인공지능 챗봇에 대한 선문답 알고리즘의 데이터 ‒ 심리치료 상담챗봇을 중심으로 ‒

Abstract

This study is an examination about whether mental health chatbots using leading-edge AI technology in the fourth industrial revolution can incorporate a selection of algorithms producing human interactive data for Seon encounter dialogues (SEDs). This comes from the finding that such similarities as simplicity, goal-directedness, and optimization exist between SEDs and algorithms. Thus, “keywords” scanned from the vast SED-oriented conversational patterns could be algorithmized by using “deep learning.” It can, then, be clustered to build data sets. The possible domains of such mental “telepathy” may encompass algorithms that provide examples of non-logical mental representation, algorithms that provide metaphorical examples and algorithms that provide conversation topics (Hwadu). This data group might show individually unique dialogue patterns inclusive of psychotherapy effectiveness. The current chatbot metrics rely on natural language processing within the established database system; however, SED-specific algorithms reply with more creative keywords. This goes beyond conventional quasi-stereotyped response patterns. This means that SED algorithm『powered mental health chatbot can offer more effective counseling based on much more data, making them access the wisdom of meditation masters in the traditional Seon Buddhism.