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

Session recommendation method based on convolutional self-attention network

A recommendation method and attention technology, applied in neural learning methods, biological neural network models, marketing, etc., can solve the problems of poor model robustness, large number of fully connected decoder parameters, long training time, etc., to improve performance. Effect

Pending Publication Date: 2021-01-22
ZHEJIANG UNIV
View PDF0 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] 2) The conventional fully connected decoder has a huge amount of parameters, long training time, and poor model robustness

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
  • Session recommendation method based on convolutional self-attention network
  • Session recommendation method based on convolutional self-attention network
  • Session recommendation method based on convolutional self-attention network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0050] The overall framework of the method in this embodiment is as follows figure 2 shown. In order to facilitate the understanding and unification of writing later, this section gives a formulaic description of some terms involved in the following. The relevant mathematical symbols and their meanings are shown in Table 1.

[0051] Table 1 Conversation recommendation related mathematical symbols and meanings

[0052]

[0053] The present invention is based on the session recommendation method of convolutional self-attention network specifically comprising the following steps:

[0054] Step 1. Get a vector representation of each item

[0055] 1.1) For a given input session, use the item embedding matrix emb to convert the input item sequence [x 0 ,x 1 ,...,x t-1 ,x t ] index is mapped to a sequence of real-valued vectors in a low-dimensional space, and the item embedding representation is obtained.

[0056] 1.2) In order to supplement the location sequence informat...

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 session recommendation method based on a convolutional self-attention network. The method comprises the following steps: 1) expressing each article in a session as a low-dimensional vector, wherein the low-dimensional vector is formed by adding article embedding and position embedding; 2) performing sequence modeling and intention modeling on the low-dimensional vector, capturing sequence information of the session by the sequence modeling, and capturing key intention information of the session by the intention modeling; and 3) based on the obtained splicing sequenceinformation and the key intention information, selectively predicting whether a user clicks the repeated articles or the non-repeated articles in the next step. Compared with the prior art, the invention has the advantages that firstly, the interdependence among different segments in the session can be captured, and the article representation sensitive to the session segment is obtained; and then,a bidirectional linear decoder is used, so that the parameter quantity of the model is reduced, and the performance and robustness of the model are improved. Finally, Gaussian shift is used to improve an attention layer, and a Gaussian weight factor is calculated, so that the performance of the repeated recommendation decoder is improved.

Description

technical field [0001] The invention relates to the application of the neural network method in the session recommendation technology, in particular to a technical method for capturing the local segment features of the session by using the convolution operation and enriching the information amount of the weight factor by using the Gaussian offset. Background technique [0002] In the context of the big data era, "information overload" is a common problem. How to obtain valuable information from complex data is a key problem in the development of big data technology. Recommender Systems (RS for short) is an effective method to solve information overload. The recommendation system refers to the use of historical interaction information between consumers and websites to model consumers and their interaction information, tap consumer interests, and then filter and evaluate massive choices, and finally personalize for consumers Recommended technical system. [0003] Traditiona...

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): G06Q30/06G06Q30/02G06N3/08G06N3/04
CPCG06Q30/0631G06Q30/0202G06N3/08G06N3/045
Inventor 张寅汪千缘
Owner ZHEJIANG UNIV
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