Towards Efficient Target-Level Machine Unlearning Based on Essential Graph

TL;DR

This paper introduces a novel target unlearning method that efficiently removes specific partial targets from trained models by leveraging an essential graph structure, improving scalability and effectiveness over existing approaches.

Contribution

It proposes an innovative target unlearning scheme using an essential graph to selectively remove partial target information, addressing limitations of previous instance-level unlearning methods.

Findings

Effective removal of partial targets demonstrated across multiple datasets.

Significant reduction in unlearning time compared to baseline methods.

Maintains model performance while erasing targeted information.

Abstract

Machine unlearning is an emerging technology that has come to attract widespread attention. A number of factors, including regulations and laws, privacy, and usability concerns, have resulted in this need to allow a trained model to forget some of its training data. Existing studies of machine unlearning mainly focus on unlearning requests that forget a cluster of instances or all instances from one class. While these approaches are effective in removing instances, they do not scale to scenarios where partial targets within an instance need to be forgotten. For example, one would like to only unlearn a person from all instances that simultaneously contain the person and other targets. Directly migrating instance-level unlearning to target-level unlearning will reduce the performance of the model after the unlearning process, or fail to erase information completely. To address these concerns, we have proposed a more effective and efficient unlearning scheme that focuses on removing partial targets from the model, which we name "target unlearning". Specifically, we first construct an essential graph data structure to describe the relationships between all important parameters that are selected based on the model explanation method. After that, we simultaneously filter parameters that are also important for the remaining targets and use the pruning-based unlearning method, which is a simple but effective solution to remove information about the target that needs to be forgotten. Experiments with different training models on various datasets demonstrate the effectiveness of the proposed approach.