Social network spam filtering method based on distributed matrix decomposition feature extraction

A social network and matrix decomposition technology, applied in the field of network garbage filtering, can solve the problems of coexistence of true and false data, massive and rapid dissemination, unbounded quantity, unreliability, etc.

CN104063516AActive Publication Date: 2014-09-24FUZHOU UNIVERSITY
6 Cites 1 Cited by

Patent Information

Authority / Receiving Office
CN · China
Current Assignee / Owner
Publication Date
2014-09-24

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention relates to the technical field of network spam filtering, in particular to a social network spam filtering method based on distributed matrix decomposition feature extraction. The social network spam filtering method based on distributed matrix decomposition feature extraction comprises the following steps that (1) a user and attribute matrix of a social network is established; (2) feature extraction is conducted on the user and attribute matrix of the social network based on distributed matrix decomposition; (3) potential feature vectors are classified, and whether the potential feature vectors are social network spam or not is judged. The method is beneficial to efficiently filtering spam data in the social network.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention relates to the technical field of network garbage filtering, in particular to a social network garbage filtering method based on distributed matrix decomposition feature extraction. Background technique

[0002] At present, social networks have played an important role in people's lives and have an impact that cannot be underestimated on people's information acquisition, thinking and life. Through social networks, users can communicate through chat rooms, create personal homepages to share favorite information, maintain more direct contact with friends, create a large social circle and find lost friends, follow other people's homepages and share Wait. However, the data in social networks has the coexistence of true and false, massive, rapid dissemination, and unbounded quantity. Due to the coexistence and massiveness of true and false data, there are a lot of unreliable content in social networks, such as fraudulent advertisements, hate...

Examples

Embodiment Construction

[0068] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0069] The present invention is based on the social network garbage filtering method of distributed matrix decomposition feature extraction, such as figure 1 shown, including the following steps:

[0070] Step S1: Construct social network user-attribute matrix. The construction method of social network user-attribute matrix is ​​as follows:

[0071] Assuming that there are n users and m attributes, the social network user-attribute matrix is ​​constructed as a set of n users in the known social network and m attribute sets , A ij means user u i pair attribute v j The metric value of , thus constructing the social network user-attribute matrix A ∈ R n×m :

[0072]

[0073] Since there are various attributes in a social network, and each user usually has only a few attributes, the social network user-attribute matrix ...