摘要
隨著人工智慧技術的迅速進步，社會工作領域正面臨前所未有的數位化轉型。本研究旨在探討機器學習技術在弱勢群體風險評估與分級中的應用，並分析其對提升社會工作效率、改善服務品質及增強風險預防能力的影響。透過整合衛生福利部智慧決策平台的實證數據，以及對國際間相關人工智慧應用案例的比較分析，研究結果顯示，機器學習演算法能有效識別高風險個案，並將風險評估的準確性提升至85%以上，同時顯著減輕社工人員的行政負擔。然而，技術的應用亦引發了倫理考量、資料隱私保護及專業判斷與自動化決策之間的平衡等挑戰。本研究建議建立完善的人工智慧治理框架，強化跨領域合作機制，並持續優化演算法的公平性與可解釋性，以實現社會工作智慧化轉型的可持續發展目標。
關鍵詞：人工智慧、機器學習、社會工作、風險評估、弱勢群體
Abstract
The rapid advancement of artificial intelligence technology is catalyzing a significant digital transformation within the field of social work. This research investigates the implementation of machine learning technology in risk assessment and classification for vulnerable populations, examining its impact on efficiency, service quality, and risk prevention capabilities. Integrating empirical data from the smart decision-making platform of Taiwan’s Ministry of Health and Welfare with a comparative analysis of international case studies, the study reveals that machine learning algorithms can effectively identify high-risk cases. The results demonstrate an accuracy rate exceeding 85%, significantly alleviating the administrative workload of social workers. Nonetheless, the deployment of such technology presents challenges, particularly regarding ethics, data privacy, and the balance between professional judgment and automated decision-making. Consequently, this study advocates for a comprehensive AI governance framework, enhanced cross-disciplinary collaboration, and the continuous optimization of algorithmic fairness to ensure the ethical and sustainable transformation of social work practice.
Keywords: Artificial intelligence, Machine learning, social work, Risk assessment, Vulnerable populations