# Chase Anonymisation: Privacy-Preserving Knowledge Graphs with Logical Reasoning

arXiv: https://arxiv.org/abs/2410.12418 · HTML: https://mlsanity.com/paper/2410.12418/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 49 references that are in the map.
- First posted on arXiv 16 Oct 2024.

## Builds on (most cited first, 8 of 8)

- [Knowledge Graphs](https://mlsanity.com/paper/2003.02320/) (2020)
- [AI in Finance: Challenges, Techniques and Opportunities](https://mlsanity.com/paper/2107.09051/) (2021)
- [Knowledge Graphs: Opportunities and Challenges](https://mlsanity.com/paper/2303.13948/) (2023)
- [The Vadalog System: Datalog-based Reasoning for Knowledge Graphs](https://mlsanity.com/paper/1807.08709/) (2018)
- [Synthetic Data Applications in Finance](https://mlsanity.com/paper/2401.00081/) (2023)
- [Private Graph Data Release: A Survey](https://mlsanity.com/paper/2107.04245/) (2021)
- [SoK: Differential Privacy on Graph-Structured Data](https://mlsanity.com/paper/2203.09205/) (2022)
- [Publishing Community-Preserving Attributed Social Graphs with a Differential Privacy Guarantee](https://mlsanity.com/paper/1909.00280/) (2019)

## Built on by (most cited first, 1 of 1)

- [PrivGemo: Privacy-Preserving Dual-Tower Graph Retrieval for Empowering LLM Reasoning with Memory Augmentation](https://mlsanity.com/paper/2601.08739/) (2026)

## Abstract (arXiv)

We propose a novel framework to enable Knowledge Graphs (KGs) sharing while ensuring that information that should remain private is not directly released nor indirectly exposed via derived knowledge, maintaining at the same time the embedded knowledge of the KGs to support business downstream tasks. Our approach produces a privacy-preserving KG as an augmentation of the input one via controlled addition of nodes and edges as well as re-labeling of nodes and perturbation of weights. We introduce a novel privacy measure for KGs, which considers derived knowledge, a new utility metric that captures the business semantics we want to preserve, and propose two novel anonymisation algorithms. Our extensive experimental evaluation, with both synthetic graphs and real-world datasets, confirms the effectiveness of our approach.