Abstract: The promulgation of Federated Learning standardization means that the technology of Federated Learning will towards more mature, standardized and industrialized direction, laying a foundation for all walks of life to jointly build the Federated ecosystem.

 

On June 28th, co-sponsored by the Ministry of Industry and Information Technology of the People’s Republic of China and the Beijing municipal government, the 23rd China International Software Expo is launching in Beijing Exhibition Centre. On the “artificial intelligence open-source forum” in the next day, China Artificial Intelligence Open Source Software Development Alliance (AIOSS) released four group standards including《Information Technology service – Federated Learning – Reference Architecture》and《Application Case Collection of Chinese Artificial Intelligence Open Source Software》.

AIOSS is an organization that gathers industry, academia and research institutes to jointly promote the development of artificial intelligence open-source software in China, established on July 1, 2018, by China Electronics Standardization Institute, with the support of Information and Software Services Division of Ministry of Industry and Information Technology of People’s Republic of China.

This means that “Federated Learning”, an emerging artificial intelligence technology, is further concerned by the society. Its core is to solve the issue of “Data Silos” and data privacy protection in the application of artificial intelligence industry. Artificial intelligence can’t survive without big data. However, most industries are now faced with small data, data fragmentation, data silos, which has become a bottleneck restricting the development of artificial intelligence. How to promote shared development while avoiding the monopoly of data and platform and respecting and protecting personal privacy is a major problem facing the industrialization of artificial intelligence. Federated Learning is the key technology to solve this industry problem. The advantage of Federated Learning is that it can ensure that the data of the parties involved in local and independent to achieve AI collaboration. Federated Learning enables enterprises and institutions to meet the needs of user privacy protection and data security and jointly improve the level of AI application.

It is reported that the standard formulation is led by WeBank. As the first advocate of Federated Learning in China, WeBank is also leading the efforts to promote the formulation of IEEE international standards for Federated Learning. The development of federal learning standardization will provide technical specifications for the implementation of federal learning and help the community to build a federal ecological basis for cooperation. Professor. Yang Qiang, international expert in artificial intelligence, Chief AI Officer of WeBank, they jointly released the standard at the meeting. Professor. Yang Qiang said, “the introduction of the first group standard for Federated Learning means that the technology of federal learning will become more mature, standardized and industrialized, laying a foundation for all sectors to jointly build the federated ecosystem.“

When Federated Learning becomes the tool level of industrialization, WeBank’s AI team has developed the world’s first industrial-scale Federated Learning technology framework Federated AI Technology Enabler (FATE) which is open-source. Not only does it offer a set of out-of-the-box federated learning algorithms, but it also provides developers with a template for implementing Federated Learning algorithms and systems, so that most traditional algorithms can be adapted to the Federated Learning framework, thus quickly joining the federated ecosystem. Recently, WeBank donated FATE to the Linux Foundation, a well-known non-profit organization in the world, to reach out to the open-source community and further contribute the federal learning technology research results to developers around the world. Recently, WeBank donated FATE to the global famous non-profit organization, Linux foundation to reach out to the open-source community and further contribute the Federated Learning technology research results to developers around the world.

At present, China’s economy is transitioning from a phase of rapid growth to a stage of high-quality development, is in the transformation of the mode of development, optimizing the economic structure, transform growth research period. In the key nodes, AI will play an important role in how to protect the privacy of data on the premise of achieving wider cooperation, cross-border integration, open group of wisdom, is the era of challenge. We look forward to seeing federal learning become the next driver of AI industrialization, bringing about profound changes in the “AI + Industry.”