Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Method and device for recommending items to user and storage medium

A project and user technology, applied in the field of recommending projects to users, can solve problems such as increased data volume, false positive samples, slow recommendation model training speed, etc., to achieve the effect of enhancing user experience and ensuring timeliness

Active Publication Date: 2020-06-05
TENCENT TECH (SHENZHEN) CO LTD
View PDF8 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, in the case of multi-product advertisements, that is, one advertisement displays multiple products, one of the multiple products is randomly selected as a training sample in the prior art, which transmits wrong information to the recommendation model, making the recommendation model less effective ; or split multiple pro

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method and device for recommending items to user and storage medium
  • Method and device for recommending items to user and storage medium
  • Method and device for recommending items to user and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0030] The item recommendation advertisement is based on the neural network algorithm, which inputs user characteristics and item characteristics into the neural network, and outputs a value between 0 and 1 to indicate the probability of the item being clicked. In this paper, a method for recommending items to users is proposed, which is based on the user-item dual-tower structure, because the user-item dual-tower structure can meet the timeliness requirements of online real-time recommendation, that is, from millions to tens of millions The top 100 related projects are recommended in the project library. Whether the evaluation is relevant is measured by the user's click rate on the item. The higher the click rate, the better the effect of the recommended item, and vice versa, the worse the effect of the recommended item.

[0031] The following description provides specific details for a thorough understanding and practice of various embodiments of the present disclosure. It ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to a method and device for recommending items to a user and a storage medium. The invention provides a recommendation model training method for recommending items to a user. Themethod comprises the following steps: acquiring a plurality of samples, each of the plurality of samples comprising a user feature, a project group and a label corresponding to the project group, theproject group comprising two or more projects, and the label indicating whether a user selects the project group; generating a respective project embedding vector for each project in the project group; performing weighted averaging on the generated project embedding vectors to obtain a comprehensive embedding vector of the project group; generating a user embedding vector of the user based on theuser characteristics; and training a recommendation model for recommending items to the user by using the obtained comprehensive embedding vector, the user embedding vector of the user and the label corresponding to the item group. The recommendation model can realize accurate recommendation of projects.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence and machine learning, in particular to a method, device and storage medium for recommending items to users. Background technique [0002] Dynamic Product Ads (Dynamic Product Ads) DPA is an advertisement that can use recommendation algorithms to display products of interest to users based on user preferences and interests. It is essentially a matching process between items and users. The existing commodity recommendation algorithm is an algorithm based on a neural network, which inputs user characteristics and commodity characteristics into the neural network, and outputs a number between 0 and 1 to represent the probability that the commodity is clicked by the user. For the case where an advertisement only displays one product, the probability of the product being clicked can be directly determined. However, in the case of multi-product advertisements, that is, one advert...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06Q30/06G06K9/62G06Q30/02
CPCG06Q30/0631G06Q30/0251G06F18/2411
Inventor 潘颖吉
Owner TENCENT TECH (SHENZHEN) CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com