Service pushing method and service pushing system

Abstract

The invention discloses a service pushing method and a service pushing system. The method comprises the steps that user characteristic vectors influencing the service order probability are determined; the user characteristic vectors of multiple users who have already ordered service are selected as samples and put into a logistic regression model to estimate the weight of each user characteristic variable in the user characteristic variables of the ordered users, and the weight vector of each user characteristic variable corresponding to each ordered user is generated; based on the weight vector of each ordered user, a benchmark weight vector corresponding to the user characteristic variables aiming at all users is predicted; the benchmark weight vector is put into the logistic regression model, and the values of the user characteristic vectors which can make the value of a Sigmoid function in the logistic regression model within the preset probability range are calculated; based on the calculated values of the user characteristic vectors, service is pushed to corresponding users. According to the technical scheme, the service pushing method and the service pushing system have high pushing precision.

Description

Technical field [0001] The invention relates to the field of telecommunications service push, in particular to a service push method and a service push system. Background technique [0002] In the prior art, the following two methods are often used for service push: [0003] First, establish a user interest model by obtaining service information, user identity, subscription service information, and online log information, and obtain the association relationship between user interest and telecommunication service and user interest model based on user interest model, telecommunications service information, and subscription service information , Get a list of recommended services, and recommend services to users. [0004] Second, according to the communication records of the user, determine the density factor of the user and each contact of the user. The density factor is used to indicate the frequency of communication between the user and each contact; to obtain the business informati...

AI Technical Summary

Problems solved by technology

[0005] The above-mentioned first push method requires the establishment of a user interest model, the integration of information in various systems, cross-system information fusion, Uniform Resource Locator (Uniform Resource Locator referred to as URL) analysis, stream processing and other calculations, which is difficult to develop , the cycle is long, the system efficiency is not high, and the range of recommendation results based on the user interest model is relatively coarse, which affects customer experience

[0006] The above-mentioned second push method uses the density factor of the user and each contact as the recommendation basis, and the push basis is single, and the effect is not good

Images