A sample generation method and device for a user satisfaction prediction model

Abstract

Embodiments of the present invention provide a sample generation method and device for a user satisfaction prediction model, which relate to the computer field and can reduce the complexity of samples. The method includes: obtaining sample data of the network; performing normalization processing on the network data to generate a first network matrix; inputting the first network quality matrix into the first self-encoding neural network to obtain a second network quality matrix; The information matrix is ​​input into the second self-encoding neural network to obtain the second user information matrix; the first self-encoding loss function is calculated according to the first network quality matrix and the second network quality matrix; the first self-encoding loss function is calculated according to the first user information matrix and the second user information matrix The second self-encoding loss function: calculate the total loss function, and extract the corresponding target second network quality matrix and target second user information matrix as samples of the user satisfaction prediction model according to the minimum value of the total loss function. The embodiment of the present application is applied to sample generation of a user satisfaction prediction model.

Description

technical field [0001] Embodiments of the present invention relate to the field of computers, and in particular to a method and device for generating samples of a user satisfaction prediction model. Background technique [0002] The rapid development of mobile communication networks has experienced the transition from voice to data era, and user satisfaction is the primary issue for network operators to strive for success in the future. Once the user is dissatisfied with the product or service, the business will be terminated, and it may cause a chain reaction, so that the surrounding users will be affected by the unsatisfactory evaluation, and the related communication business will also be terminated. Therefore, improving user satisfaction is an important means to enhance the competitiveness of communication network service products in the market. [0003] Most of the existing user satisfaction prediction technologies simply integrate the data of the business support syst...

AI Technical Summary

Problems solved by technology

First of all, the existing technology does not explain how to integrate the data of the BSS domain and the OSS domain. If the data of the BSS domain and the OSS domain are directly used as the sample data learned by the machine learning algorithm, not only the sample data has high complexity, but also many Useless data increases the complexity of the user satisfaction prediction model. In addition, useless data will cause interference and reduce the accuracy of the results

Secondly, in the process of predicting the user satisfaction prediction model, the existing methods mainly select the sample data of the user satisfaction prediction model through sampling and traversal, but selecting sample data through sampling will reduce the accuracy of the prediction results. When traversing and selecting sample data, the sample data will have high complexity

Therefore, the problem of high complexity of the sample data selected in the prior art becomes more and more prominent

Images