Alibaba DAMO Academy, the global research initiative by Alibaba Group, announced it has made the source code of its latest federated learning platform FederatedScope accessible to the open-source community. FederatedScope is a comprehensive learning platform with easy-to-use packages.
With the rise of machine learning in the digital era, gathering training data to build advanced AI models is challenging. The process could pose potential privacy concerns. To address such a challenge, federated learning – a way of privacy-preserving computation – has emerged. By coordinating the training of micro-tasks across different end devices, intermediate training results – rather than raw user data – is fed back to the cloud server to alleviate privacy concerns. Yet it still enables data analytics and machine learning tasks across end devices.
“By sharing our self-developed federated learning technologies with the open-source community, we hope to promote the research and industrial deployment of privacy-preserving computation in different sectors. This includes healthcare and smart mobility that usually involves sensitive user data. It also requires strict privacy protection practices,” said Bolin Ding, Research Scientist at Alibaba DAMO Academy.
Providing flexible support and comprehensive tools
In addition, with a newly-implemented event-driven framework, FederatedScope provides flexible support and comprehensive tools. It includes a collection of benchmark datasets, well-known model architectures, advanced federated learning algorithms, easy-to-use automatic tuning functionalities and friendly interfaces. These enable researchers and developers to quickly build and customise task-specific federated learning applications in areas. These include computer vision, natural language processing, speech recognition, graph learning and recommendation.
For privacy protection in particular, the platform also offers differential privacy and multi-party computation to meet different requirements of privacy protection. “We believe privacy-preserving computation is an important and essential trend,” added Ding. “Training AI models without compromising privacy is critical. That’s why we have devoted a lot of resources to drive the research of federated learning. We hope that by sharing our source codes and technology platform, we can support global developers in the community. Furthermore, encourage more innovation in this emerging field.”
According to Gartner, 60% of large organisations are expected to use one or more privacy-enhancing computation techniques by 2025. Earlier this year, Alibaba DAMO Academy also revealed its forecast of the leading trends that would shape the tech industry in the years ahead, and privacy-preserving computation is one of the top 10 trends. According to the forecast, in the next three years, we are expecting to “witness groundbreaking improvements in the performance and interpretability of privacy-preserving computation.”
Enterprise Times: What this means for business.
Last month, the EU agreed to a new Digital Services Act (DSA). The act sets out a new standard for the accountability of online platforms with enhanced privacy requirements. Enterprises have to deal with maturing international privacy and data protection legislation. In addition, avoid any loss of customer trust resulting from privacy incidents. Therefore, Gartner expects 60% of large organizations to use one or more privacy-enhancing computation techniques by 2025.
This explains why Alibaba is going down the open-source code route to share learning via its federated learning platform FederatedScope. Privacy will remain a major issue for businesses in the foreseeable future. There’s also a skills shortage of the data scientists required by companies such as Alibaba to develop the needed automated tools.
By going the open-source code route, Alibaba hopes to stimulate the promotion of research and deployment of privacy-preserving computation. Such tools will be desperately needed by enterprises to ensure they remain on the right side of these new laws.