Tencent and WPP are joining forces to help Chinese advertisers make the most of data-driven and artificial intelligence.
They are coming together to ensure there won’t be any data silos or complicated data sharing practices, despite the tighter data privacy protection regulations.
WPP and Tencent Cloud Big Data are already running campaigns of multiparty federated learning using the Cloud Security Privacy Computing Platform.
Federated Learning is an open-source machine-learning framework that dismantles data silos to unlock the full potential for AI. Based on the principle that no underlying data should be disclosed, different parties can perform joint modeling by exchanging encrypted intermediate result and data enrichment legally.
“At the beginning of this year, we upgraded Shen Dun Federated Computing to Tencent Cloud Secure Privacy Computing, which is based on the Tencent Angel PowerFL privacy computing framework, with the protection technology of private data such as Federated Learning (FL), Secure Multiparty Computation (MPC), and Trusted Execution Environment (TEE),” said Naruto Guo, head of Tencent Cloud’s security privacy computing product.
“Customized privacy protection transformation is carried out for algorithms such as machine learning and data analysis to create privacy computing products with full links of data applications. The upgraded product allows for fast completion of privacy computing tasks while the underlying data stays local. While ensuring data security, it can also maximize the value of data and solve the problem of data silos faced by enterprises.”
He adds: “Cooperation with WPP is a very good starting point in the field of advertising and marketing. This advertising cooperation is possible thanks to the platform’s excellent security and performance. It ensures data security and maximizes the enterprise data’s value. The follow-up promotion and application also require joint efforts of all partners in the industry, and we look forward to creating more outstanding cases in the future.”
How will federated Learning help advertisers?
Pernod Ricard’s implementation process required multiple parties to input their own audience profiles and device IDs in the respective modeling environments. These are based upon the high-quality seed audiences for its brands Martell, and the Glenlivet.
After completing the model training, all parties perform federated scoring and reasoning for devices based on the platform’s hidden query function.
Federated learning’s core principle is to not disclose the underlying data. All parties participated in joint modeling by exchanging encrypted intermediate result and finally generating the exact model expected by Martell, The Glenlivet, through feature engineering and algorithm tuning.
The model then applies the audience group score and selection to the activation process for the mainstream publishers in this market.
The results showed that federated learning is significantly more effective for long-duration campaigns for brands with large amounts of CRM historical data, such as Martell.
The Glenlivet is a relatively new brand. The model generated by federated Learning using limited CRM data has attained high-quality exposure. It is twice as good as the control group in short-duration campaign campaigns.