𝐏𝐫𝐢𝐯𝐚𝐜𝐲 𝐏𝐫𝐞𝐬𝐞𝐫𝐯𝐚𝐭𝐢𝐨𝐧 𝐚𝐧𝐝 𝐌𝐮𝐥𝐭𝐢𝐦𝐨𝐝𝐚𝐥 𝐃𝐚𝐭𝐚 𝐅𝐞𝐝𝐞𝐫𝐚𝐭𝐢𝐨𝐧: 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡 𝐂𝐡𝐚𝐥𝐥𝐞𝐧𝐠𝐞𝐬 𝐚𝐧𝐝 𝐎𝐩𝐩𝐨𝐫𝐭𝐮𝐧𝐢𝐭𝐢𝐞𝐬 𝐟𝐫𝐨𝐦 𝐀𝐜𝐚𝐝𝐞𝐦𝐢𝐜 𝐚𝐧𝐝 𝐈𝐧𝐝𝐮𝐬𝐭𝐫𝐢𝐚𝐥 𝐏𝐞𝐫𝐬𝐩𝐞𝐜𝐭𝐢𝐯𝐞𝐬
![](https://www.fit.uet.vnu.edu.vn/wp-content/uploads/2023/04/z4270942061161_95a06850f97be171f22b90310f683da8.jpg)
![📌](https://static.xx.fbcdn.net/images/emoji.php/v9/t5/1.5/16/1f4cc.png)
![📌](https://static.xx.fbcdn.net/images/emoji.php/v9/t5/1.5/16/1f4cc.png)
![📌](https://static.xx.fbcdn.net/images/emoji.php/v9/t5/1.5/16/1f4cc.png)
![📌](https://static.xx.fbcdn.net/images/emoji.php/v9/t5/1.5/16/1f4cc.png)
𝐀𝐛𝐬𝐭𝐫𝐚𝐜𝐭: In this Talk, Dr. Xuan-Son Vu would like to present a small part of his research in privacy preserving ML and the challenges in multimodal data federation settings. User privacy is not only limited to a certain type of data but any data, regardless of their form, their modality, or their quantity. Due to this nature in privacy preservation, it is very challenging to detect user data, their privacy concern, from which, we protect user data accordingly while optimizing for data utilities. Throughout his research, Dr. Xuan-Son Vu has been addressing these problems in (1) privacy concern detection methods; (2) privacy guarantee data analysis; and (3) privacy guarantee methods for data sharing. He will try his best to relate these methods in real-world applications and discuss on how they can be addressed from industrial perspectives with real-world use cases.
𝐁𝐢𝐨: Xuan-Son Vu (Member, IEEE) received a B.S. degree in information systems from the University of Engineering and Technology, VNU, Hanoi, Vietnam in 2011 and an M.Sc. degree from Kyungpook National University, South Korea, in 2014. He earned his PhD. in computer science from Umeå University, Sweden in 2020 with a focus on privacy-preserving machine learning with big data. Before joining Umeå University, from 2015 to 2016, he was a full-time member of UKPLab, TU Darmstadt, Germany. He is currently a senior researcher at WASP Media & Language, UMU and Senior AI/ML lead at Devr INC, where he focuses on research and development of privacy enhancing tech in machine learning. His research interest has been primarily focused on privacy enhancing tech – both for free text and multimodal data to power downstream applications. Dr. Xuan-Son Vu’s awards and honours include the 3rd place for best paper awards at CICLING 2019, best student paper award and best inquisitive mind award at CICLING 2018. He also received the best paper award at the 40th Conference of the Korea Information Processing Society in 2014. He is also a co-founder in multiple projects such as AIcovidVN, AIHUB.ML, MLOpsVN.