PII/PHI的认知服务 [英] Cognitive Services with PII/PHI

问题描述

我们公司正在考虑将Azure认知服务与chatbot实施结合使用.但是，人们非常担心认知服务存储的数据(用于例外情况，培训等)可能包含PII/PHI数据.例如，用户询问什么 是我们对癌症的医疗保障"没有个人信息，而问题我们为我的女儿玛丽·珍妮提供的癌症医疗覆盖范围是什么?"会.

Our company is contemplating using Azure Cognitive Services with chatbot implementation. However, there are a lot of concerns that data stored by Cognitive Services (for exceptions, training, etc) may contain PII/PHI data. For example, user asking "What is our medical coverage for cancer" does not have personal information while question "What is our medical coverage for cancer for my daughter Mary Jane" does.

认知服务是否提供从存储的信息中剥离PII/PHI元素的选项?

Do Cognitive Services provide an option to strip out PII/PHI elements from stored information? Are there discussions on similar topics that provide guidance on how other companies handle this situations?

推荐答案

我不确定您使用的是哪种认知服务产品.但是我们确实有内容适中"的提示，在我们的Bot Service V4中.请参考以下文档，其中有一个模板可以帮助您解决问题.https://docs.microsoft.com/zh-CN/azure/bot-service/bot-builder-enterprise-template-overview- detail?view = azure-bot-service-4.0#content-moderator

I am not sure which Cognitive service product you are using. But we do have "Content Moderate" in our Bot Service V4. Please refer to following document, there is a template can help you figure out your problem. https://docs.microsoft.com/en-us/azure/bot-service/bot-builder-enterprise-template-overview-detail?view=azure-bot-service-4.0#content-moderator

此致

雨桐

这篇关于PII/PHI的认知服务的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！