A Privacy-Preserving Workflow Publishing Method Maintaining Critical Path Availability

Abstract

The invention discloses a privacy protection workflow publishing method for maintaining the availability of critical paths, comprising the following steps: dividing zero-time access edges, full-time access edges and partial access edges in the workflow based on the critical paths between pairs of target modules; introducing Workflow (k, ε)-critical path anonymous privacy protection model, anonymizes the critical path between target module pairs, perturbs the edge weights on the Top-k path, realizes (k, ε)-critical path anonymity while maintaining lineage The structure of the workflow graph remains unchanged; in order to solve the problem of anonymity strength loss caused by the number of paths between target modules is lower than the k value, path splitting is performed based on module decomposition, and the composite modules on the anonymous path are found and split to achieve path splitting, so that the workflow can be split. The release graph strictly satisfies (k,ε)‑critical path anonymity. While preventing critical path based privacy attacks, it also takes into account the maintenance of critical path topology availability.

Description

technical field [0001] The invention relates to a data privacy protection publishing method, which is object-oriented and is a lineage workflow, and realizes the anonymity protection of key paths in the workflow. Background technique [0002] With the deepening of data sharing applications, there is an increasingly urgent need to share and publish lineage workflows that describe the principles of data generation and evolution. The lineage workflow is usually represented in the form of a directed acyclic graph (DAG, Directed acyclic graph). The execution cost and transmission cost of . By analyzing the critical path of the workflow and the functional modules on the path, the shortest time required to complete the entire workflow can be measured, thereby controlling the time and cost. one. The shared release of the lineage workflow leads to the disclosure of the privacy of the workflow structure including the critical path structure, and then the attacker learns the key act...

AI Technical Summary

Problems solved by technology

[0004] (1) While ensuring that there are at least k candidate paths with the same length, the concealment of the true length of the path is not taken into account. If the attacker knows that a special node is located on the critical path, there is still a risk of critical path leakage;

[0005] (2) When the number of paths between target nodes is lower than the anonymity strength k, the weight perturbation method that maintains the graph structure is difficult to ensure that the leakage probability of the key path in the release graph does not exceed 1 / k

Images