The Possibility of Matrix Decomposition as Anonymization and Evaluation for Time-sequence Data

Mimoto, Tomoaki; Kiyomoto, Shinsaku; Hidano, Seira; Basu, Anirban; Miyaji, Atsuko

doi:10.1109/pst.2018.8514189

Cited by 8 publications

(3 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Chen et al experimentally evaluated the privacy increase for such attackers by adding noise to the dimension-reduced data. A success rate of the Re-identification attack [19] was used for the privacy evaluation. This attack links corresponding records x i ∈ X and x i ∈ X for the original data X and anonymized data X .…”

Section: B Privacy Of Data Collaborationmentioning

confidence: 99%

Toward the Application of Differential Privacy to Data Collaboration

Yamashiro,

Omote,

Imakura

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Federated Learning, a model-sharing method, and Data Collaboration, a non-model-sharing method, are recognized as data analysis methods for distributed data. In Federated Learning, clients send only the parameters of a machine learning model to the central server. In Data Collaboration, clients send data that has undergone irreversibly transformed through dimensionality reduction to the central server. Both methods are designed with privacy concerns, but privacy is not guaranteed. Differential Privacy, a theoretical and quantitative privacy criterion, has been applied to Federated Learning to achieve rigorous privacy preservation. In this paper, we introduce a novel method using PCA as a dimensionality reduction method, aiming to apply Differential Privacy to Data Collaboration. Experimental evaluation using the proposed method show that differentially-private Data Collaboration achieves comparable performance to differentially-private Federated Learning.

show abstract

Section: B Privacy Of Data Collaborationmentioning

confidence: 99%

Toward the Application of Differential Privacy to Data Collaboration

Yamashiro,

Omote,

Imakura

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…3 to address the privacy and evaluate the effect in Sect. 5. This paper is expanded from [7]. The matrix factorization algorithm is changed and stochastic gradient descent (SGD) based matrix factorization, which is widely used in many fields such as recommender system, is applied.…”

Section: Privacy Definition Against An Authorized Usermentioning

confidence: 99%

“…Rank r can be treated as the parameter of anonymization by matrix factorization because the accuracy of dataset X = UV T depends on rank r, so that r is the parameter of our algorithm and we set r = 10, 20, 30, 40. We set larger values in the experiments in [7] but the results of the case r > 40 are saturated. The probabilities of re-identification and linkage attack are shown in Table 3.…”

Section: Effects Of Matrix Factorization Itselfmentioning

confidence: 99%

Anonymization Technique Based on SGD Matrix Factorization

Mimoto

Hidano

Kiyomoto

et al. 2020

IEICE Trans. Inf. & Syst.

Self Cite

View full text Add to dashboard Cite

Time-sequence data is high dimensional and contains a lot of information, which can be utilized in various fields, such as insurance, finance, and advertising. Personal data including time-sequence data is converted to anonymized datasets, which need to strike a balance between both privacy and utility. In this paper, we consider low-rank matrix factorization as one of anonymization methods and evaluate its efficiency. We convert time-sequence datasets to matrices and evaluate both privacy and utility. The record IDs in time-sequence data are changed at regular intervals to reduce re-identification risk. However, since individuals tend to behave in a similar fashion over periods of time, there remains a risk of record linkage even if record IDs are different. Hence, we evaluate the reidentification and linkage risks as privacy risks of time-sequence data. Our experimental results show that matrix factorization is a viable anonymization method and it can achieve better utility than existing anonymization methods.

show abstract

Safety Analysis of High-Dimensional Anonymized Data from Multiple Perspectives

Yamazoe

Omote

2020

Network and System Security

View full text Add to dashboard Cite

The Possibility of Matrix Decomposition as Anonymization and Evaluation for Time-sequence Data

Cited by 8 publications

References 16 publications

Toward the Application of Differential Privacy to Data Collaboration

Toward the Application of Differential Privacy to Data Collaboration

Anonymization Technique Based on SGD Matrix Factorization

Safety Analysis of High-Dimensional Anonymized Data from Multiple Perspectives

Contact Info

Product

Resources

About