의료기관 콜상담 데이터 기반 노쇼 위험 신호 탐색 연구: 미용 성형 분야의 시술 확정군과 노쇼군의 텍스트마이닝

초록

No-show behavior in medical institutions reduces operational efficiency, causes financial loss, and lowers reservation utilization. While prior work has predicted no-shows using demographics, reservation variables, and communication interventions, this study examined linguistic cues and decision processes in real consultations using call-center data from a cosmetic plastic surgery institution. STT consultation texts were collected and grouped into procedure confirmed (n=1,859) and no-show (n=363). Using TF-IDF topic analysis and semantic network analysis, we compared language patterns between groups. The confirmed group emphasized treatment and scheduling terms and decision-oriented communication, while the no-show group more often used postponement and promotion-seeking language, comparison-shopping expressions, and schedule uncertainty. These signals may help detect no-show risk early and support targeted call-center interventions.

키워드