Feature sample real-time acquisition method and system

Abstract

The embodiment of the invention discloses a feature sample real-time acquisition method which comprises the following steps: when a sample production node monitors a result message of a model result message system, the sample production node obtains metadata required by a feature sample from a distributed configuration system according to a session ID of the result message, and obtains all features required by the feature sample from the feature library, constructs a cache sample according to the metadata and all the features, and stores the cache sample in a storage system; each real-time behavior acquisition node monitors a user behavior message system, and when the real-time behavior acquisition node monitors the behavior message of the user behavior message system, the real-time behavior acquisition node obtains a cache sample from a cache system according to the session ID of the behavior message and generates a sample label; and the sample label and the cache sample are integrated into a feature sample and the feature sample is stored in the feature sample storage system. The embodiment of the invention also discloses a feature sample real-time acquisition system. The method can well support real-time collection of feature samples.

Description

technical field [0001] The invention relates to the technical field of machine learning, in particular to a method and system for real-time collection of feature samples. Background technique [0002] Feature sample collection is an important part of machine learning model training, and the quality of data determines the effect of the algorithm model. In recommender system-related technologies, samples are mostly obtained through off-line calculations. The production efficiency of samples in this way is low, the production of samples is complicated, and the accuracy of the obtained samples is also low. Contents of the invention [0003] The purpose of the present invention is to provide a method and system for real-time collection of feature samples, which can not only well support the real-time collection of feature samples to improve the online training effect of the model, but also solve the problems of low offline production efficiency, complicated production and low a...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to provide a method and system for real-time collection of feature samples, which can not only well support the real-time collection of feature samples to improve the online training effect of the model, but also solve the problems of low offline production efficiency, complicated production and low accuracy of feature samples. question

Images