A data processing method, device and storage medium

Abstract

The present application discloses a data processing method, device, and storage medium. The method includes: acquiring log data of a target object within a preset time period; The time difference between the Nth exposure time of the target object and the Nth click time for the target object; wherein, N is a positive integer; when the time difference is less than or equal to the first threshold, calculate the target object The expected exposure duration of the target object; when the expected exposure duration of the target object is less than the preset duration threshold and the time difference is greater than the second threshold, the exposure data corresponding to the Nth exposure time and the Nth click time The corresponding click data is determined as abnormal data; and the abnormal data is deleted from the log data. The data processed by the method of the present application is used as the training data, and the target click rate prediction model obtained has a higher prediction accuracy.

Description

technical field [0001] The present application relates to the field of computer technology, and in particular to a data processing method, device and storage medium. Background technique [0002] Internet advertising is developing rapidly in the Internet field. Compared with traditional media, Internet advertising spreads faster. As a new advertising medium, it has become the best choice for small and medium-sized enterprises to expand their popularity. During the rapid development of Internet technology, some companies take advantage of the rapid dissemination of Internet information to place advertisements, mainly to expand the promotion of advertisements, and the prediction of advertisement clicks is more important. [0003] The personalized advertising recommendation industry is booming. If the recommendation algorithm is not accurate enough, it will not only affect the user experience, but also have a negative impact on the revenue of the advertising platform. In order...

AI Technical Summary

Problems solved by technology

[0006] The general ranking formula for advertising recommendation is "bid * pctr * pcvr * quality factor". If other variables remain unchanged, pCTR is estimated to be high, so that advertisers can obtain exposure with a lower bid, achieving "low bid, high exposure" The speculative purpose of the platform will correspondingly cause a loss of revenue for the platform. At the same time, because the accuracy of the recommended advertisements will be affected, the user experience will also be compromised.

[0007] It can be seen that there are large errors in the data used to train the advertising click-through rate prediction model in the prior art, which will lead to a decrease in the prediction accuracy of the advertising click-through rate

Images