Unlock instant, AI-driven research and patent intelligence for your innovation.

Key-value pair model safety training and reasoning method based on safety multi-party calculation

A technology of secure multi-party computing and reasoning methods, which is applied in the field of security training and reasoning of key-value pair models based on secure multi-party computing, and can solve problems such as internal data not being open to the outside world and data value not being fully reflected

Active Publication Date: 2021-10-22
FUDAN UNIV
View PDF3 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, at present, most enterprises and organizations consider data security and personal privacy issues, and are very cautious about data sharing. Even internal data is not open to the outside world, resulting in isolated data islands, resulting in the inability to obtain the value of data. full expression

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Key-value pair model safety training and reasoning method based on safety multi-party calculation
  • Key-value pair model safety training and reasoning method based on safety multi-party calculation
  • Key-value pair model safety training and reasoning method based on safety multi-party calculation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0055] The present invention will be described in detail below in conjunction with the accompanying drawings.

[0056] exist figure 1 Above, participants 1~n want to jointly train or predict a key-value pair model, but they do not want other participants to know the data they own;

[0057] (1) According to the required model algorithm, construct an optimized calculation flow graph, and hand over the model parameters to all participants in the form of secret sharing;

[0058] (2) The participant performs secret sharing of its own local key-value pair data and transmits it to other participants. So far, the entire model and input data are held among all participants in the form of secret sharing;

[0059] (3) All participants cooperate to perform calculations according to the calculation flow graph model, in which basic mathematical calculation operators, comparison operators, and GET query operators are called. The specific operation process of the three models of calculatio...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention belongs to the technical field of network space safety, and particularly relates to a key-value pair model safety training and reasoning method based on safety multi-party calculation. The method comprises the basic steps of abstracting calculation in training and reasoning of common key-value pair models such as a decision tree model and a basic mathematical statistical model, and designing a unified key-value pair data query and calculation primitive; and designing a safe and efficient calculation mechanism for the training and reasoning algorithm of the model, constructing a corresponding calculation flow graph, and carrying out driving calculation according to the topological sorting result of the calculation flow graph in combination with the key value pair primitives. The method has the advantages that a unified key-value pair query and calculation protocol is abstracted, and any customized key-value pair model algorithm is supported; the optimized training and reasoning mechanism has higher operation efficiency; the design implementation is performed in combination with a secret sharing protocol, and the privacy safety of input data is ensured.

Description

technical field [0001] The invention belongs to the technical field of cyberspace security, and in particular relates to a key-value pair model security training and reasoning method based on secure multi-party computing. Background technique [0002] With the rapid development of Internet technology and information technology, people have gradually entered the era of big data. Data is an important resource in this era. How to reasonably and fully utilize its value is an important issue in today's era. Generally, efficient and secure data sharing can provide producers (data users) with large amounts of timely high-value data, effectively supporting social production and life. Therefore, it is the key core technology in the current era of data as a factor of generation, and it needs to be studied in depth. Break through its technical bottleneck. However, at present, most enterprises and organizations consider data security and personal privacy issues, and are very cautious a...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F16/2458G06F17/18G06F21/62G06K9/62
CPCG06F16/2462G06F17/18G06F21/6245G06F18/24323
Inventor 韩伟力吴豪奇姜子丰汪家璇
Owner FUDAN UNIV