Method and device for generating user description text based on text generation network

Abstract

The embodiments of this specification provide a method and device for generating user description text based on a text generation network. The method includes: inputting various features of a target user into a first encoder, and obtaining each initial corresponding to each feature through the first encoder The user feature vector is encoded based on the self-attention mechanism, and the encoded state vector is obtained; the encoded state vector is input into the retrieval model, and K sentences are retrieved from the artificial knowledge base through the retrieval model, and the words corresponding to each word contained in the K sentences are determined. encoding vector, determine each attention coefficient according to the output feedback vector of the decoder and the word encoding vector, and perform weighted summation of each word encoding vector according to each attention coefficient to obtain a semantic representation vector; input the encoding state vector and semantic representation vector The decoder generates the user description text of the target user through the decoder, and the hidden state of the decoder is used as the output feedback vector. Can improve the quality of the resulting text.

Description

technical field [0001] One or more embodiments of this specification relate to the computer field, and in particular, to a method and apparatus for generating user description text based on a text generation network. Background technique [0002] Since there is an associated relationship between the user characteristics of the user and the user category of the user, the user can be classified based on the user characteristics of the user. User characteristics may be the user's age, education, income and other data, and the user category may include multiple preset categories, for example, whether there is a repayment risk, whether there is a money laundering risk, and the like. Usually, it is not convincing to only give the user characteristics of the user and the user category of the user. Therefore, after obtaining the user characteristics of the user, it is necessary to generate a user description text based on the user characteristics. The user description text includes ...

AI Technical Summary

Problems solved by technology

[0003] In the prior art, the user description text is often generated based on the text generation network, that is to say, the user description text is generated by machine learning. Since the number of training samples used to train the text generation network is small, the quality of the text obtained by this method is not good. good

Images