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

Cloud exhibition content recommendation method, system and equipment based on generative confrontation network

A content recommendation and network technology, applied in biological neural network models, neural learning methods, instruments, etc., can solve the problems of sparse samples and features, modeling, and insufficient display of materials, achieve accurate classification results, and solve data sparseness , to facilitate the calculation of the updated effect

Active Publication Date: 2021-11-09
TURING AI INST NANJING CO LTD
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0009] However, in the user-centered recommendation system, it does not reflect how to interact with users in a natural and transparent way, understand the real needs and preferences of users, and through natural interaction with users, the recommendation system estimates, extracts, and also provides User feedback, affecting the user's satisfaction, preference, demand, interest, activity patterns and other implicit states, so as to start from the user's preference, the behavior process of user decision-making reasoning, and make the best decision for the user.
[0010] At the same time, the goals of optimization embodied in existing technologies are all short-term benefits, such as click-through rate and viewing time, and it is difficult to model long-term benefits; the most important thing is to predict user interest, but the models are all based on logged feedback training , the samples and features are extremely sparse, and a large number of materials have not been fully displayed. At the same time, there are still a large number of new materials and new users pouring in, and there is a large amount of bias. In addition, users' interests change drastically, behaviors are diverse, and there are many noises

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
  • Cloud exhibition content recommendation method, system and equipment based on generative confrontation network
  • Cloud exhibition content recommendation method, system and equipment based on generative confrontation network
  • Cloud exhibition content recommendation method, system and equipment based on generative confrontation network

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach

[0090] Such as Figure 6 As shown, the present invention uses a discriminator to judge the authenticity of the simulated user behavior. The specific implementation is as follows:

[0091] S2-23. Compare the score sequence of user browsing behavior data output by the score calculation model with the behavior sequence output by the deep learning model LSTM, wherein the score sequence of user browsing behavior data at this time is converted into the input form of the discriminator User browsing behavior sequence;

[0092] It should be noted that, in order to effectively solve the data sparsity problem of training samples in the prior art, before re-dividing the training samples in the score calculation model according to the comparison results, it also includes

[0093] First, the LSTM hidden layer vector contained in the LSTM output behavior sequence of the deep learning model is connected to the Softmax classifier to obtain the classification results of user browsing behavior ...

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 present invention provides a cloud exhibition content recommendation method based on a generative confrontation network, including the first step, constructing a generator for generating user browsing behavior score sequence; the second step, constructing a discriminator, and verifying the authenticity of the score sequence Judging to obtain valid user browsing behavior data; the third step is to recommend unique valid user browsing behavior data as the target result of the determined output. The present invention provides cold-start recommendations according to the registration characteristics of new users by modeling user characteristics and content characteristics, and simultaneously utilizes the similarity of characteristics of different users to provide recommendations based on user characteristics and uses content characteristics to make recommendations to reflect users from multiple perspectives The way of preference is to achieve a combined display that brings more latitude to users, so as to interact and obtain user feedback.

Description

technical field [0001] The invention relates to the technical field of digital recommendation for an online cloud exhibition system, specifically a method, system and equipment for cloud exhibition content recommendation based on a generative confrontation network. Background technique [0002] The cloud exhibition system relies on the network technology platform to operate an exhibition project with an organic combination of text, images, and videos, regardless of time, venue, number of people, and quantity of goods. Its essence is based on the Internet to build a digital information integrated display space with cloud computing, big data, mobile Internet technology, social communities, and various entities in the exhibition industry chain, thus forming a comprehensive and three-dimensional new exhibition The model, which is also an extension and supplement to offline exhibitions, is a derivative product of offline physical exhibitions. It is known as "the never-ending exhi...

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
Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/9535G06K9/62G06N3/04G06N3/08
CPCG06F16/9535G06N3/08G06N3/044G06F18/2415G06F18/214
Inventor 龙利民陈功李强
Owner TURING AI INST NANJING 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