텍스트마이닝을 활용한 작업관련성 근골격계질환 연구동향 분석

Abstract

Domestic musculoskeletal disorders have been on the rise since the 1997 IMF crisis, as rapid changes in work environments and practices, such as mechanization, automation, and specialization, have been implemented by various companies to enhance productivity. Furthermore, with increasing awareness and concern among workers, the government has been conducting harmful factor and causal investigations related to musculoskeletal disorders caused by burdensome work tasks through the amendment of Article 39 of the Industrial Safety and Health Act since 2003.
 To ensure the proactive prevention of work-related musculoskeletal disorders and minimize work-related losses for workers' safety, there is a continuous need for systematic and comprehensive research. To address this, the present study conducted a trend analysis to explore new perspectives for preventing industrial accidents.
 The literature collection comprised a total of 653 domestic theses and academic papers from 2003 to 2022 that included the keyword 'musculoskeletal disorder,' gathered through search engines such as the Korean Citation Index (KCI) and the Research Information Sharing Service (RISS). The data were categorized into two periods: the first period from 2003 to 2012 and the second period from 2013 to 2022. Text mining was employed to analyze the frequency (TF) and network connections, which were subsequently visualized.
 The results revealed that the top keywords in work-related musculoskeletal disorder research throughout the entire period (2003 to 2022) were 'work,' 'musculoskeletal disorder,' and 'symptoms,' consistently recognized as essential research topics in both the first and second periods. In the first period, the most frequently discussed factors influencing workers' musculoskeletal disorders included 'posture,' 'physical condition,' 'environment,' 'job stress,' 'time,' 'age,' and the term 'survey' was prominent among the top keywords. In the second period, keywords like 'comparison,' 'development,' 'measurement,' 'method,' and 'investigation' emerged.
 N-Gram analysis indicated that the simultaneous occurrence of the 'musculoskeletal disorder-symptoms' keyword was the most frequent throughout the entire period, followed by high frequencies of 'musculoskeletal disorder,' 'pain,' 'awareness,' and 'burden.' In the first period, body part-related keywords like 'hand-wrist,' 'shoulder-back,' 'wrist-finger,' and 'back-leg' were prominent. In the second period, keywords such as 'workload,' 'safety-health,' 'factor-investigation,' 'risk-assessment,' and 'occupational accident-insurance' were observed.
 This study attempted the first analysis of theses and academic papers on work-related musculoskeletal disorders using text mining, a big data analysis method. The findings contribute to suggesting future research directions on work-related musculoskeletal disorders, holding significant implications for the field.