초등영어 예비교사들의 AI 챗봇 개발 활동 연구

A study on the development of artificial intelligence(AI) chatbot by pre-service student teachers

펼치기

This study has evaluated 9 AI chatbots developed by a class of 19 pre-service teacher students in order to find out whether those chatbots work properly and pre-service student teachers’ attitudes to developing chatbot experiences. The summary of the study is as follows. First, four types of evaluation are used: 1) asking questions specific to the entity, 2) evaluating whether prompts work properly, 3) evaluating whether or not the partial utterances were accepted depending on the work of prompts and finally, 4) performing the actual learning task. According to the analysis, 1) the chatbot had a 57.4% success rate in setting up the entity. 2) Prompt operation in relation to the entity showed a success rate of 50%. 3) Initiating prompts by part of speech was 57.4% successful. 4) Finally, six chatbots have succeeded in performing learning task. Second, 9 out of 10 items in the active and collaborative learning domain show more than 3.0 and positive recognition. Six out of seven items in the learning strategy category have scored at least 3.5, indicating that the collaborative task had students actively communicate with peers and contribute to effectively completing the task. In addition, the average time spent for finishing the task was 34 hours or more, which is a surprising result considering that the total semester hours for the course work are 30 hours per semester. The findings of this study suggest that it’s the only beginning of a collective effort to implement AI in education. We are supposed to face the music and give it a try.