온라인 리뷰 데이터와 Kano 모델을 기반으로 미술관 사용자 요구 분석 －중국미술관(中国美术馆)을 중심으로 -

Abstract

This study proposes a method to identify visitor needs at the National Art Museum of China by integrating online review data with text mining under the Kano model framework. Using LDA topic modeling, experience dimensions were extracted, followed by aspect-based sentiment analysis with a BERT model. Finally, the Penalty-Reward Contrast Analysis (PRCA) was applied to classify Kano attributes and determine optimization priorities. Results show that visitors are most concerned with five dimensions: exhibition quality, exhibit types and content, transportation and entry experience, service facilities and visiting experience, and exhibition environment and architectural features. Exhibition quality was identified as an expectation attribute, where positive experiences moderately increase satisfaction but negative ones have stronger effects. The other four dimensions were classified as basic attributes, where negative reviews significantly reduce overall ratings. Based on IA and AP indicators, optimization priorities were ranked. The findings demonstrate the feasibility of using user-generated content combined with text mining and the Kano model to analyze and improve museum visitor experience.