Data source quality evaluation method, device and equipment and computer readable storage medium

Abstract

The invention provides a data source quality evaluation method, apparatus and device, and a computer readable storage medium. The method, apparatus and device are applied to a first participant device. The method comprises the steps of obtaining first feature data held by a first participant device and a second encryption feature sent by a second participant device, wherein the second encryption feature is calculated by the second participant device based on second feature data in a to-be-evaluated data source; training a preset first model based on the first feature data and the second encrypted feature to obtain a trained first model; acquiring test data held by the first participant device, and determining a test result based on the test data and the trained first model; and based on the test result, determining an evaluation result of the to-be-evaluated data source. According to the method and the device, the passive party data source quality can be evaluated through joint modeling, and the model effect of joint modeling is improved.

Description

technical field [0001] The present application relates to artificial intelligence technology, and in particular to a data source quality assessment method, device, electronic equipment, computer readable storage medium and computer program product. Background technique [0002] With the general trend of all walks of life gradually strengthening data privacy protection, federated learning, a technology that can cooperate with multi-party data to establish machine learning while protecting data privacy, has become one of the keys to cooperation between enterprises / industries. Longitudinal federated learning is to take out the part of users and data that have the same participant users but different user data characteristics for joint modeling when the data characteristics of the participants are small but the user overlap is large. By improving the model performance, it can Enable participant users to provide better service to customers. [0003] In the vertical federation sc...

AI Technical Summary

Problems solved by technology

However, in actual scenarios, sometimes the active party has less labeled data, and the active party and the passive party can match less labeled data. The less data makes it impossible to evaluate the data quality or the evaluation results cannot reflect the real data quality, resulting in joint construction. Die's model doesn't work well

Images