DPI clustering and user expansion method

Abstract

The invention discloses a DPI clustering and user expansion method, and the method comprises the steps: determining N different and unclassified DPIs, wherein each DPI comprises an access user group;using an Apriori algorithm in an association rule to discover the number C of non-repeated frequent item sets in the N DPIs; traversing each frequent item set in the C frequent item sets, and selecting a high-quality frequent item set K by taking the user cardinal number and the frequent item set in the marketing advertisement direction as standards; calculating DPI support degrees for DPI and theaccess user group of each non-reference frequent item set in the K high-quality frequent item sets; calculating all the DPI support degrees, and sorting the DPI support degrees according to the sizeof the support degrees; and according to the support degree of the DPI and the number of the users accessing the DPI, determining whether to add the users accessing the DPI as potential target users into a target group of advertisement putting until the number of the users is expanded to a satisfactory range.

Description

technical field [0001] The present invention relates to the field of big data processing and information technology, and more specifically, relates to a DPI clustering and user expansion method based on an association rule mining algorithm and user DPI access conditions. Background technique [0002] With the development of big data and information technology, there are more and more scenarios for expanding the number of users in the marketing of operators, especially in the field of advertising and marketing, it is often very difficult to find user groups that meet marketing methods from a large number of user groups. [0003] In the Internet industry, users can be easily grouped by clicking, downloading, using and other behaviors of web pages or APPs, and then exposure and delivery of advertisements. In the traditional marketing using operators, it is difficult to easily obtain a high-quality potential marketing user group. [0004] In addition, in the prior art, when use...

AI Technical Summary

Problems solved by technology

[0006] ①. It is difficult to obtain a large number of potential users, because the cost of finding new target groups from all unknown target groups is very high, and the number of new users is difficult to guarantee;

[0007] ②. The efficiency of finding new users through user characteristics is too low to achieve rapid expansion of users, because the above-mentioned characteristics acquisition channels are cumbersome and the correctness cannot be guaranteed

Images