토론봇의 윤리적 편향성 방지를 위한 고려 요인 = The Consideration Factors of Ethical Bias for Debate Bot

다국어 초록 (Multilingual Abstract)

With the development of artificial intelligence technology, attempts to use chatbot in education are increasing. Debate Bot with natural language processing technology can communicate naturally with user by recognizing users' Intentions and context. Debate Bot can be used for debate education. Learning debate can improve students' communication skills, data analysis skills, and critical thinking skills. It is difficult to conduct the debate at the school site due to time limitation and participation issues. Debate Bot can solve these problems by providing users with automated responses. However, various ethical issue such as data bias can arise because what chatbot say is provided by algorithm based on collected data. Based on this problem, this study suggested principle for developing debate chatbot. As results, the most important principle was to eliminate designer’s bias on collecting data. And inhuman like conversation needs to be removed for make the communication be more human like.