Data processing method, data processing device, data processing system, and electronic equipment

Abstract

The invention discloses a data processing method, a data processing device, a data processing system, and an electronic equipment. The data processing method comprises steps that a user social characteristic calculating device is used for receiving a social characteristic calculation request of at least one prediction model of a target user set transmitted by a social characteristic calculation requester; according to the social characteristic calculation request of the at least one prediction model, N to-be-calculated social characteristics associated with the at least one prediction model are determined; according to corresponding relations between stored social characteristics and social characteristic calculation modes, K social characteristic calculation modes corresponding to the N to-be-calculated social characteristics are determined; by taking the social characteristic calculation modes as calculation units, the social characteristic value of every to-be-calculated social characteristic of the N to-be-calculated social characteristics corresponding to the K social characteristic calculation modes is calculated; and the calculated social characteristic values are returned to the social characteristic calculation requester. Social characteristic calculation time is accelerated, and calculation resources are saved.

Description

technical field [0001] The invention relates to the technical field of the Internet, in particular to a data processing method, device, system and electronic equipment. Background technique [0002] In the Internet and telecommunications business, in order to better understand user needs, improve user experience, and increase business revenue, business operators will build a user behavior analysis system. and behavior records, establish data mining models, establish user portraits, and predict user behaviors. The business operator will provide different business service models or adopt different business promotion and marketing strategies according to user behaviors. [0003] In user behavior analysis, the user's social attributes (such as the number of friends, the age of friends, the way friends use services and their preferences, etc.) will also reflect the user's attributes and behaviors to a large extent. More and more user behavior analysis systems also use various so...

AI Technical Summary

Problems solved by technology

[0004] The social feature calculation method of the existing social feature-based user behavior prediction system is as follows: the original input of the prediction model is the user’s social behavior records (call records, message records, follow behavior, etc.) and user attributes (including basic attributes and behavior attributes) ;According to the input data, build a graph data set with users as nodes and user relationships as edges, the attributes of nodes are user attributes, and the attributes of edges are user relationship attributes; multiple social feature values ​​for predicting user behavior are defined based on social database , as the input of the prediction model; the prediction model training stage will calculate the social feature value vector of each sample user, and generate the corresponding prediction model. In the above social feature calculation method, the defined social features are independent of each other, and the social features need to Calculation, social feature calculation is slow

Images