A double-clustering recommendation method and system based on reinforcement learning

A recommendation method and reinforcement learning technology, applied in the field of recommendation, can solve problems such as unsatisfactory recommendation effect and poor user experience

Active Publication Date: 2021-07-30
CHENDU PINGUO TECH CO LTD
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In order to solve the above problems, the present invention provides a double-clustering recommendation method and system based on reinforcement learning, which uses double-clustering, dimensionality reduction and reinforcement learning methods to solve the problem that the current user-item recommendation method is not ideal. Bad experience

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 double-clustering recommendation method and system based on reinforcement learning
  • A double-clustering recommendation method and system based on reinforcement learning
  • A double-clustering recommendation method and system based on reinforcement learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0046] In this example, if figure 1 As shown, a biclustering recommendation method based on reinforcement learning includes the following steps:

[0047] S1. Establish a user-item matrix;

[0048] Use the behavior of the user to click on the recommended items (filters, stickers) to construct the user-item matrix. Assume that the user has N=5 and the items M=12, and construct an N×M user-item matrix, as shown in Table 1 .

[0049] Table 1:

[0050] Item 1 Item 2 Item 3 Item 4 item 5 Item 6 Item 7 Item 8 Item 9 item 10 Item 11 Item 12 user 1 1 0 0 1 0 1 1 0 1 1 1 0 user 2 1 1 1 1 0 0 0 1 0 1 1 0 user 3 1 1 1 1 1 1 1 0 0 1 1 0 user 4 1 1 1 1 0 0 0 1 0 1 1 0 user 5 1 1 0 0 0 1 1 1 0 1 1 0

[0051] S2. Performing biclustering on the user-item matrix to obtain a plurality of biclustering clusters;

[0052] Using the BiBit double clustering method, first enco...

Embodiment 2

[0105] A biclustering recommendation system based on reinforcement learning, including:

[0106] A double-clustering module, configured to double-cluster the user-item matrix to obtain a double-cluster cluster;

[0107] A state mapping module, configured to map the biclusters to a low-dimensional space to obtain a state set;

[0108] The reinforcement learning module is used to obtain the best state transition sequence according to the state collection through reinforcement learning;

[0109] The recommendation module is used to generate a recommendation list of users to be recommended.

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 double-clustering recommendation method and system based on reinforcement learning, which belongs to the technical field of recommendation, and includes steps: S1. Establishing a user-item matrix; S2. Performing double-clustering on the user-item matrix to obtain multiple Bi-clustering clusters; S3. Mapping the bi-clustering clusters to a low-dimensional space to obtain a state set; S4. Obtaining an optimal state transition sequence through reinforcement learning according to the state set obtained in S3; S5. Generating a recommendation list; Aiming at user-item recommendation scenarios, the present invention uses double clustering, dimensionality reduction and reinforcement learning methods to significantly improve the recommendation effect of user-item scenarios, and at the same time solves the problem of cold start, making the user experience better; in the online case, When a user triggers a new state by searching or providing geographic location and time information, a new recommendation list can be quickly and autonomously learned through reinforcement learning, so as to better perceive the user's preferences in the current state and meet the user's current needs.

Description

technical field [0001] The present invention relates to the technical field of recommendation, in particular to a double-clustering recommendation method and system based on reinforcement learning. Background technique [0002] In the case of information explosion, how to provide personalized content for different users has promoted the recommendation system to obtain significant development in recent years. [0003] At present, collaborative filtering is an important means in the recommendation system. For the user-item recommendation method, in massive scenarios, traditional collaborative filtering algorithms often generate recommendations based on users or items based on a single dimension. Due to the sparseness of the matrix And high dimensionality often leads to the recommendation system ignoring many important factors, resulting in unsatisfactory recommendation effects and poor user experience. In addition, there is a very practical problem in the recommendation syste...

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): G06Q30/06G06K9/62
CPCG06Q30/0631G06F18/23
Inventor 王丹徐滢
Owner CHENDU PINGUO TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap