Metacosm scene-oriented data privacy protection method and system, and storage medium

Abstract

The invention discloses a meta universe scene-oriented data privacy protection method and system, and a storage medium. The method comprises the following steps: uploading non-private data to a data storage module in a virtual world as a first local private data set to train a first local model; private data with sensitive information is stored locally and serves as a second local private data set to train a second local model; constructing a meta-universe cross-chain federated machine learning framework, wherein the meta-universe cross-chain federated machine learning framework comprises a task publisher, a first client for training a first local model through a first local private data set, and a second client for training a second local model through a second local private data set; according to the task publisher, a first client is arranged in a virtual world, and a second client is arranged in a real world; and the first client and the second client update and aggregate the first local model parameter and the second local model parameter based on a cross-chain aggregation method to obtain a new global model.

Description

technical field [0001] The invention relates to the technical field of communication signal demodulation, and more particularly, to a data privacy protection method, system and storage medium oriented to a metaverse scenario. Background technique [0002] Federated learning (federated learning) is an emerging artificial intelligence technology. Its design goals are to ensure information security when big data is shared, protect terminal data and personal data privacy; Efficient machine learning model training between multiple computing nodes. Federated learning realizes "data availability is invisible" and "data does not go out" through iterative training operations. In the process of model training, the interactive data is encrypted and operated, which achieves efficient model training to a certain extent and protects participation to the greatest extent. It can solve the data privacy problem caused by the need to access data for traditional machine learning models. At the...

AI Technical Summary

Problems solved by technology

[0006] The disadvantages of current existing technologies are as follows: the AI ​​model in the metaverse needs to be trained with data generated in the real world and the virtual world to obtain better model performance, but if all local private data is uploaded to the virtual world for training, On the one hand, uploading real world data to the virtual world storage module will have a large communication overhead and storage overhead, on the other hand, it will also lead to the risk of privacy leakage and data abuse of private data

Images