Data Sensitivity and Domain Specificity in Reuse of Machine Learning Applications
研究了机器学习应用复用时面临的数据敏感性和领域特异性挑战，识别出四种基于不同复用策略的应用类型，并总结了开发部署的教训。
Abstract Data sensitivity and domain specificity challenges arise in reuse of machine learning applications. We identify four types of machine learning applications based on different reuse strategies: generic, distinctive, selective, and exclusive. We conclude with lessons for developing and deploying machine learning applications.