자연어 처리(NLP) 기반 텍스트마이닝을 적용한 한국체육측정 평가학회지 연구동향 분석: 국문초록을 중심으로

초록

This study utilized Topic Modeling (LDA) to analyze the research trends of The Korean Journal of Measurement and Evaluation in Physical Education and Sport Science. To achieve the research objectives, we selected publications from the first issue of 1999 to the first issue of 2023 in the Journal of the Korean Society of Measurement and Evaluation in Physical Education. Among them, 59 Korean abstracts from 1999 to 2004, which were impossible to crawl, were manually transcribed, and 520 research papers’ Korean abstracts were chosen for analysis. The preprocessed data from the Korean abstracts underwent top keyword frequency analysis, N-gram analysis, Topic Modeling (LDA), topic occupancy rate, and correlation between the year and topic occupancy rate using Python 3.5 and SPSS 27. The results showed that ‘Topic1: Development of physical performance measures and evaluation criteria’, ‘Topic3: Analysis of physical characteristics by sports event’, ‘Topic4: Physical Examination’, and ‘Topic5: Measurement and analysis tool fitness’ displayed a negative correlation, but it was not significant. The ‘Topic6: The research on physical education in school’ topic had consistent fluctuations in the occupancy rate over time, but it was not statistically significant. On the other hand, the ‘Sports analysis and Big Data Utilization’ topic showed a continuous increase in occupancy rate as the year increased, and a significant positive correlation was observed. Especially, the restrictions on clinical experiments due to the spread of COVID-19 appeared to have increased the research on ‘Sports Analysis and Big Data’. and it is expected that there will be a consensus among researchers for the systematization and scientific analysis of sports game analysis techniques using big data in the future.

키워드