A Collaborative Filtering Method Fused with Content and Behavior for Personalized Recommendations

A collaborative filtering and behavioral technology, applied in data processing applications, character and pattern recognition, buying and selling/lease transactions, etc., can solve problems such as dependencies and data sparsity, improve efficiency and real-time performance, improve calculation accuracy, and improve recommendation. quality effect

Active Publication Date: 2021-10-08
TONGJI UNIV
View PDF13 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, relying too much on user ratings will face serious data sparsity problems

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 Collaborative Filtering Method Fused with Content and Behavior for Personalized Recommendations
  • A Collaborative Filtering Method Fused with Content and Behavior for Personalized Recommendations
  • A Collaborative Filtering Method Fused with Content and Behavior for Personalized Recommendations

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0078] Such as figure 1 As shown, a collaborative filtering method for fusion content and behavior for personalized recommendation, the method includes the following steps:

[0079] (1) Feature input: feature input includes item-attribute matrix representing item content and user-item browsing matrix, user-item collection matrix, user-item purchase matrix, item evaluation label matrix and user-item rating matrix representing user behavior ;

[0080] (2) Content-based item clustering: combine item-attribute matrix and item evaluation label matrix, calculate item similarity, and cluster items;

[0081] (3) Rating prediction and feature filling: For the unrated items in the user-item rating matrix, find the nearest neighbor items in the clusters with high similarity to the unrated items, and according to the ratings of the nearest neighbor items, unrated The item performs rating prediction and populates the user-item rating matrix;

[0082] (4) Behavior-based user clustering: ...

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 relates to a collaborative filtering method for merging content and behavior for personalized recommendation. The method includes the following steps: (1) feature input: including an item-attribute matrix representing item content and a user behavior matrix representing user behavior; (2) Content-based item clustering: Calculate item similarity and cluster items; (3) Score prediction and feature filling: Score prediction for unrated items, populate user-item scoring matrix; (4) Behavior-based User clustering: According to the item clustering results and the user-item rating matrix, the users are clustered; (5) Rating prediction and item recommendation: determine the cluster where the target user is located, find the nearest neighbor user set, and classify the target user The unrated items are predicted, and finally the top N items with the highest predicted ratings are recommended to the target users. Compared with the prior art, the present invention effectively solves the problems of data sparsity and cold start, and has high recommendation efficiency.

Description

technical field [0001] The invention relates to a collaborative filtering method in personalized recommendation, in particular to a collaborative filtering method for fusing content and behavior. Background technique [0002] Since Resnick and Varian proposed the concept of recommender systems in the 1990s, recommender systems have attracted extensive attention from academia and industry due to their important theoretical and application values, and have developed into an independent research field. In academia, there are many famous research institutions abroad, which have produced a large number of breakthrough academic research and academic papers. In the industry, practical recommendation systems serve various fields such as Internet news, e-commerce, and social networking. Major Internet companies at home and abroad also have teams specializing in recommendation systems. [0003] The common sorting and popular recommendation of each site is also a form of recommendatio...

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): G06K9/62G06Q30/06
CPCG06Q30/0631G06F18/23213
Inventor 马云龙刘敏袁菡殷蓉
Owner TONGJI UNIV
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