Webpage advertisement putting device and method based on multilayer random hidden feature model
A technology of advertising placement and hidden features, applied in the field of advertising, can solve the problems of low advertising recommendation efficiency and sparse data, and achieve the effect of efficient and accurate advertising recommendation, wide application, and accurate delivery.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0050] Embodiment 1: see figure 1 , a web page advertisement delivery device based on a multi-layer random latent feature model, the device includes:
[0051] The advertising data collection module 110 is used to collect and store user's advertising behavior data. The advertisement behavior data refers to the interaction record between the user and the advertisement.
[0052] The data conversion module 120 is used to convert the advertising behavior data into a target matrix R, which is a high-dimensional sparse matrix with M rows and N columns, where M represents the number of users and N represents the number of advertisements. The target matrix R refers to the matrix that uses the interaction record between the user and the advertisement to convert, and the element r in the matrix u,i Represents the element in the uth row and ith column of the matrix, if the uth user has browsed the ith advertisement, then r u,i = 1, otherwise the value is a missing value.
[0053] The ...
Embodiment 2
[0063] Example 2: see figure 2 , a web page advertisement delivery method based on a multi-layer random latent feature model, characterized in that: the method includes the following steps:
[0064] S1, collecting and storing user's advertising behavior data.
[0065] S2, converting the advertising behavior data into a target matrix and storing it for later use.
[0066] S3, based on the target matrix, randomly generate weights and offsets, use the activation function to generate a user behavior feature matrix, and then use the user behavior feature matrix to derive the advertisement feature matrix.
[0067] The weights include the first-layer weight matrix A and the multi-layer weight matrix W, which are used to add weighted items to the user behavior feature matrix in the activation function.
[0068] The bias includes the first layer bias vector b and the multi-layer bias vector d, which are used to add a bias item to the user behavior feature matrix in the activation fu...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
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