Privacy Preserving Nearest Neighbor Search

Shaneck, Mark; Kim, Yong‐Dae; Kumar, Vipin

doi:10.1007/978-0-387-88735-7_10

Cited by 40 publications

(36 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A two-party protocol for finding k-Nearest Neighbors was given in [SKK06], and improved from quadratic to linear communication complexity in [QA08]. Our protocol for finding the nearest neighbor is a more efficient protocol for the special case k = 1.…”

Section: Introductionmentioning

confidence: 99%

Improved Garbled Circuit Building Blocks and Applications to Auctions and Computing Minima

Kolesnikov¹,

Sadeghi

Schneider

2009

Cryptology and Network Security

195

183

View full text Add to dashboard Cite

Abstract. We consider generic Garbled Circuit (GC)-based techniques for Secure Function Evaluation (SFE) in the semi-honest model. We describe efficient GC constructions for addition, subtraction, multiplication, and comparison functions. Our circuits for subtraction and comparison are approximately two times smaller (in terms of garbled tables) than previous constructions. This implies corresponding computation and communication improvements in SFE of functions using our efficient building blocks. The techniques rely on recently proposed "free XOR" GC technique. Further, we present concrete and detailed improved GC protocols for the problem of secure integer comparison, and related problems of auctions, minimum selection, and minimal distance. Performance improvement comes both from building on our efficient basic blocks and several problemspecific GC optimizations. We provide precise cost evaluation of our constructions, which serves as a baseline for future protocols.

show abstract

Section: Introductionmentioning

confidence: 99%

Improved Garbled Circuit Building Blocks and Applications to Auctions and Computing Minima

Kolesnikov¹,

Sadeghi

Schneider

2009

Cryptology and Network Security

195

183

View full text Add to dashboard Cite

show abstract

“…With the development of data mining and provide the various methods for privacy preserving, [8][9][10][11][12][13][14][15][16] Advantages and disadvantages of these methods and how to implement them better has been much discussed. Most of the proposed methods based on perturbation, randomization or anonymity.…”

Section: Related Workmentioning

confidence: 99%

“…In papers [12,15,16] be seen some methods with this approach that how do these algorithms is such that try distribute the calculations between data servers and finally get the results of calculation instead of data. Major disadvantage of this approach is that many of these algorithms have to disclosure new instances.…”

Section: Related Workmentioning

confidence: 99%

A Novel Algorithm for Privacy Preserving Distributed Data Mining

Molaei¹,

Jalali²,

Vadiatizadeh³

2013

IJCA

View full text Add to dashboard Cite

With the development of data mining science and information technology, distributed data mining was considered. Distributed data mining was submitted with different purposes, such as increasing the accuracy of result and using data from multiple sources. With the development of distributed data mining, challenges in this field were introduced soon. The main challenge was the issue of privacy preserving. In the past few years, solutions for this problem have been proposed but each one has had a weakness.Given the importance of knowledge, today's organizations need to Distributed Data Mining, Our goal in this article is to provide an approach that able to preserve the privacy in distributed mining. Our approach can be implemented on most of the algorithms. Proposed approach has been useed data encryption and some of techniques that used in security. General TermsPrivacy preserving distributed data mining algorithm.

show abstract

“…Related studies have been published focusing on not only simple analyses such as database queries with very specific inclusion/exclusion criteria but also sophisticated algorithms for prediction analysis including logistic regression [13,14], support vector machine (SVM) [15,16], knearest neighborhood [17], Cox regression [18], and tensor factorization [19]. However, most studies involve restrictive assumptions originating from the requirement that data should be integrated in a matrix format, either common feature events assumption for horizontally-partitioned data or common patient records assumption for verticallypartitioned data.…”

Section: Introductionmentioning

confidence: 99%

Privacy-Preserving Predictive Modeling: Harmonization of Contextual Embeddings From Different Sources (Preprint)

Huang¹,

Lee²,

Wang³

et al. 2017

Preprint

View full text Add to dashboard Cite

Background: Data sharing has been a big challenge in biomedical informatics due to privacy concerns. Contextual embedding models have demonstrated very strong representative capability to describe medical concepts (and their context), and they have shown promise as an alternative way to support deep learning applications without the need to disclose original data. However, contextual embedding models acquired from individual hospitals cannot be directly combined because their embedding spaces are different and naive pooling renders combined embeddings useless.Objective: We present a novel approach to address these issues to promote sharing representation without sharing data. We can build a global model from representations learned from local private data without sacrificing privacy and synchronize information from multiple sources. Methods:We propose a methodology that harmonizes different local contextual embeddings into a global model. We use Word2Vec to generate contextual embeddings from each source and Procrustes to fuse different vector models into one common space by using a list of corresponding pairs as anchor points. With harmonized embeddings, we performed prediction analysis. Results:We used sequential medical events extracted from the Medical Information Mart for Intensive Care III database to evaluate the proposed methodology in predicting the next likely diagnosis of a new patient using either structured data or unstructured data. Under different experimental scenarios, we confirmed that the global model built from harmonized local models achieves more accurate prediction than local models and global model built from naive pooling.Conclusions: Such aggregation of local models using our unique harmonization can serve as the proxy for a global model, combining information from a wide range of institutions and information sources. It allows information unique to a certain hospital to become available to other sites, increasing the fluidity of information flow in health care.

show abstract

Privacy Preserving Nearest Neighbor Search

Cited by 40 publications

References 40 publications

Improved Garbled Circuit Building Blocks and Applications to Auctions and Computing Minima

Improved Garbled Circuit Building Blocks and Applications to Auctions and Computing Minima

A Novel Algorithm for Privacy Preserving Distributed Data Mining

Privacy-Preserving Predictive Modeling: Harmonization of Contextual Embeddings From Different Sources (Preprint)

Contact Info

Product

Resources

About