Our project called "Constructing an infrastructure for secure data sharing and distributing, accelrating integrated bigdata utilization" is 5.5-year national project, which started in 2015. This is a sub-project of "Advanced Core Technologies for Big Data Integration" run by JST CREST.
An infrastructure where providers can trust data on, users can get results from and make use of them trustfully, is required today. We do not relay on "Anonimization" or "Encryption during transmission", and aim to construct an open-source library for data mining which handles data on an encrypted state. We are optimizing it from the aspects of both cryptology and computer architecture, making it more than 1000 times faster than today's scheme.
FCMalloc: A Fast Memory Allocator for Fully Homomorphic Encryption
MCMalloc: A Scalable Memory Allocator for Multithreaded Applications on a Many-Core Shared-Memory Machine
lbsRecommend: Privacy-Preserving Recommendation for Location-Based Service
OPSI-CA: Outsourced Private Set Intersection Cardinality with Fully Homomorphic Encryption
NB-Classify: Naive Bayes with Fully Homomorphic Encryption
blkctrl: Block Control for Ext2/3/4
prv_cmp: Private Comparison library
FHE-Table-Search: FHE-Table-Search function evaluation of fully homomorphic encryption using table lookup
Subring-HE/FV: The decomposition ring based homomorphic encryption scheme
PP-CNN: Privacy Preserving CNN Inference over Homomorphic Encryption
DAMCREM: Dynamic Allocation Method of Computation REsource to Macro-Tasks for Fully Homomorphic Encryption Applications
BRPS for FHE: Bootstrapping and Relinearization Problem Solver: Solvers for Fully Homomorphic Encryption