스포츠 애널리틱스와 인공지능에 관한 체계적 문헌 검토 : 기술에 따른 연구 변화와 과제 = A Systematic Review on the Sports Analytics and Artificial Intelligence: Research Trends and Challenges by Technology

다국어 초록 (Multilingual Abstract)

Purpose This study was conducted to analyze changes in the latest AI research within the field of sports analytics, as AI becomes commonplace and research is actively pursued across all disciplines. The purpose of this study was to systematically analyze research trends in sports analytics utilizing artificial intelligence (AI) in the Korean academics, present research changes according to technological advancements, and suggest future research agenda. Methods Using the framework of a systematic literature review on AI research in the sports field conducted in January 2024, literature related to domestic professional sports was searched on DBPia, Kyobo Scholar, and Google Scholar between January 8, 2025, and January 8, 2025, using a combination of keywords such as 'sports,' 'physical education,' 'AI,' 'machine learning,' 'artificial intelligence,' 'machine learning,' and 'sports analytics.' A total of 28 publications were analyzed. Results The results of this study are as follows: By sport, baseball (8 cases) was the most frequent, followed by general research (7 cases), soccer (5 cases), and ssireum (Korean wrestling) (4 cases). In terms of analysis methods, game data analysis (13 cases) was the most common, followed by case/literature analysis (6 cases) and review/trend analysis (4 cases). Regarding the unit of analysis, league-level analysis (8 cases) was the most prevalent, followed by individual (6 cases) and industry (5 cases). As for the research objective, practical suggestions (9 cases) were the most common, followed by game result prediction (6 cases) and research trend analysis (4 cases). A systematic review revealed that the most frequently used AI algorithms were Random Forest, XGBoost, and Gradient Boost, each appearing 6 times. It was discussed that sports analytics applying AI is evolving as a representative example of promoting the convergence scientific value of the sports industry, and the accumulation of key agendas and empirical results based on sports game data analysis is important. Conclusion Future research should move beyond simply presenting accuracy through algorithms and focus on analyzing the impact of game data analysis on the sports industry and its future directions from an integrated perspective. Furthermore, discussions should encompass not only artificial intelligence as an engineering technology but also the impact of sports analytics from a convergence perspective, including social implications and the promotion of fan engagement, among other diverse viewpoints.