• The Journal of the Korea Contents Association
• /
• v.21 no.7
• /
• pp.144-156
• /
• 2021

This study was conducted to explore important factors to consider when designing voice-based psychological counseling agents amid the increasing use of conversational agents in counseling and psychotherapy. 48 participants selected their preferred agent's voice among four types (young women and men, middle-aged women and men) and had a conversation with a psychological counseling agent. They also evaluated the reasons for voice selection, mood changes, perception of the agent's characteristics, and counseling outcomes. As a results, the agent's voice type selected according to the user's gender was not statistically significant. However, the qualitative analysis showed 'comfort' of the voice was an important factor. Next, the user's mood improved significantly after the conversation with the agent, which confirmed the intervention effect. Finally, it was found that expertness and attractiveness perceptions toward the agent contributed to the counseling outcomes. The implications of the study and suggestions for future research were discussed.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2002.05a
• /
• pp.32-36
• /
• 2002

심리상담의 영역이 확대되어 감에 따라 오프라인 상담뿐만 아니라 온라인 형태의 상담이 급속히 발전하고 있다. 인터넷의 활용도가 증가함에 따라 컴퓨터를 의사소통의 매개로 활용한 사이버상담 형태를 통한 상담도 체계적으로 개발되어야 할 필요성이 있다. 그러나 현재 개설되어 운영되고 있는 사이버상담은 내담자에게 적합한 맞춤상담이 불가능하며 또한 자가치유가 가능한 자가치유시스템이 없는 실정이다. 따라서 본 논문에서는 지능형 에이전트를 이용하여 내담자에게 적합한 맞춤상담이 가능한 방법을 제안함과 동시에 과거 상담사레 데이터베이스를 바탕으로 이전의 상담사레들을 신경망의 BP학습알고리즘을 이용하여 학습을 시킨 후 자가치유가 가능한 자가치유시스템을 설계하는 방법을 제안한다.