본 연구는 미국의 프로야구 경기장을 방문한 관람객의 온라인 리뷰를 바탕으로 텍스트 마이닝과 IPA(Importance-Performance Analysis) 기법을 적용하여 프로야구 경기장 스포츠스케이프에 대한 관람객의 만족도를 파악하는 것을 목표로 한다. 기존 프로야구 경기장 스포츠스케이프와 만족도에 관한 연구는 설문 조사를 기반으로 진행되어왔으며, 이러한 설문 조사 기법은 시간이 오래 걸리고 큰 비용이 요구된다는 한계점이 있다. 이에 따라 본 연구에서는 텍스트 마이닝 방법론과 IPA를 결합한 형태의 방법론을 채택한다. 본 연구를 진행하기 위해 Python Selenium으로 구글 지도에서 미국 프로야구 경기장 15곳의 리뷰를 수집하였다. BERTopic 토픽모델링을 진행하여 스포츠스케이프의 주요 요인을 도출하였으며, 토픽모델링 결과에서 스포츠스케이프 요인에 대한 내용이 부족하거나 토픽 수가 적절하지 못한 결과를 제외하여 최종적으로 4개의 구장의 스포츠스케이프 요인을 도출하였다. Google Cloud Natural Language API를 활용한 감성분석 통해 만족도 점수를 추정하고, 딥러닝 방법론을 통해 만족도 점수를 계산하여 최종적으로 스포츠스케이프 요인에 대한 Importance-Performance Map을 작성하였다. 본 연구의 결과를 바탕으로 프로야구 구단은 관람객 경험을 향상시키기 위한 구체적인 방향을 제시하고, 특히 만족도가 낮은 영역의 개선을 통해 관람객 충성도와 경기장 방문율을 높이는 데 기여할 것으로 기대된다.|This study aims to understand spectator satisfaction with the sportscape of professional baseball stadiums by applying text mining and Importance-Performance Analysis (IPA) techniques based on online reviews of visitors to professional baseball stadiums in the United States. Existing studies on the sportscape and satisfaction of professional baseball stadiums have been conducted based on surveys, and these survey techniques have the limitation of being time-consuming and costly. Therefore, this study adopts a methodology that combines text mining methodology and IPA. To conduct this study, we used Python Selenium to collect reviews of 15 American professional baseball stadiums from Google Maps. We conducted BERTopic topic modeling to derive the main factors of the sportscape, and excluded the results that lacked the content of the sportscape factors or the number of topics from the topic modeling results, and finally derived the sports cape factors of six stadiums. We estimated the performance score through sentiment analysis using Google Cloud Natural Language API, calculated the importance score through deep learning methodology, and finally created the Importance-Performance Map for the sportscape factors. Based on the results of this study, it is expected that professional baseball clubs will provide specific directions to improve the visitor experience and contribute to increasing visitor loyalty and stadium visitation rates by improving areas of low satisfaction.
Alternative Abstract
This study aims to understand spectator satisfaction with the sportscape of professional baseball stadiums by applying text mining and Importance-Performance Analysis (IPA) techniques based on online reviews of visitors to professional baseball stadiums in the United States. Existing studies on the sportscape and satisfaction of professional baseball stadiums have been conducted based on surveys, and these survey techniques have the limitation of being time-consuming and costly. Therefore, this study adopts a methodology that combines text mining methodology and IPA. To conduct this study, we used Python Selenium to collect reviews of 15 American professional baseball stadiums from Google Maps. We conducted BERTopic topic modeling to derive the main factors of the sportscape, and excluded the results that lacked the content of the sportscape factors or the number of topics from the topic modeling results, and finally derived the sports cape factors of six stadiums. We estimated the performance score through sentiment analysis using Google Cloud Natural Language API, calculated the importance score through deep learning methodology, and finally created the Importance-Performance Map for the sportscape factors. Based on the results of this study, it is expected that professional baseball clubs will provide specific directions to improve the visitor experience and contribute to increasing visitor loyalty and stadium visitation rates by improving areas of low satisfaction.
