Developing efficient methods to remove private document data from well-trained AI models for document classification, while maintaining high performance on retained data.
