Specific user mining method and system based on naive Bayesian algorithm

Abstract

The invention discloses a specific user mining method and system based on a naive Bayesian algorithm, and relates to the technical field of networks. The method comprises the following steps that behavior information indexes of users within a set time period are collected from a server, and part of the behavior information indexes are selected to serve as characteristic indexes of the users; part of users are sampled from the collected users to serve as sampled users, and a naive Bayesian classifier is constructed through a Bayesian kit according to the characteristic indexes of the sampled users; the characteristic indexes, collected after the set time period, of the users serve as input variables of the naive Bayesian classifier, and the probability that the users are specific users is obtained. By means of the Bayesian kit, the naive Bayesian classifier is constructed according to the characteristic indexes selected from the behavior information indexes of the users, an effective characteristic index combination can be found, the specific users can be accurately recognized, and the construction efficiency of the naive Bayesian classifier is high.

Description

[0001] The present invention relates to the field of network technology, in particular to a specific user mining method and system based on a naive Bayes algorithm. Background technique [0002] With the rapid development of the live broadcast industry, the competition among various live broadcast platforms has become increasingly fierce. How to quickly and effectively dig out specific users from all users of live broadcast websites, for example, how to effectively remove potential paying users from all live broadcast websites The users are unearthed so that operators can make further refined marketing plans for specific users, improve user experience, and increase the conversion rate of specific users. This has become an urgent problem for various live broadcast websites. At present, due to the explosive growth of the number of users of live broadcast websites, user behaviors tend to be diversified and complicated, and user behaviors change rapidly, which directly leads to multip...

AI Technical Summary

Problems solved by technology

At present, due to the explosive growth of the number of users of live broadcasting websites, user behavior tends to be diversified and complicated, and user behavior changes rapidly, which directly leads to multi-dimensional user data, fast update of user data, and large data volume, which leads to the establishment of specific user data. The process of the prediction model and the model itself are relatively complex, and the modeling cycle is long, which leads to the low accuracy and efficiency of the model for specific user predictions

Images