Data processing method and device, electronic device, computer readable medium

Abstract

The disclosure relates to a data processing method and device, electronic equipment, and a computer-readable medium. Belonging to the field of Internet technology, the data processing method includes: obtaining training data by counting historical user behavior data, wherein the training data includes basic information of training stores and corresponding training user information, and the basic information of training stores includes training store information. According to the training data, the neural network model is trained; according to the basic information of the candidate store and the trained neural network model, the corresponding candidate user information is obtained, wherein the basic information of the candidate store includes the candidate store geographic location information. The present disclosure enables geographically constrained search result recall.

Description

technical field [0001] The present disclosure relates to the technical field of the Internet, and in particular, to a data processing method and device, electronic equipment, and a computer-readable medium. Background technique [0002] Keyword advertising is a very important form of search advertising. After a user is given a search query (query), the user returns some relevant sorted advertising result content. The most important part of the entire keyword advertisement is the recall of the advertisement, which is to determine the advertisement candidates that meet the relevance, after which the results sorting and other processes will be carried out. Both the relevance and quantity of the recall results play a crucial role in the quality of the overall advertising results. [0003] In the O2O (Online To Offline, online to offline / online to offline) scene, the search advertisement recall link has unique relevance requirements and is very sensitive to geographical location...

AI Technical Summary

Problems solved by technology

[0010] In the O2O scenario, there will be distance constraints, but the absolute threshold is not easy to determine. A larger distance threshold will lead to high computational performance pressure in subsequent stages such as sorting, and a smaller distance threshold may miss recalling some candidate advertisements that originally met the conditions.

Images