Method and system for dynamic collection and release of check-in data based on b-dp differential privacy mechanism

Abstract

The present invention provides a method and system for dynamically collecting and publishing sign-in data based on the B-DP differential privacy mechanism, including an initialization process, in which the disturbance probability matrix Q and the estimated distribution of sign-in data are initialized by the data acquisition end, and broadcast to the data provider; the disturbance process , the data provider generates the check-in data, and uses the probability matrix Q to disturb the check-in state; the aggregation process, the data collection end collects the check-in data, counts the check-in data after a period of time, and obtains the estimated disturbance distribution of the current time slice; In the construction process, the data acquisition end uses the inverse estimation method to obtain the original estimated distribution of the check-in data of the current time slice, and obtains the estimated distribution to be released through correction; in the release process, the data acquisition end sends the current estimated distribution to be released to The data user end; the update process is compared and judged by the data acquisition end. If the start update condition is met, a new disturbance probability matrix Q is obtained by using the B‑DP mechanism and broadcast to the data provider end.

Description

technical field [0001] The invention belongs to the field of information security, and in particular relates to a method and system for dynamically collecting and publishing sign-in data based on a B-DP differential privacy mechanism. Background technique [0002] With the rapid development of the mobile Internet, the rapid advancement of positioning technology and the widespread popularity of mobile devices, location-based service (LBS, Location Based Service) applications, including social platforms such as Foursquare, Gowalla and Facebook, generate a large amount of check-in data every day . Check-in data generally includes information such as the user's visit time, location, check-in point POI (Points Of Interest) attributes, mood, and evaluation of its visit location, so the check-in data becomes a carrier of the user's life track and interest tendency. This "user" is the so-called check-in person, and through the check-in service, he can share the POIs he visits (such...

AI Technical Summary

Problems solved by technology

[0006] (1) The differential privacy protection method that prioritizes privacy is often unacceptable in practical applications;

[0007] (2) For some differential privacy protection methods that take into account usability, the privacy protection is often very weak, or even almost no privacy at all

However, this trade-off is also difficult to avoid weakening or leaking privacy

Images