We shall realize a society that can improve sales performance and expand customer experience safely and securely by enabling cross-domain marketing using Artificial Intelligence (AI) while preserving privacy.

Purpose of Establishment

In services such as e-commerce and onsite shopping, user behavior data are utilized to extract behavior patterns and to realize marketing that is adapted to the individual user. However, item recommendation across services and companies requires user ID exchange between services and companies, which has difficulty in terms of protecting privacy and interests.


Therefore, in order to conduct marketing that connects services and companies in different domains while protecting privacy and interests, we will develop a secure and easy-to-use “persona marketing AI“, which exchanges personas (*1) generated by each service instead of ID exchange. We will continue R&D and demonstrative experiments for realizing “persona marketing AI” with support from JST CREST (*2).


Takahiro Hara

(Consortium Chairman, A Research Director of JST CREST “Artificial Intelligence”, Professor of Osaka University)


(*1) A persona is a virtual user type extracted from user behavior data using AI. 
(*2)This R&D and demonstrative experiments are promoted under the research subject “Persona Model Transfer Techniques for Activity Prediction of Users in Heterogeneous Domains“, which has been accepted in the public project entitled “Development and Integration of Artificial Intelligence Technologies for Innovation Acceleration (Artificial Intelligence)” conducted by the Core Research for Evolutional Science and Technology (CREST) of the Japan Science and Technology Agency (JST).