中醫學中的問診是其中一個重要的診斷方法，而了解病患的體質可以協助中醫師更精確地評估個體的身體狀況，從而為病患提供針對性的醫療建議。為了評估個體的體質特徵，中國醫藥大學中醫學院的蘇奕彰教授提出了中醫體質量表，該量表可以根據不同的問題將人體的體質特徵分成不同的類別。
然而，若在看診前提供病患填寫中醫體質量表，可能會因為題目過多且繁瑣，導致病患填寫時不耐煩，甚至亂填，這將嚴重影響醫師對體質的正確評估。因此，開發一種自然語言處理的體質問診系統具有重要的實踐價值。
中醫體質自然語言問診系統利用自然語言處理和經微調後的BERT模型，對使用者輸入的內容進行分析和判斷，自動生成相關問診問題，最終判斷使用者的體質屬性。
本研究旨在設計和實現一個能根據使用者回答自動判斷體質的問診系統，以提供更便捷的中醫體質檢測服務。該系統的架構結合了自然語言處理和決策樹算法，通過線上交互的方式為使用者檢測體質。

In Traditional Chinese Medicine (TCM), questioning is one of the essential diagnostic methods, and understanding a patient's constitution assists TCM practitioners in accurately assessing their physical condition and providing tailored medical advice. To evaluate an individual's constitutional characteristics, Professor Su Yi-Chang, who is working at the University of Traditional Chinese Medicine in Taiwan, proposed the Body Constitutions Questionnaire(BCQ). This questionnaire categorizes an individual's constitutional traits into different categories based on various questions.
However, providing patients with the lengthy and cumbersome BCQ to fill out before the consultation may lead to impatience or even random responses, which can potentially affect the accurate assessment of the patient's constitution by TCM practitioners. Therefore, the development of a natural language processing-based constitution questioning system holds significant practical value.
The Traditional Chinese Medicine Natural Language Body Constitutions Consultation System utilizes natural language processing and a fine-tuned BERT (Bidirectional Encoder Representations from Transformers) model to analyze and interpret user inputs, automatically generating relevant questioning prompts, and ultimately determining the user's constitutional attributes.
This study aims to design and implement an inquiry system that can automatically determine a person's constitution based on their responses, in order to provide a more convenient traditional Chinese medicine constitutional testing service. The system architecture combines natural language processing and decision tree algorithms to detect a user's constitution through online interaction.

致謝 I
摘要 II
ABSTRACT III
目錄 IV
圖目錄 VII
表目錄 IX
第一章、 緒論 1
1-1 研究背景 1
1-2 研究動機與目的 3
1-3 論文架構 4
第二章、 相關研究與文獻探討 5
2-1 相關文獻 5
2-1-1 自然語言處理在醫療領域的應用 5
2-1-2 基於決策樹的聊天機器人 6
2-2 中醫體質量表 6
2-3 BERT模型 8
2-4 ALBERT 10
2-5 決策樹 11
2-5-1 決策樹演算法 11
2-5-2 基於決策樹之中醫體質線上問診系統 12
第三章、 研究方法 13
3-1 系統架構概述 13
3-2 微調BERT模型 14
3-2-1 資料收集 14
3-2-2 資料預處理 15
3-2-3 模型訓練及比較 16
3-3 應用決策樹 20
3-4 系統實作細節 23
3-4-1 Line Bot實作與整合 23
3-4-2 部署於雲端 24
3-4-3 成品展示 26
第四章、 實驗設計與結果討論 33
4-1 實驗設計 33
4-2 實驗結果分析 33
4-2-1 BERT模型效果分析 33
4-2-2 體質分類準確性 36
4-3 問題討論 39
第五章、 結論與未來展望 41
5-1 結論 41
5-2 未來展望 42
參考文獻 43
附錄 45
附錄一 頻率語詞得分情形 45
附錄二 強度語詞得分情形 46

[1]A. Ben Abacha and P. Zweigenbaum, "MEANS: A medical question-answering system combining NLP techniques and semantic Web technologies," Information Processing & Management, vol. 51, no. 5, pp. 570-594, 2015/09/01/ 2015, doi: https://doi.org/10.1016/j.ipm.2015.04.006.
[2]L.-L. Chen et al., "BCQ+: A Body Constitution Questionnaire to Assess Yang-Xu: Part II: Evaluation of Reliability and Validity," Forschende Komplementärmedizin / Research in Complementary Medicine, vol. 16, no. 1, pp. 20-27, 2009, doi: 10.1159/000197770.
[3]S.-L. Chen, Y.-T. Liu, K.-C. Hsueh, and P.-L. Tang, "Body constitution of traditional Chinese medicine caused a significant effect on depression in adult women," Complementary Therapies in Clinical Practice, vol. 42, p. 101288, 2021/02/01/ 2021, doi: https://doi.org/10.1016/j.ctcp.2020.101288.
[4]C.-C. Chern, Y.-J. Chen, and B. Hsiao, "Decision tree–based classifier in providing telehealth service," BMC Medical Informatics and Decision Making, vol. 19, no. 1, p. 104, 2019/05/30 2019, doi: 10.1186/s12911-019-0825-9.
[5]K. Chung, H. Cho, and J. Park, "A Chatbot for Perinatal Women's and Partners' Obstetric and Mental Health Care: Development and Usability Evaluation Study," JMIR Medical Informatics, vol. 9, p. e18607, 03/03 2021, doi: 10.2196/18607.
[6]A. Conneau, D. Kiela, H. Schwenk, L. Barrault, and A. Bordes, "Supervised learning of universal sentence representations from natural language inference data," arXiv preprint arXiv:1705.02364, 2017.
[7]J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, "Bert: Pre-training of deep bidirectional transformers for language understanding," arXiv preprint arXiv:1810.04805, 2018.
[8]C. Lab. "CKIP Transformers." https://ckip-transformers.readthedocs.io/en/stable/index.html (accessed.
[9]Z. Lan, M. Chen, S. Goodman, K. Gimpel, P. Sharma, and R. Soricut, "Albert: A lite bert for self-supervised learning of language representations," arXiv preprint arXiv:1909.11942, 2019.
[10]J.-D. Lin, L.-L. Chen, J.-S. Lin, C.-H. Chang, Y.-C. Huang, and Y.-C. Su, "BCQ–: A Body Constitution Questionnaire to Assess Yin-Xu: Part I: Establishment of a Provisional Version through a Delphi Process," Forschende Komplementärmedizin / Research in Complementary Medicine, vol. 19, no. 5, pp. 234-241, 2012, doi: 10.1159/000343580.
[11]J.-D. Lin, J.-S. Lin, L.-L. Chen, C.-H. Chang, Y.-C. Huang, and Y.-C. Su, "BCQs: A Body Constitution Questionnaire to assess Stasis in traditional Chinese medicine," European Journal of Integrative Medicine, vol. 4, no. 4, pp. e379-e391, 2012/12/01/ 2012, doi: https://doi.org/10.1016/j.eujim.2012.05.001.
[12]J.-S. Lin et al., "BCQ–: A Body Constitution Questionnaire to Assess Yin-Xu: Part II: Evaluation of Reliability and Validity," Forschende Komplementärmedizin / Research in Complementary Medicine, vol. 19, no. 6, pp. 285-292, 2012, doi: 10.1159/000346060.
[13]J. Pennington, R. Socher, and C. Manning, "GloVe: Global Vectors for Word Representation," Doha, Qatar, October 2014: Association for Computational Linguistics, in Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pp. 1532-1543, doi: 10.3115/v1/D14-1162. [Online]. Available: https://doi.org/10.3115/v1/D14-1162
[14]I. W. Puspitasari, F. R. Rinawan, W. G. Purnama, H. Susiarno, and A. I. Susanti, "Development of a Chatbot for Pregnant Women on a Posyandu Application in Indonesia: From Qualitative Approach to Decision Tree Method," Informatics, vol. 9, no. 4, p. 88, 2022. [Online]. Available: https://www.mdpi.com/2227-9709/9/4/88.
[15]J. R. Quinlan, "Induction of Decision Trees," Machine Learning, vol. 1, pp. 81--106, 1986.
[16]Y.-C. Su, L.-L. Chen, J.-D. Lin, J.-S. Lin, Y.-c. Huang, and J.-S. Lai, "BCQ+: A Body Constitution Questionnaire to Assess Yang-Xu: Part I: Establishment of a First Final Version through a Delphi Process," Forschende Komplementärmedizin / Research in Complementary Medicine, vol. 15, no. 6, pp. 327-334, 2008, doi: 10.1159/000175938.
[17]A. Wang, A. Singh, J. Michael, F. Hill, O. Levy, and S. R. Bowman, "GLUE: A multi-task benchmark and analysis platform for natural language understanding," arXiv preprint arXiv:1804.07461, 2018.
[18]林幼淳, "兩種中醫體質量表之一致性評估," 碩士, 公共衛生學系碩士班, 中國醫藥大學, 台中市, 2016. [Online]. Available: https://hdl.handle.net/11296/3w7wtu
[19]葉士誠, "基於決策樹之中醫體質線上問診系統," 碩士, 資訊工程學系, 國立東華大學, 花蓮縣, 2021. [Online]. Available: https://hdl.handle.net/11296/feyr69
[20]蘇奕彰 and 陳麗麗, "中醫體質量表及分型指標之建立(II)," 2005.
[21]蘇奕彰, 陳麗麗, and 賴俊雄, "中醫體質量表及分型指標之建立," 2004.