A user behavior pattern analysis method and device

Abstract

The invention discloses a user behavior pattern analysis method and device. The user behavior pattern analysis method comprises the steps of acquiring card swiping data of a user, building N data points with user IDs according to the card swiping data, and clustering the N data points to obtain k target clusters, wherein the data points are multi-period distribution vectors of card swiping frequencies of a single user within a set time period; seeking user information corresponding to the user IDs according to the user IDs of the data points in the target clusters, and determining label information used for marking a user behavior pattern of each target cluster according to the user information; building an incidence relation between a central point of the target cluster and the label information; and determining that the central point of the target cluster is a vector model representing the user behavior pattern. On the basis of the vector model, other users are found out by using the model to match, a probable type of the user can be judged, some suspicious users who probably and similarly possess the behavior can be identified, and the user behavior pattern analysis method and device has enough screening capability.

Description

technical field [0001] The present invention relates to the field of communication technologies, in particular to a user behavior pattern analysis method and device. Background technique [0002] In most applications of card swiping records for electronic access control in the prior art, only simple search operations can be completed, for example, according to the user's unique identifier (ID) or time period information, the set of records that meet the conditions are screened out from the data records, and then Carry out manual analysis, drawing, operation, research and judgment, etc. In some applications, the use of access control card swiping records is limited to the statistics of simple indicators such as the number of card swipes and the frequency of card swipes during key periods. For example, for the use of library card swiping records, the reader's behavior is limited to providing statistical data on the daily average card swiping frequency, which is used to analyz...

AI Technical Summary

Problems solved by technology

Due to the lack of corresponding analysis algorithms and auxiliary tools, the analysis method of access control card swiping behavior based only on the simple number of card swipes per day or the number of swipes in special periods has the problems of single focus, limited analysis ability and poor effect

[0003] Therefore, in the face of a large amount of access control data, the existing analysis capabilities are limited. Simply relying on simple access frequency analysis cannot have sufficient screening capabilities, and cannot conduct in-depth analysis of access control card swiping data. It is impossible to identify and judge the user’s identity based on the user’s card swiping records. Possible types, unable to truly describe the user's behavior pattern

[0004] At present, there is no better implementation plan for modeling and analyzing the behavior patterns of card swiping users through the card swiping records of access control

Images