Unlock instant, AI-driven research and patent intelligence for your innovation.

A construction method and device for an item recommendation model based on a hybrid neural network, and an item recommendation method

A hybrid neural network and item recommendation technology, applied in the field of data mining, can solve problems such as poor recommendation effect, achieve the effect of improving recommendation effect and performance

Inactive Publication Date: 2021-07-02
WUHAN UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] In view of this, the present invention provides a method and device for constructing an item recommendation model based on a hybrid neural network, and an item recommendation method to solve or at least partially solve the technical problem of poor recommendation effect existing in methods in the prior art

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
  • A construction method and device for an item recommendation model based on a hybrid neural network, and an item recommendation method
  • A construction method and device for an item recommendation model based on a hybrid neural network, and an item recommendation method
  • A construction method and device for an item recommendation model based on a hybrid neural network, and an item recommendation method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0068] This embodiment provides a method for constructing an item recommendation model based on a hybrid neural network, please refer to figure 1 , the method includes:

[0069] Step S1: Filter the comment information according to the pre-built user-item scoring matrix, and preprocess the filtered comment information, where each row in the user-item scoring matrix is ​​used to represent the user-item scoring information The user features in , and each column in the rating matrix is ​​used to represent the item features in the user-item rating information.

[0070] Specifically, the user-item rating matrix is ​​constructed from users' previous ratings on items. In recommender systems, models that predict user preferences based on user-item rating matrices are generally called latent semantic models. Each row in the rating matrix is ​​called the user's preference feature (ie user feature), and each column is called the item feature. For example, user A likes books on mathemat...

Embodiment 2

[0129] This embodiment provides a device for constructing an item recommendation model based on a hybrid neural network, please refer to figure 2 , the device consists of:

[0130] The comment information preprocessing module 201 is used to filter the comment information according to the pre-built user-item scoring matrix, and preprocess the filtered comment information, wherein each row in the user-item scoring matrix is ​​used for Represents the user features in the user-item rating information, and each column in the rating matrix is ​​used to represent the item features in the user-item rating information;

[0131] The feature learning module 202 is used to learn the contextual features related to the item in the preprocessed comment information by using the convolutional neural network, and is used to learn the user features and item features in the user-item rating information by using the convolutional neural network;

[0132] The feature fusion module 203 is used to fu...

Embodiment 3

[0170] This embodiment provides a method for item recommendation, which includes:

[0171] The comment information and rating information corresponding to the item to be recommended are input into an item recommendation model based on a hybrid neural network constructed to obtain the recommendation result.

[0172] In order to illustrate the project recommendation method in the present invention more clearly, a specific process will be used to illustrate it below, please refer to figure 1 . The item recommendation model constructed by the present invention includes an embedding layer, a convolutional layer, a pooling layer, a fusion layer, a connection layer, a neural interaction layer and a prediction layer.

[0173] First collect training data, and judge whether it is unstructured text data, if yes, obtain the project description document set after preprocessing, if not, further judge whether it is structured project features.

[0174] After obtaining the preprocessed proj...

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 discloses a construction method and device of an item recommendation model based on a hybrid neural network, and an item recommendation method. The construction method first filters comment information, preprocesses the filtered comment information, and then uses convolution The neural network learns the contextual features related to the item in the preprocessed review information, the user features and item features in the rating information; then the item features in the user-item rating information and the contextual features in the review information are fused and interacted, Then the learned user features and fused item features are integrated into the multi-task learning framework, and jointly trained to obtain an item recommendation model based on a hybrid neural network. The present invention can more accurately learn the implicit feature vectors of users and items by integrating the two heterogeneous data of rating information and comment information into a unified model, thereby achieving the purpose of improving the performance of the recommendation system and improving the recommendation effect.

Description

technical field [0001] The invention relates to the technical field of data mining, in particular to a method and device for constructing an item recommendation model based on a hybrid neural network, and an item recommendation method. Background technique [0002] The development of network information technology, while satisfying people's demand for information, has also caused the problem of information overload. In the face of massive amounts of information, it is difficult or costly for users to find the information they are interested in. Therefore, how to quickly and efficiently select the information you are interested in from the massive amount of information has become a major problem in the information age. The recommendation system emerged as the main method to solve the problem of information overload. It analyzes the user's historical activity information, mines the user's preference, provides the user with personalized information, products or services, and s...

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): G06F16/9536G06N3/04
CPCG06F16/9536G06F2216/03G06N3/045
Inventor 李晶刘东华杜博常军高榕吴玉佳
Owner WUHAN UNIV
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com