A method and device for pushing information

Abstract

An information pushing method comprises: in response to a current behavior of a user, extracting an attribute feature from a session that the current behavior belongs to, inputting the attribute feature to a preset user behavior lifecycle model, and outputting a current phase of the user in a user behavior lifecycle, the user behavior lifecycle comprising a phase that a target object is not clear, a phase that the target object is clearly selected, and a phase that the target object is locked (201); inputting attribute features that are extracted from to-be-pushed objects and the user to a preset hit probability model corresponding to the current phase of the user in the user behavior lifecycle, and outputting hit probabilities of the to-be-pushed objects, hit probability models corresponding to different phases being different (202); selecting at least one to-be-pushed object in descending order of the hit probabilities of the to-be-pushed objects (203); and pushing the selected to-be-pushed object (204). The method can reduce, as much as possible, the possibility that a user performs repeated searches, improves user experience, and reduce resources of a searching server and a pushing server. Further disclosed is an information pushing device.

Description

technical field [0001] The present application relates to the field of computer applications, in particular to an information push method and device. Background technique [0002] On an e-commerce website, users can obtain the target objects they follow or like by entering search keywords in the search box on the site. Specifically, the search server first searches for objects that match the search keywords entered by the user based on a matching algorithm (such as a text matching algorithm), then sorts the searched objects, and finally displays the searched objects on the pages of the website according to the ranking , allowing users to select their preferred audience. [0003] In addition to providing the searched object to the user, the e-commerce website may also push other objects similar to the searched object to the user. Specifically, the push server first takes the searched object as a benchmark, and obtains objects similar to the searched object according to a co...

AI Technical Summary

Problems solved by technology

[0004] In the process of implementing this application, the inventors of this application have found that there are at least the following problems in the prior art: Unlike the search server that provides users with searched objects, the push server can generally only push a limited number of push objects to users

However, this sorting method only takes into account the correlation between the pushed object and the searched object, and the pushed object itself may not be the target object that the user pays attention to or likes, so the user needs to rebuild the search keyword to search again, and The repeated search process will not only reduce the user experience, but also excessively consume the resources of the search server and the push server

Images