Electrotechnical and Computer Engineering
Vol. 39 No. 06 (2024): Proceedings of Faculty of Technical Sciences
MEDICAL CHATBOT BASED ON ENCODER-DECODER ARCHITECTURE
Abstract
The goal of this thesis is to create conversational chatbot which utilizes advanced machine learning algorithms and NLP techniques to predict diagnosis or provide treаtment recommendations based on patient's symptoms. This paper presents experimental results that show that performances are better when the chatbot is trained with a data set in which there is an alternative paraphrasing of questions and where for similar questions it receives generalized answers, rather than answers that are very narrowly specialized and personalized for the patient's stated question.
References
[1] Sutskever, O. Vinyals, and Q. V. Le, “Sequence to Sequence Learning with Neural Networks”.
[2] P. F. Brown et al., “A STATISTICAL APPROACH TO MACHINE TRANSLATION,” vol. 16, no. 2, 1990.
[3] S. Hochreiter and J. Schmidhuber, “Long Short-Term Memory,” Neural Comput., vol. 9, no. 8, pp. 1735–1780, 1997, doi: 10.1162/neco.1997.9.8.1735
[4] X. He et al., “MedDialog: Two Large-scale Medical Dialogue Datasets.” arXiv, Jul. 07, 2020.
doi: http://arxiv.org/abs/2004.03329
[5] https://www.kaggle.com/datasets/tusharkhete/dataset-for-medicalrelatedchatbots