Hierarchical commodity information filtering and recommending method
Hierarchical commodity information filtering and recommending method
 CN 105,809,474 A
 Filed: 02/29/2016
 Published: 07/27/2016
 Est. Priority Date: 02/29/2016
 Status: Active Application
First Claim
1. a stratification merchandise news filtered recommendation method, it is characterised in that comprise the steps:
 A1;
for commending system, structure one layering Poisson model；
A2, it is the vectorial z of K to each group of validated user commodity to structure length_{ui}, each of which component z_{uik}～
Poisson (θ
_{uk}β
_{ik}), scoring is sized to the inner product of corresponding user preference vector and item property vector, and wherein K is the length of item property vector sum user preference vector, z_{ui}For often organizing user, the commodity auxiliary vector that length is K to structure, θ
_{u}For user preference vector, β
_{i}For item property vector, k is the sequence number of component, and u is user'"'"'s sequence number, and i is commodity sequence number；
The method that A3, employing variation are inferred carries out approaching Posterior distrbutionp, utilizes coordinate rise method successive ignition until convergence, derives all hidden variablesAPPROXIMATE DISTRIBUTION；
Wherein the implication of each parameter is as follows;
β
is β
_{i}Set, θ
represents θ
_{u}Set；
ξ
_{u}Meeting the scale parameter in Gamma distribution for user preference vector, ξ
represents ξ
_{u}Set, η
_{i}Meeting the scale parameter in Gamma distribution for item property vector, η
is η
_{i}Set, z variable represents z_{ui}Set；
A4, prediction often organize user'"'"'s commodity to scoring,User can being carried out final recommendation according to the sequence of score size, wherein subscript T represents vector transposition, is row vector by column vector transposition.
Chinese PRB Reexamination
Abstract
The invention relates to a hierarchical commodity information filtering and recommending method which comprises the following steps: building a hierarchical Poisson model; establishing a vector lengthened by K for each pair of valid user'"'"'s commodities wherein given scores correspond to the dot products of a user'"'"'s preferential vector and a commodity attribute vector; using a variation inference method to approach posterior distributions; using a coordinate ascent method to do iterations until convergence; inferring approximate distributions of all hidden variables; predicting the score of each pair of valid user'"'"'s commodities; providing final recommendations to users according to their scores. The method distinguishes itself from other by firstly the generation of sparse representations for commodities and users, secondly the accurate fitting of the long tail effects on commodities and users, thirdly a weight descending effect on unscored users and commodities, fourthly rapid inference to sparse matrixes for scoring and finally good expandability for large scale scoring sets.

12 Citations
Multimedia messages recommend method and device  
Patent #
CN 108,388,624 A
Filed 02/12/2018

Current Assignee

A kind of commercial articles clustering confirmation method and device  
Patent #
CN 108,053,268 A
Filed 12/29/2017

Current Assignee

A kind of user characteristics extracting method and relevant apparatus  
Patent #
CN 107,862,532 A
Filed 09/22/2016

Current Assignee

A kind of processing method and system of commodity evaluation score  
Patent #
CN 108,510,297 A
Filed 02/28/2017

Current Assignee

The multitask recommended method and system of joint Bayesian inference and weighting refusal sampling  
Patent #
CN 110,210,944 A
Filed 06/05/2019

Current Assignee

A kind of item recommendation method and device  
Patent #
CN 109,509,051 A
Filed 09/12/2018

Current Assignee

Cooperation recommending system based on user partial interest digging  
Patent #
CN 101,271,559 A
Filed 05/16/2008

Current Assignee

Product information sending method and device  
Patent #
CN 102,467,709 A
Filed 11/17/2010

Current Assignee

Project correlation fused and probabilistic matrix factorization (PMF)based collaborative filtering recommendation system  
Patent #
CN 102,129,463 A
Filed 03/11/2011

Current Assignee

Collaborative filtering recommendation method based on user interest groups  
Patent #
CN 103,279,552 A
Filed 06/06/2013

Current Assignee

SYSTEM AND METHOD FOR MODELING DEMAND AND OPTIMIZING PRICES WITH IMMUNITY TO OUTOFSTOCK EVENTS  
Patent #
US 20130325551A1
Filed 06/05/2013

Current Assignee
N/A

Individualized recommendation method based on user preferences and commodity properties  
Patent #
CN 103,824,213 A
Filed 03/13/2014

Current Assignee

10 Claims

1. a stratification merchandise news filtered recommendation method, it is characterised in that comprise the steps:

A1;
for commending system, structure one layering Poisson model；A2, it is the vectorial z of K to each group of validated user commodity to structure length_{ui}, each of which component z_{uik}～
Poisson (θ
_{uk}β
_{ik}), scoring is sized to the inner product of corresponding user preference vector and item property vector, and wherein K is the length of item property vector sum user preference vector, z_{ui}For often organizing user, the commodity auxiliary vector that length is K to structure, θ
_{u}For user preference vector, β
_{i}For item property vector, k is the sequence number of component, and u is user'"'"'s sequence number, and i is commodity sequence number；The method that A3, employing variation are inferred carries out approaching Posterior distrbutionp, utilizes coordinate rise method successive ignition until convergence, derives all hidden variablesAPPROXIMATE DISTRIBUTION；
Wherein the implication of each parameter is as follows;
β
is β
_{i}Set, θ
represents θ
_{u}Set；
ξ
_{u}Meeting the scale parameter in Gamma distribution for user preference vector, ξ
represents ξ
_{u}Set, η
_{i}Meeting the scale parameter in Gamma distribution for item property vector, η
is η
_{i}Set, z variable represents z_{ui}Set；A4, prediction often organize user'"'"'s commodity to scoring,User can being carried out final recommendation according to the sequence of score size, wherein subscript T represents vector transposition, is row vector by column vector transposition.


2. stratification merchandise news filtered recommendation method as claimed in claim 1, it is characterised in that:
 in step A1, structure layering Poisson distribution is the situation for hidden feedback.

3. stratification merchandise news filtered recommendation method as claimed in claim 1, it is characterised in that step A1 includes:

A1 1;
each user u is constructed the user preference vector θ
that length is K_{u}, the potential feature of this user of this vector representation, wherein each component θ
_{uk}～
Gamma (a, ξ
_{u}), namely each component meets Gamma distribution and the parameter ξ
in this distribution_{u}Being defined as the liveness of this user, namely the commodity of customer consumption account for the ratio of all commodity and ξ
_{u}～
Gamma (a '"'"', a '"'"'/b '"'"'), namely user'"'"'s liveness equally also meets Gamma distribution；
Wherein parameter a represents that user preference vector meets the form parameter in Gamma distribution, and a '"'"' expression '"'"' abovementioned scale parameter meets the form parameter in Gamma distribution, and a '"'"'/b '"'"' represents that abovementioned scale parameter meets the scale parameter in Gamma distribution；A1 2;
the item property vector β
that length is K is constructed for every commodity i_{i}, this vector has again showed that the potential feature of these commodity, each of which component β
_{ik}～
Gamma (c, η
_{i}), namely each component meets Gamma distribution and the parameter η
in this distribution_{i}Being defined as the popularity of these commodity, the user namely consuming these commodity accounts for the ratio of all users and η
_{i}～
Gamma (c '"'"', c '"'"'/d '"'"'), namely commodity popularity equally also meets Gamma distribution；
；
Wherein parameter c represents that item property vector meets the form parameter in Gamma distribution, and c '"'"' represents that abovementioned scale parameter meets the form parameter in Gamma distribution, and c '"'"'/d '"'"' represents that abovementioned scale parameter meets the scale parameter in Gamma distribution；A1 3;
the scoring often organizing user'"'"'s commodity pair supposes to meet Poisson distribution, namelyParameter in this distribution is equal to the inner product of user preference vector and item property vector.


4. the stratification merchandise news filtered recommendation method as described in claim 1 or 2 or 3, it is characterised in that in step A2, it is assumed that z_{ui}Each component is separate.

5. the stratification merchandise news filtered recommendation method as described in claim 1 or 2 or 3, it is characterized in that being in that in step A3, by coordinate rise method, namely assume that the distribution of other hidden variables is known, maximize the KL divergence between the Posterior distrbutionp about the distribution of current goal variable and true distribution, by successive ignition until convergence, then draw the approximate Posterior distrbutionp of parameter.

6. stratification merchandise news filtered recommendation method as claimed in claim 4, is characterized in that being in that in step A4, at auxiliary variable z_{ui}Help under, the distribution of the full terms of each hidden variable is as follows:

7. stratification merchandise news filtered recommendation method as claimed in claim 5, it is characterized in that the coordinate all parameters of rise method iteration include:
 assume that the Posterior distrbutionp of hidden variable can be analyzed to the distribution product of each Independent Vector, namely assuming separate between all hidden variables (it is practically impossible to), its form is as follows;
 assume that the Posterior distrbutionp of hidden variable can be analyzed to the distribution product of each Independent Vector, namely assuming separate between all hidden variables (it is practically impossible to), its form is as follows;

8. stratification merchandise news filtered recommendation method as claimed in claim 7, is characterized in that:
 for each user, the parameter γ
of its preference weight distribution_{uk}And the parameter κ
of liveness distribution_{u}Iteration successively by the following step;
 for each user, the parameter γ

9. stratification merchandise news filtered recommendation method as claimed in claim 7, is characterized in that:
 for each commodity, the parameter lambda of its property distribution_{ik}And the parameter τ
of popularity distribution_{i}Iteration successively by the following step;
 for each commodity, the parameter lambda of its property distribution_{ik}And the parameter τ

10. stratification merchandise news filtered recommendation method as claimed in claim 7, is characterized in that:
 auxiliary variable z_{ui}Posteriority multinomial distribution in parameter phi_{ui}Update as follows;
${\mathrm{\φuiProportional;\mathrm{exp}\{\mathrm{psi;({\mathrm{gamma;ukshp){\mathrm{loggamma;ukrte+\mathrm{psi;({\mathrm{lambda;ikshp){\mathrm{loglambda;ikrte\},}}_{}^{}}}_{}^{}}}}_{}^{}}}_{}^{}}}}_{}$ Wherein Ψ
is double;
two gamma functions.
 auxiliary variable z_{ui}Posteriority multinomial distribution in parameter phi_{ui}Update as follows;
Specification(s)