Secure and Private Zero Knowledge Data Collaboration Spaces

该项目旨在提供安全和保密的零知识数据协作空间作为服务。同态加密（HE）允许在不解密的情况下处理加密数据，甚至不知道解密密钥。这使得数据所有者能够在不泄露数据内容的情况下提供对私人或机密数据的访问，从而消除潜在合作的现有障碍。该项目将提供安全的协作空间作为一种服务，这将消除协作项目的许多技术障碍，使广泛的参与者能够专注于数据科学。直接在加密数据上进行计算的能力允许共享机密数据，同时保持符合机密性要求，例如GDPR。例子包括使用客户数据来制定营销策略；使用临床数据作为公共或学术基因组分析项目的一部分；或者与合作伙伴公司共享机密实验数据，例如减少制药公司对体内试验的需求；银行可以共享真实世界的客户交易数据，以便训练准确的欺诈检测模型。因此，它允许在公共和不太可信的环境中进行协作，而不会泄露任何数据或任何计算结果；所有内容都是加密的，只有数据或结果的所有者才能看到它。这是一个相对较新的发展，很少有其他商业选择提供这种能力。这也将促进和支持IT/软件或服务提供商的增长，他们依赖于真实数据来开发他们的产品。例如，它将使软件提供商能够开展基于机器学习（ML）的项目，以更有效地检测健康状况，但由于隐私和安全问题，他们需要访问现实世界的临床试验或基因组数据，这些数据目前无法访问。

This project aims to provide secure and confidential zero-knowledge data collaboration spaces as a service.Homomorphic Encryption (HE) enables processing on encrypted data without decrypting it, even without knowing the decryption key. This enables data owners to provide access to private or confidential data without disclosing the content of the data thereby removing existing barriers to potential collaborations. This project would provide secure collaboration spaces as a service which would remove many of these technical barriers to collaboration projects by, enabling a wide range of participants to focus on the data science.The ability to compute directly on encrypted data enables sharing of confidential data, while remaining compliant to confidentiality requirements, such as GDPR. Examples include use of customer data to derive marketing strategies; to use clinical data as part of public or academic genomic analysis projects; or to share confidential experimental data with partner companies, for example to reduce the need for in-vivo testing by pharmaceutical companies; banks could share real-world customer transaction data to enable training of accurate fraud detection models. Therefore, it allows for collaboration in public and less trusted environments, without leaking any of the data or any of the results of the computation; everything remains encrypted so that only the owner of the data or results can see it. This is a relatively new development, there are very few alternate commercial options providing this capability.This would also enable and support growth of IT/Software or service providers who are dependent on real data to develop their products. For example, it would enable software providers to work on machine learning (ML)-based projects to detect health conditions more efficiently, but who need access to real-world clinical trial or genomic data that is currently not accessible, due to privacy and security concerns.