REAL RELATIONSHIP MATCHING METHOD AND APPARATUS FOR SOCIAL PLATFORM USERS, AND READABLE STORAGE MEDIUM

摘要

Disclosed are a real relationship matching method for social platform users, a data processing apparatus and a computer-readable storage medium. The method comprises the steps of: acquiring account information about users on a social platform, and establishing a user vector model according to the account information about each user and a TransE model (S10); acquiring textual interaction information between the users on the social platform, and establishing a textual relationship prediction model according to the textual interaction information between the users and a convolutional neural network model (S20); and projecting an output result of the user vector model and an output result of the textual relationship prediction model into the same space for linear combination and output to obtain a real relationship prediction result of the users on the social platform (S30). The method of analysing by combining user account information and interactive text information between users can accurately mine the real relationship of online users from a huge number of users upward of a hundred million in a social platform.