α世代運動行銷5.0-青少年運動推廣效率與人工智慧運用

αGeneration sports marketing 5.0: Enhancing adolescent sports promotion efficiency with artificial intelligence applications

緒論：本文整合來自四大政府出版物或資料庫的政府資料開放，分別為「學生參與各級教育」、「現有運動情況調查」、「運動統計」、「民眾運動消費支出調查」，進行資料再用，共計22個變項或指標。除總結「社會生態模型」探討社會運動、學校運動及家庭運動，亦整合「青少年身體活動推廣模型」之使能因子、傾向因子及增強因子，共計兩個模型剖析臺灣青少年運動推廣與行銷的關鍵議題。方法：經整合四大政府資料開放、共計22個指標，本文運用倒傳遞類神經網路預測「我國民眾運動消費支出調查」並補充2022年數據資訊；我們使用網絡資料包絡分析法衡量2018至2022年臺灣22個縣市青少年運動推廣效率，接著採用拔靴法截斷迴歸識別影響青少年運動推廣的關鍵因素；最終運用倒傳遞類神經網絡資料包絡分析法預測2023年青少年運動推廣效率。結果：本文衡量2018至2022年臺灣22個縣市青少年運動推廣效率，並結合人工智慧技術預測2023年青少年運動推廣效率。趨勢分析顯示，臺北市和花蓮縣連續六年保持最佳效率，九個縣市呈現效率增長趨勢，十一個縣市呈現效率下降趨勢。「接收信息頻率」、「了解足夠運動時間」、「體育課程喜好程度」及「步行至運動場地的時間」是青少年運動發展的關鍵因素。結論：本文構建青少年運動推廣效率模型，並衡量與預測青少年運動推廣效率和趨勢分析。我們證明青少年運動推廣效率模型中的關鍵因素，並驗證運動發展的重要性。本文啟動運動科技運用，提出運動推廣前瞻性思維與未來研究建議，並相對應α世代的運動行銷5.0。

Introduction: This study integrates open government data from four major sources: ''Students' Participation in All Levels of Education,'' ''Survey of the Existing Sport Situation,'' ''Sports Statistics,'' and ''Survey of Sports Consumer Expenditure in Taiwan,'' reusing a total of 22 variables or indicators. In addition to summarizing the ''Social Ecological Model'' to examine the social, school, and family sports environments, this study incorporates the ''Youth Physical Activity Promotion Model'' to analyze enabling, predisposing, and reinforcing factors. These two models were applied to explore key issues in adolescent sports promotion and marketing in Taiwan. Methods: By integrating the four datasets and 22 indicators, this study used a back-propagation neural network to predict the ''Survey of Sports Consumer Expenditure in Taiwan'' and supplemented data for 2022. Network Data Envelopment Analysis (DEA) was employed to evaluate adolescent sports promotion efficiency in 22 counties and cities in Taiwan from 2018 to 2022. Bootstrap truncated regression was then applied to identify key factors influencing adolescent sports promotion. Finally, a back-propagation neural network DEA was used to predict adolescent sports promotion efficiency for 2023. Results: This study assessed adolescent sports promotion efficiency across 22 counties and cities in Taiwan from 2018 to 2022 and integrated artificial intelligence techniques to forecast efficiency for 2023. Trend analysis revealed that Taipei City and Hualien County consistently maintained the highest efficiency for six consecutive years. Nine counties and cities showed an increasing trend in efficiency, while eleven counties and cities exhibited a decline. Key factors identified in sport-for-development efficiency for adolescents included ''frequency of receiving information,'' ''awareness of sufficient exercise time,'' ''preference for physical education classes,'' and ''walking time to exercise facilities. Conclusion: This study constructs a model for evaluating and predicting adolescent sports promotion efficiency, identifying key factors that contribute to sport-for-development efforts. It validates the importance of sport-for-development and introduces the application of sports technology, offering actionable recommendations and forward-looking insights for future research in sports promotion. This aligns with the goals of Sports Marketing 5.0 for the Alpha Generation.