結合正向行為支持之聊天機器人於改善國小身心障礙學生社交技巧

Integrating Positive Behavioral Support Strategy into a Chatbot Application for Improving the Social Skills of Elementary Students with Disabilities

黃瑜陞(Yu-Sheng Huang)

指導教授 ： 張家瑋

國立臺中科技大學/資訊與流通學院/資訊工程系碩士班/碩士(2023年)

摘要

本研究以Rasa框架建構聊天機器人，並比較Rasa透過不同組件訓練後的準確率；此聊天機器人以身心障礙學生較缺乏的社交技巧而導致的問題行為為目標，以正向行為支持理論為基礎結合聊天機器人讓學生進行替代行為的練習，最後使用科技接受模型所建構的問卷做結果分析，而本研究旨在探討透過聊天機器人讓學生進行練習是否具有成效、學生對於聊天機器人是否有興趣以及聊天機器人是否能減輕教師的教學負擔。
研究結果顯示學生使用前後的測驗具有正向的顯著差異，學生的使用興趣問卷平均分數為 4.21，而教師在科技接受模型的感知有用性構面平均分數為 3.73 整體而言研究問題的結果都是成立的。

關鍵字

正向行為支持 ； Rasa ； 聊天機器人 ； 科技接受模型

並列摘要

This study builds a chatbot using the Rasa framework and compares the accuracy of Rasa trained with different components. The goal of this chatbot is to address problem behaviors caused by a lack of social skills in students with disabilities. Developing the chatbot based on the positive behavior theory. Allowing students to use the chatbot for practice and finally analyzing the results using a questionnaire constructed from the technology acceptance model (TAM). The results showed a significant positive difference in the results before and after the students used the chatbot. The average score of the students' interest toward using questionnaire is 4.21, and the average score of the teachers' perceived usefulness dimension in the technology acceptance model is 3.73. Overall, the results of the research questions are confirmed.

並列關鍵字

positive behavior support ； Rasa ； chatbot ； TAM

參考文獻

[1] S. Meshram, N. Naik, M. VR, T. More and S. Kharche, “Conversational AI: Chatbots,” International Conference on Intelligent Technologies (CONIT), pp. 1-6, 2021.

[2] V. Dhanasekar, Y. Preethi, V. S, P. J. I. R and B. P. M, “A Chatbot to promote Students Mental Health through Emotion Recognition,” Third International Conference on Inventive Research in Computing Applications (ICIRCA), pp. 1412-1416, 2021.

[3] A. Gupta, M. A. Raj, K. Singh and R. Deshmukh, “REDE - Detecting human emotions using CNN and RASA,” International Conference for Advancement in Technology (ICONAT), pp. 1-6, 2022.

[4] H. C. van Cuylenburg and T. N. D. S. Ginige, “Emotion Guru: A Smart Emotion Tracking Application with AI Conversational Agent for Exploring and Preventing Depression,” International Conference on UK-China Emerging Technologies (UCET), pp. 1-6, 2021.

[5] S. Narynov, Z. Zhumanov, A. Gumar, M. Khassanova and B. Omarov, “Chatbots and Conversational Agents in Mental Health: A Literature Review,” 21st International Conference on Control, Automation and Systems (ICCAS), pp. 353-358, 2021.