. GitHub. Retrieved from [link].
Paradigmxyz/flux (2023). Flux: Graph-based LLM power tool for exploring many completions in parallel. Retrieved from [link]
Ramesh, A., Dhariwal, P., Nichol, A., Chu, C., and Chen, M. (2022). Hierarchical Text-Conditional Image Generation with CLIP Latents. arXiv:2204.06125, 1(2), 3.
Sharma, D. Kaushal, S. Kumar, H. and Gainder, S. (2022) "Chatbots in Healthcare: Challenges, Technologies and Applications," 4th International Conference on Artificial Intelligence and Speech Technology (AIST), Delhi, India, 2022, pp. 1-6, DOI: 10.1109/AIST55798.2022.10065328.
Teo, S. (2023). How I Won Singapore’s GPT-4 Prompt Engineering Competition: A deep dive into the strategies I learned for harnessing the power of Large Language Models (LLMs). Towards Data Science. Retrieved from [link]
Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., Chi, E. H., Le, Q. V., and Zhou, D. (2023). Chain-of-Thought Prompting Elicits Reasoning in Large Language Models. In Proceedings of the 36th Conference on Neural Information Processing Systems, 35, 24824-24837. arXiv:2201.11903v6.
Bodenheimer, T., Davis, C., and Holman, H. (2007). Helping Patients Adopt Healthier Behaviors. Clinical Diabetes, 25(2), 66-70.
Dhuliawala, S., Komeili, M., Xu, J., Raileanu, R., Li, X., Celikyilmaz, A., and Weston, J. (2023). Chain-of-verification reduces hallucination in large language models.
Furtado, E. S., Furtado, L. S. (2023). Como ensinar Aspectos de IHC de forma desplugada? O uso de um espaço cultural como metáfora de interação. Revista Brasileira de Informática na Educação, [S. l.], v. 31, p. 488–510, 2023.
Goh, E., Bunning, B., Khoong, E., Gallo, R., Milstein, A., Centola, D., & Chen, J. H. (2023). ChatGPT Influence on Medical Decision-Making, Bias, and Equity: A Randomized Study of Clinicians Evaluating Clinical Vignettes. medRxiv : the preprint server for health sciences, 2023.11.24.23298844.
Heisler, M. and Resnicow, K. (2008). Helping Patients Make and Sustain Healthy Changes: A Brief Introduction to Motivational Interviewing in Clinical Diabetes Care. Clinical Diabetes, 26(4), 161–165.
Huang, M., Zhu, X., and Gao, J. (2020). Challenges in Building Intelligent Open-Domain Dialog Systems. ACM Transactions on Information Systems (TOIS), 38(3), 1-32. arXiv:1905.05709v3.
Liu, P., Yuan, W., Fu, J., Jiang, Z., Hayashi, H., & Neubig, G. (2023). Pre-train, prompt, and predict: A systematic survey of prompting methods in natural language processing. ACM Computing Surveys, 55(9), 1-35. arXiv:2107.13586
Muñoz, D., Pedell, S., and Sterling, L. (2022). Understanding Confidence of Older Adults for Embracing Mobile Technologies. In Proceedings of the 34th Australian Conference on Human-Computer Interaction (pp. 38-50).
Nikitina, S., Callaioli, A., and Baez, M. (2018). Smart conversational agents for reminiscence. In 1st Workshop on Software Engineering for Cognitive Services (pp. 52–57). ACM, NY, USA.
Nigh, M. (2023). ChatGPT3 Prompt Engineering. GitHub. Retrieved from [link].
Paradigmxyz/flux (2023). Flux: Graph-based LLM power tool for exploring many completions in parallel. Retrieved from [link]
Ramesh, A., Dhariwal, P., Nichol, A., Chu, C., and Chen, M. (2022). Hierarchical Text-Conditional Image Generation with CLIP Latents. arXiv:2204.06125, 1(2), 3.
Sharma, D. Kaushal, S. Kumar, H. and Gainder, S. (2022) "Chatbots in Healthcare: Challenges, Technologies and Applications," 4th International Conference on Artificial Intelligence and Speech Technology (AIST), Delhi, India, 2022, pp. 1-6, DOI: 10.1109/AIST55798.2022.10065328.
Teo, S. (2023). How I Won Singapore’s GPT-4 Prompt Engineering Competition: A deep dive into the strategies I learned for harnessing the power of Large Language Models (LLMs). Towards Data Science. Retrieved from [link]
Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., Chi, E. H., Le, Q. V., and Zhou, D. (2023). Chain-of-Thought Prompting Elicits Reasoning in Large Language Models. In Proceedings of the 36th Conference on Neural Information Processing Systems, 35, 24824-24837. arXiv:2201.11903v6.
Publicado
25/06/2024
Como Citar
SILVA, Vitória; FURTADO, Elizabeth Sucupira; OLIVEIRA, Juliana; FURTADO, Vasco.
Engenharia de Prompts em Assistentes Conversacionais para Promoção de Autocuidado baseados em Modelos Amplos de Linguagem. In: SIMPÓSIO BRASILEIRO DE COMPUTAÇÃO APLICADA À SAÚDE (SBCAS), 24. , 2024, Goiânia/GO.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2024
.
p. 377-388.
ISSN 2763-8952.
DOI: https://doi.org/10.5753/sbcas.2024.2252.
