Project Deliverables
Work Package 2 – Requirements Analysis
D2.1 describes the results of the Anonymous Big Data workshop, the nonconfidential part of the project partners’ use cases together with the accuracy and privacy requirements.
D2.2 includes details on quantitative measures for accuracy and privacy of the generated synthetic data with respect to the original data.
This document is a supplement to D2.1 and includes use cases that were collected in the context of the master’s thesis “A new approach for marketing analytics in an increasing environment of data regulation – synthetic data.”
Work Package 3 – Generative DNN Architectures
D3.2 reviews published privacy-preserving techniques for deep learning.
D3.3 provides details about the architectures that will be implemented in our virtual data lab.
Work Package 4 – Virtual Data Lab (VDL)
Work Package 5 – Model Development
Work Package 6 – Empirical Validation
Papers
Vamosi S., Platzer M. and Reutterer T. (2022). AI-based Re-identification of Behavioral Clickstream Data. arXiv preprint arXiv:2201.10351
Vamosi S., Reutterer T., and Platzer M. (2022). A deep recurrent neural network approach to learn sequence similarities for user-identification. Decision Support Systems, 113718.
Eigenschink P., Vamosi S., Vamosi R., Sun Ch., Reutterer T. and Kalcher, K. (2021). Deep Generative Models for Synthetic Data. WU Working/discussion paper.
Platzer M. and Reutterer T (2021) Holdout-Based Empirical Assessment of Mixed-Type Synthetic Data. Front. Big Data 4:679939. doi: 10.3389/fdata.2021.679939
Talks
AI-based re-identification of behavioral data – presentation at European Marketing Academy Conference (EMAC) 2021