Multiparty cooperative filtering method and system

Abstract

The invention provides a multidata cooperative filtering method which comprises the following steps of: on the basis of a K-near neighbor set of each first parameter released by another participant participating in the multiparty cooperative filtration, obtaining the K-near neighbor set of each first parameter which is identical to each released first parameter by calculating the metric value of neighboring according to the owned data; aiming at each first parameter, determining the intersection of the released K-near neighbor set and the obtained K-near neighbor set; selecting a first parameter of a first quantity from the obtained K-near neighbor set except for the first parameter in the intersection to form a first set; selecting a first parameter of a second quantity from the released K-near neighbor set except for the first parameter in the intersection to form a second set; and estimating an estimation score between the second parameter and the first parameter by using the owned data according to the intersection, the first set and the second set.

Description

technical field [0001] The present invention relates to a collaborative filtering technology in the field of information search, and more specifically, to a multi-party collaborative filtering technology that takes into account privacy protection. Background technique [0002] Facing the massive amount of information on the Internet, how users can quickly and effectively obtain the part of the information they are interested in has always been a research hotspot in Internet applications. Usually the solution is based on two main ideas: the user's active search and the system's active recommendation. Now, especially in the field of Internet e-commerce, active recommendation technology has been widely used as a core marketing method. The most representative, practical and widely used recommendation technology is collaborative filtering (Collaborative Filtering) technology. Collaborative filtering is based on collecting and processing the interest information of a group of us...

AI Technical Summary

Problems solved by technology

[0008] As mentioned above, both references 1 and 3 require a third-party server, which makes these two methods lack sufficient flexibility and convenience in practical applications

In addition, the privacy protection method in Reference 1 is only to simply anonymize users, but in fact, this simple anonymization cannot really protect user privacy in many cases, especially when some users have some special points of interest. In the case of , the real identity of the user can be deduced according to some abnormal points in the scoring data, which leads to the leakage of user privacy

Reference 3 protects privacy through random perturbation, which has the general disadvantage of random perturbation, that is, the problem of outlier exposure

Reference 2 does not require the participation of a third party, but is based on secure multi-party computing, which requires frequent communication between the participating parties during the calculation process, making the participants too coupled, poor operability in practical applications, and secure communication The adoption of the protocol will greatly reduce the computational efficiency

Images