Shen Yan, a professor at Peking University National Development Council: "Big Data Killing" should promote algorithm audit

In the digital era, data has become an important production factors, and its role in improving production efficiency is increasingly prominent. The "Digital China Development Report (2021)" released a few days ago shows that from 2017-2021, my country's digital economy has ranked second in the world for many years, accounting for 9.9%of the world. China's data volume is huge and rapidly growing. From the development of the platform economy to data analysis and governance, they are facing challenges.

"The general direction of platform data governance is to promote innovation on the basis of protecting the rights and interests of consumers and other data subjects, and maintain international competitiveness." On July 31st, Shen, a professor at Peking University and Deputy Director of Peking University Digital Financial Research Center, Shen The launch of the "Platform Economic Innovation and Governance" seminar and "Platform Economy: Innovation, Governance and Prosperity" in the launch conference of the launch conference was directly at the site. How does intelligent and artificial intelligence interact for data decision -making? How is data security and privacy protection? "

As an important production factor, Shen Yan compared the data circulation to "dancing with shackles". In her opinion, there are series of laws and regulations in China to protect data security and personal privacy. The circulation of data elements also needs to follow relevant laws and regulations. "In fact, China's relevant regulations are comparable to that of Europe." Shen Yan pointed out.

How to promote the circulation of data production factors? Shen Yan believes that different from the traditional factor transaction "first confirmation and then circulation", data transactions should "focus on cultivating data -based data service transactions, promote data services based on data value, and promote the primitive primitive based on clear data ownership. Data transaction.

In addition, Shen Yan suggested to promote public data sharing, and use the "classification of data licenses may permit" to solve the marginal issue of data circulation, and it is also convenient for data to "accountability".

With the development and application of data technology, data analysis brings convenience to decision -making, and there are also some "hidden" problems, such as big data to kill and hidden algorithm discrimination. These "discrimination" from artificial intelligence are not so easy to be found, and new requirements have also been made for supervision.

In this regard, Shen Yan suggested to promote algorithm audits. "Algorithm audit is not a value judgment, but to increase the transparency of rules." For example, when using platform software, let the regulators, consumers, platform parties, and services be understood at this time. Asymmetric, so as to adjust the rights and interests of all parties, and be relatively fair.

"Pay attention to the evaluation of input and output and results, not the public algorithm, and make the code transparent." Shen Yan briefly explained the path of the algorithm audit. From the content of the content, you can audit the "prediction or optimization goals", "data description", "algorithm technology and reasons", "algorithm operation effect", "specific arrangements for personal information protection and data security" It can include discrimination, effectiveness, transparency, direct impact, security and acquisition.

Shen Yan specifically mentioned that in the process of application and development of the platform, the decision -making matching of algorithms should be made. "".

At the moment, data security and privacy protection have attracted much attention from personal perspective and national level. In terms of technology, there are currently related discussions such as dense computing, de -identification and desensitization technology.

How to further promote data security and privacy protection? In Shen Yan's opinion, the system construction of more than technology, privacy protection and privacy computing is urgent, such as the process of internal design privacy protection mechanisms for enterprises, and regulatory departments to set up regulatory supporting measures. She emphasized that data security and privacy protection should focus on technological development rather than administrative means to solve problems.

In the end, Shen Yan suggested that whether it is the control of data production factor, algorithm management or privacy, and data security protection, it is necessary to protect consumers and should take into account the rights and interests of other subjects, strengthen the transparency of information on both ends of the platform, and reduce information asymmetry. At the same time, we should also pay attention to mobilizing the enthusiasm of the platform to innovate data governance and give full play to industry self -discipline.

Daily Economic News

- END -