This article has been reviewed according to Science X's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility:

fact-checked

proofread

Study addresses privacy-preserving collaborative data collection and analysis with many missing values

Privacy-preserving collaborative data collection and analysis with many missing values
Example application of the proposed privacy-preserving data collection method. Credit: IEEE Transactions on Dependable and Secure Computing (2022). DOI: 10.1109/TDSC.2022.3174887

To control pandemics like the novel coronavirus infection (COVID-19), data such as the age, gender, family composition, and medical history of infected individuals are required. While patients themselves may provide this information to medical institutions, these details are highly confidential.

If the data is properly handled for , it can be shared with researchers worldwide without identifying the infected individual, which can help clarify the state of the pandemic and more accurately predict its progression.

There may be missing values in the information provided by patients, and existing methods do not take these missing values into account when collecting personal data while ensuring privacy. This is leading to a significant reduction in the accuracy of data analysis.

Differential Privacy, the privacy protection metric addressed in this paper, is adopted by many organizations, including Apple, Google, Microsoft, and LINE. Numerous methods have been proposed to collect and analyze based on Differential Privacy. However, none of the existing methods take into account the presence of missing values.

When considering medical data, as during the COVID-19 pandemic, it is conceivable that different hospitals can obtain different information, and many patients may feel comfortable providing only some data after privacy protection processing. Under the current methodology, the accuracy of analysis is greatly reduced in such scenarios, which has prevented sufficient data analysis for pandemic mitigation.

Professor Sei has demonstrated that using the Copula model, primarily used in the finance field, can restore the true statistical model from data processed by Differential Privacy technology even in situations with many missing values, enabling highly accurate data analysis. Of course, he mathematically proves that each individual's is strictly protected at the exact same level as existing methods.

In real society, data typically has various missing elements. By using the proposed method, not only but also various societal and with missing values can be safely analyzed with high accuracy. Therefore, this research is expected to have a significant impact on society.

The study is published in the journal IEEE Transactions on Dependable and Secure Computing.

More information: Yuichi Sei et al, Privacy-Preserving Collaborative Data Collection and Analysis with Many Missing Values, IEEE Transactions on Dependable and Secure Computing (2022). DOI: 10.1109/TDSC.2022.3174887

Provided by The University of Electro-Communications
Citation: Study addresses privacy-preserving collaborative data collection and analysis with many missing values (2023, July 7) retrieved 23 February 2024 from https://techxplore.com/news/2023-07-privacy-preserving-collaborative-analysis-values.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.

Explore further

AI technology for safer integrated analysis of data held by multiple organizations

1 shares

Feedback to editors