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

User similarity calculation method

A kind of user similarity, calculation method technology, applied in the direction of calculation, computer parts, special data processing applications, etc., can solve the problem of lack of consideration, achieve short time, good personalized recommendation service, improve accuracy and universality. Effect

Active Publication Date: 2020-09-08
SHANGHAI UNIV
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, it only considers the positive impact of clustering on similarity, and does not consider the influence of other factors that affect similarity, such as user frequency, time difference of operation behavior, and location distance, on similarity.

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
  • User similarity calculation method
  • User similarity calculation method
  • User similarity calculation method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0049] In the traditional collaborative filtering recommendation algorithm, the calculation of user similarity is relatively simple, and the similarity is generally judged directly according to the user's historical behavior. The user similarity calculation method of the present invention can more comprehensively consider the factors that affect the user similarity. Firstly, the time complexity is reduced by simple clustering and grouping of user information, and the positive correlation between similarity and login time and the negative correlation that decays over time are considered into a formula at the same time, and the denominator is normalized Processing; by increasing the weight of self-information and the weight of negative frequency correlation for users, the accuracy of similarity calculation can be comprehensively improved.

[0050] Specifically, the user similarity calculation method mainly includes: performing simple clustering and grouping according to user attr...

Embodiment approach

[0061] In another preferred embodiment, the user similarity calculation method provided by the present invention first reduces the time complexity by performing simple clustering and grouping of user information, and combines the positive correlation between similarity and login time with time At the same time, the attenuation negative correlation is considered into a formula, and the denominator is normalized; by adding the weight of self-information and the weight of frequency negative correlation to the user, and on this basis, by obtaining the user's location information, taking the distance weight into user similarity can improve the accuracy of similarity calculation more comprehensively.

[0062] Specifically, the user similarity calculation method mainly includes:

[0063] S1 performs clustering and grouping according to user attributes, and calculates the similarity sim based on static attributes attr ;

[0064] S2 calculates the similarity sim according to the user...

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 provides a user similarity calculation method, which comprises the steps of S1, performing clustering grouping according to user attributes, and calculating similarity simattr based on static attributes; S2, calculating the similarity simtime according to the fact that the user similarity is in direct proportion to the login time but is subjected to the time attenuation effect; S3, calculating the similarity simfre according to the negative correlation relationship between the user similarity and the operation behavior frequency of the user; s4, according to the uncertainty thatthe user is attracted by the product, the self-information amount is increased for the user, and the similarity siminf is obtained; and S5, superposing and normalizing the similarities calculated in the steps S1-S4 to obtain the final user similarity.

Description

technical field [0001] The present invention relates to recommendation algorithm technology, in particular to a method for calculating user similarity Background technique [0002] In the traditional collaborative filtering recommendation algorithm, the calculation of user similarity is relatively simple, and the similarity is generally judged directly according to the user's historical behavior, so the accuracy is not high. [0003] For this reason, the prior art once proposed a "Collaborative Filtering Recommendation Algorithm Integrating Penalty Factor and Time Weight", DOI is 10.19358 / j.issn.2096-5133.2020.05.004. But there are some problems in the traditional collaborative filtering algorithm, such as cold start, data sparseness and Matthew effect. Lan Yan et al. used the attenuation factor to establish a non-linear time weighting function, and gave different time weights to the score, which improved the accuracy of the recommendation. Although the above literature ta...

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 Applications(China)
IPC IPC(8): G06F16/9536G06K9/62
CPCG06F16/9536G06F18/23G06F18/22G06F18/251
Inventor 王斌张克
Owner SHANGHAI 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