CN109034941A - Products Show method, apparatus, computer equipment and storage medium - Google Patents
Products Show method, apparatus, computer equipment and storage medium Download PDFInfo
- Publication number
- CN109034941A CN109034941A CN201810607862.9A CN201810607862A CN109034941A CN 109034941 A CN109034941 A CN 109034941A CN 201810607862 A CN201810607862 A CN 201810607862A CN 109034941 A CN109034941 A CN 109034941A
- Authority
- CN
- China
- Prior art keywords
- product
- user
- data
- matrix
- rating matrix
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Products Show method, apparatus, computer equipment and storage medium proposed by the present invention, wherein method includes: the product data of the user data and some product that obtain some user, and user is obtained to the first rating matrix of product and the second rating matrix for obtaining product to user according to the product data according to the user data, the user data is data of some user to all products, and the product data are data of some product to all users；First rating matrix is input in preset first matrix of depths decomposition model and be calculated user characteristics, and second rating matrix is input to and carries out that product feature is calculated in preset second matrix of depths decomposition model.Products Show method, apparatus, computer equipment and storage medium proposed by the present invention make full use of all information of all users, accurately predict user to the interest level of different product, recommend reasonable product for user.
Description
Technical field
The present invention relates to field of computer technology, a kind of Products Show method, apparatus is especially related to, computer is set
Standby and storage medium.
Background technique
With the rapid development of society, user's recommender system is to provide a user commodity using the historical data information of user
Information and suggestion, help user to determine what product bought, and pseudo sale personnel help client to complete purchasing process.Individual character
Change and recommend to be Characteristic of Interest and the buying behavior according to user, to the interested information of user recommended user and commodity.But it is existing
Some user's recommender systems mostly use greatly implicit feedback as input, and such disadvantage is to will cause information loss；It is existing
User's recommender system is also faced with Sparse Problem, thus provide it is a kind of can make full use of all information of all users, accurately
User is predicted to the interest level of different product, recommends reasonable product to become urgent problem to be solved for user.
Summary of the invention
The main object of the present invention is to provide a kind of Products Show method, apparatus, computer equipment and storage medium, sufficiently
Using all information of all users, user is accurately predicted to the interest level of different product, is recommended for user reasonable
Product.
Products Show method of the invention, comprising:
The user data of some user and the product data of some product are obtained, and user is obtained according to the user data
The first rating matrix to product and product is obtained to the second rating matrix of user, the user according to the product data
Data are data of some user to all products, and the product data are data of some product to all users；
First rating matrix is input to and carries out that user is calculated in preset first matrix of depths decomposition model
Feature, and second rating matrix is input to and carries out that product is calculated in preset second matrix of depths decomposition model
Feature；
The user characteristics and product feature are input in preset formula be calculated the user characteristics and
The similarity value of the product feature；
Judge whether the similarity value is greater than preset value；
If so, recommending the product to the user.
Further, the generation method of first rating matrix, comprising:
All products are arranged in the preference of multiple dimensions according to user and correspond to for the user data for obtaining some user
Grading parameters construct to obtain first rating matrix.
Further, the generation method of second rating matrix is rapid, comprising:
All users are arranged in the appropriate of multiple dimensions according to product and correspond to for the product data for obtaining some product
Grading parameters construct to obtain second rating matrix.
Further, described first rating matrix is input in preset first matrix of depths decomposition model carries out
User characteristics are calculated, and second rating matrix is input in preset second matrix of depths decomposition model and is carried out
Before the step of product feature is calculated, comprising:
Data extending processing is carried out respectively to first rating matrix and the second rating matrix.
Further, institute is calculated in the described user characteristics and product feature are input in preset formula
The step of stating the similarity value of user characteristics and the product feature, comprising:
According to formulaThe similarity value of the user characteristics and the product feature is calculated, wherein pi
For the user characteristics, qjFor the product feature,For the similarity value.
Further, described the step of recommending the product to the user, comprising:
The similarity value is matched with preset recommendation grade table, the recommendation grade table includes different similarities
It is worth the corresponding relationship of range and recommendation grade；
Recommendation grade is exported according to matching result；
Recommend the product to the user according to the recommendation grade.
Further, the side for carrying out data extending processing respectively to first rating matrix and the second rating matrix
Method, comprising:
Data extending is carried out in a manner of part matrix to extract in first rating matrix and the second rating matrix.
Products Show device proposed by the present invention, comprising:
Structural unit, for obtaining the user data of some user and the product data of some product, and according to the use
User data obtains user to the first rating matrix of product and obtains product according to the product data and comment the second of user
Sub-matrix, the user data are some user to the data of all products, and the product data are some product to useful
The data at family；
First computing unit, for first rating matrix to be input in preset first matrix of depths decomposition model
It carries out that user characteristics are calculated, and second rating matrix is input in preset second matrix of depths decomposition model
It carries out that product feature is calculated；
Second computing unit carries out calculating in preset formula for being input to the user characteristics and product feature
To the similarity value of the user characteristics and the product feature；
Judging unit, for judging whether the similarity value is greater than preset value；
Execution unit, for when the similarity value is greater than preset value, then recommending the product to the user.
Computer equipment proposed by the present invention, including memory and processor, the memory are stored with computer program,
It is characterized in that, the step of processor realizes the above method when executing the computer program.
Computer readable storage medium proposed by the present invention, is stored thereon with computer program, which is characterized in that the meter
The step of above method is realized when calculation machine program is executed by processor.
The invention has the benefit that user characteristics that the Products Show method according to the present invention obtains and described
The similarity value of product feature, it is bigger when similarity value, indicate that the user is interested in the product, when the similarity value
When greater than preset value, then recommend the product to the user, to realize the much information and product characteristic for making full use of user
Information accurately predicts user to the interest level of product, recommends reasonable product for user, improve user experience.
Detailed description of the invention
Fig. 1 is the step schematic diagram of the Products Show method in one embodiment of the invention；
Fig. 2 is the step schematic diagram of the Products Show method in another embodiment of the present invention；
Fig. 3 is the structural schematic diagram of the Products Show device in one embodiment of the invention；
Fig. 4 is the structural schematic diagram of the Products Show device in another embodiment of the present invention；
Fig. 5 is the structural schematic diagram of the execution unit of the Products Show device in one embodiment of the invention；
Fig. 6 is the structural schematic block diagram of the computer equipment of one embodiment of the invention.
The embodiments will be further described with reference to the accompanying drawings for the realization, the function and the advantages of the object of the present invention.
Specific embodiment
It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not intended to limit the present invention.
Referring to Fig.1, the Products Show method in the present embodiment, comprising:
Step S1 obtains the user data of some user and the product data of some product, and according to the user data
User is obtained to the first rating matrix of product and the second rating matrix for obtaining product to user according to the product data,
The user data is data of some user to all products, and the product data are number of some product to all users
According to；
First rating matrix is input in preset first matrix of depths decomposition model and calculate by step S2
To user characteristics, and second rating matrix is input in preset second matrix of depths decomposition model and calculate
To product feature；
The user characteristics and product feature are input to and carry out that the user is calculated in preset formula by step S3
The similarity value of feature and the product feature；
Step S4, judges whether the similarity value is greater than preset value；
Step S5, if so, recommending the product to the user.
In step sl, the user data of some user and the product data of some product are obtained, wherein the said goods are
The different types of insurance products of insurance company, the product data of some above-mentioned product include that the product is applicable in all users
It takes in situation, be applicable in the multiple products characteristic information such as age and applicable occupation；User data includes the user to all products
The much informations such as the quantity of evaluation, purchase product after a purchase, the time for browsing the number of product and browsing product.It is right
In above-mentioned user data and product data, according to user to evaluation of all products after product purchase, buy the number of product
The preference of amount, the number for browsing product and the time multiple dimensions for browsing product is arranged corresponding grading parameters and is configured to
To the first rating matrix, and it is multiple according to income situation, applicable age and the applicable occupation that product is applicable in all users
The appropriate of dimension is arranged corresponding grading parameters and constructs to obtain the second rating matrix.
In step s 2, above-mentioned first rating matrix is input in preset first matrix of depths decomposition model and is counted
Calculation obtains user characteristics, wherein above-mentioned user characteristics are matrix.Above-mentioned preset first matrix of depths decomposition model needs to carry out
Training, the mode that above-mentioned preset first matrix of depths decomposition model is trained are the first rating matrix by specified amount,
And user characteristics corresponding to above-mentioned first rating matrix are trained gained, above-mentioned first matrix of depths as sample data
Decomposition model is for calculating user characteristics.Above-mentioned second rating matrix is input in preset second matrix of depths decomposition model
It carries out that product feature is calculated, wherein the said goods feature is also matrix.Above-mentioned preset second matrix of depths decomposition model
It is also required to be trained, the mode that above-mentioned preset second matrix of depths decomposition model is trained is by the second of specified amount
Rating matrix, and the product feature corresponding to above-mentioned second rating matrix are trained gained as sample data, above-mentioned
Second matrix of depths decomposition model is for calculating product feature.
In step s3, above-mentioned user characteristics and product feature are input to and carry out that calculating is calculated in preset formula
As a result, preset formula needs while inputting user characteristics and product feature, wherein above-mentioned preset formula is
piFor above-mentioned user characteristics, qjFor the said goods feature,For the similarity value of above-mentioned user characteristics and the said goods feature, meter
Calculating result isThe similarity value of i.e. above-mentioned user characteristics and the said goods feature.
In step s 4, for the similarity value of obtained above-mentioned user characteristics and the said goods feature, work as similarity value
It is bigger, indicate that the user is interested in the product.Therefore can be judged whether to need to this according to the size of similarity value
User recommends the product, therefore corresponding preset value can be set, and judges whether above-mentioned similarity value is greater than preset value.
In step s 5, when above-mentioned similarity value is greater than preset value, then recommend the product to the user, to realize
The much information and product characteristic information for making full use of user, accurately predict user to the interest level of product, for
Reasonable product is recommended at family, improves user experience.
Products Show method in the present embodiment, the generation method of first rating matrix, comprising:
All products are arranged in the preference of multiple dimensions according to user and correspond to for the user data for obtaining some user
Grading parameters construct to obtain first rating matrix.
The generating mode of above-mentioned first rating matrix is specific, including will first obtain the user data of some user, according to
Evaluation, the quantity of purchase product, the number of browsing product and the time for browsing product etc. of all products in family after a purchase
The preference of multiple dimensions is arranged corresponding grading parameters and constructs to obtain above-mentioned first rating matrix.For example, user by certain
Evaluation after a product purchase, which can be subdivided into, the differences etc. such as enjoys a lot, likes, generally liking, disliking and disliking very much
The evaluation of grade, for user to the different grades of evaluation of some product, corresponding grading parameters can be set gradually as 5,4,
3, the different grading parameters such as 2 and 1 are realized and grading parameters are arranged to the evaluation of some product according to user.For buying the production
The quantity of product, can according to purchase more than four times, purchase three times, the different purchase such as purchase is secondary, purchase is primary and does not buy
Number is bought, corresponding grading parameters are set gradually as the different grading parameters such as 5,4,3,2 and 1, is realized according to buying the production
Grading parameters are arranged in the quantity of product.Similarly, for information such as times for browsing the number of the product and browsing the product,
Different grading parameters to be arranged using the above method, details are not described herein.For user is arranged under above-mentioned multiple dimensions
Scoring vector is obtained to the grading parameters of some product, grading parameters of the user to all products are constructed to obtain above-mentioned first and are commented
Sub-matrix.
Products Show method in the present embodiment, the generation method of second rating matrix, comprising:
All users are arranged in the appropriate of multiple dimensions according to product and correspond to for the product data for obtaining some product
Grading parameters construct to obtain second rating matrix.
The generating mode of above-mentioned second rating matrix is specific, including will first obtain the product data of some product, according to this
Income situation that product is applicable in all users, the appropriate setting pair for multiple dimensions such as being applicable in the age and be applicable in occupation
The grading parameters answered construct to obtain above-mentioned second rating matrix.The selling price of some product is different, the income feelings being applicable in
The user of condition will be different, such as the price of some product is very high, and for the user of the different levels of consumption, corresponding grading parameters will
Difference, the user strong for consuming capacity set 1 for grading parameters of the product to user, the user weak for consuming capacity,
0 is set by grading parameters of the product to user.The applicable age for some product is 20 years old to 30 years old, for being in 20 years old
To the user of 30 years old age bracket, then 1 is set by grading parameters of the product to the user, for being not at 20 years old to 30 years old age
The user of section, then set 0 for grading parameters of the product to the user.Similarly, it for being applicable in the information such as occupation, can use
The above method is arranged different grading parameters, and details are not described herein.For product is arranged under above-mentioned multiple dimensions to some
The grading parameters of user obtain scoring vector, and grading parameters of the product to all users are constructed to obtain above-mentioned second scoring square
Battle array.
Referring to Fig. 2, Products Show method in another embodiment, it is described first rating matrix is input to it is preset
It carries out that user characteristics are calculated in first matrix of depths decomposition model, and second rating matrix is input to preset
Be calculated before the step S2 of product feature in second matrix of depths decomposition model, comprising:
Step 201, data extending processing is carried out respectively to first rating matrix and the second rating matrix.
It is preset above-mentioned first rating matrix to be input to for above-mentioned first rating matrix and the second rating matrix
It carries out calculating in first matrix of depths decomposition model and above-mentioned second rating matrix is input to preset second matrix of depths
Before being calculated in decomposition model, data extending can be carried out respectively to above-mentioned first rating matrix and the second rating matrix
Processing, technology transform size increase training set size, its object is to avoid the generation of over-fitting, to improve the first depth
Matrix decomposition model and the accuracy of the second matrix of depths decomposition model prediction.For example, being counted for the first rating matrix
It is first technology transform of the extraction section matrix as expansion in the first rating matrix according to the concrete mode for expanding processing, then will
First technology transform is input in the first matrix of depths decomposition model together with the first rating matrix and is calculated；It scores second
The method that matrix carries out data extending processing is identical with the above method, and details are not described herein.
Products Show method in the present embodiment, it is described that the user characteristics and product feature are input to preset formula
The step S3 of the middle similarity value for be calculated the user characteristics and the product feature, comprising:
According to formulaThe similarity value of the user characteristics and the product feature is calculated, wherein pi
For the user characteristics, qjFor the product feature,For the similarity value.
For the user characteristics and product feature of input, it is special to need to carry out to be calculated above-mentioned user according to preset formula
The similarity value for the said goods feature of seeking peace, above-mentioned preset formula areWherein piFor above-mentioned user characteristics, qjFor
The said goods feature,For the similarity value of above-mentioned user characteristics and the said goods feature, above-mentioned similarity value is used to indicate
Family is to the interested degree of product, when similarity value is bigger, illustrates that user is higher to the interested degree of product.
Products Show method in the present embodiment, the step S5 for recommending the product to the user, comprising:
Step S51 matches the similarity value with preset recommendation grade table, and the recommendation grade table includes not
With the corresponding relationship of similarity value range and recommendation grade；
Step S52 exports recommendation grade according to matching result；
Step S53 recommends the product to the user according to the recommendation grade.
The similarity value of above-mentioned user characteristics and the said goods feature for output, can also be emerging to product sense by user
The similarity value of interest is matched with preset recommendation grade table, and above-mentioned recommendation grade table includes different similarity value ranges and push away
The corresponding relationship of grade is recommended, whereinValue it is bigger, indicate user it is higher to the interested degree of the product；It is tied according to matching
Fruit exports corresponding recommendation grade, wherein recommending can be the different recommendation grades such as to highly recommend and recommend, is pushed away according to above-mentioned
It recommends grade and recommends the product to the user, realize more humane for the reasonable product of user's recommendation.
Products Show device referring to Fig. 3, in the present embodiment, comprising:
Structural unit 10, for obtaining the user data of some user and the product data of some product, and according to described
User data obtains user and obtains product to the second of user to the first rating matrix of product and according to the product data
Rating matrix, the user data are data of some user to all products, and the product data are some product to all
The data of user；
First computing unit 20, for first rating matrix to be input to preset first matrix of depths decomposition model
In carry out that user characteristics are calculated, and second rating matrix is input to preset second matrix of depths decomposition model
In carry out that product feature is calculated；
Second computing unit 30, calculates for the user characteristics and product feature to be input in preset formula
Obtain the similarity value of the user characteristics and the product feature；
Judging unit 40, for judging whether the similarity value is greater than preset value；
Execution unit 50, for when the similarity value is greater than preset value, then recommending the product to the user.
The user data of some user and the product data of some product are obtained, wherein the said goods are that insurance company is different
The insurance products of type, the product data of some above-mentioned product include the income situation, suitable that the product is applicable in all users
With the multiple products characteristic information such as age and applicable occupation；User data include the user to all products after a purchase
The much informations such as evaluation, the quantity for buying product, the number for browsing product and the time for browsing product.For above-mentioned number of users
According to and product data, structural unit 10 according to user to all products product purchase after evaluation, buy product quantity,
The corresponding grading parameters of preference setting for browsing the number of product and browsing time multiple dimensions of product construct to obtain
First rating matrix, and income situation, applicable age and the applicable multiple dimensions of occupation that all users are applicable according to product
The appropriate of degree is arranged corresponding grading parameters and constructs to obtain the second rating matrix.
First computing unit 20 by above-mentioned first rating matrix be input in preset first matrix of depths decomposition model into
User characteristics are calculated in row, wherein above-mentioned user characteristics are matrix.Above-mentioned preset first matrix of depths decomposition model needs
It is trained, the mode that above-mentioned preset first matrix of depths decomposition model is trained is the first scoring square by specified amount
User characteristics corresponding to battle array and above-mentioned first rating matrix are trained gained, above-mentioned first depth as sample data
Matrix decomposition model is for calculating user characteristics.Above-mentioned second rating matrix is input to preset second matrix of depths and decomposes mould
It carries out that product feature is calculated in type, wherein the said goods feature is also matrix.Above-mentioned preset second matrix of depths is decomposed
Model is also required to be trained, and the mode that above-mentioned preset second matrix of depths decomposition model is trained is by specified amount
Second rating matrix, and the product feature corresponding to above-mentioned people's rating matrix are trained gained as sample data,
Above-mentioned second matrix of depths decomposition model is for calculating product feature.
Above-mentioned user characteristics and product feature are input in preset formula by the second computing unit 30 to be calculated
Calculated result, preset formula needs while inputting user characteristics and product feature, wherein above-mentioned preset formula ispiFor above-mentioned user characteristics, qjFor the said goods feature,For the phase of above-mentioned user characteristics and the said goods feature
Like angle value, calculated result isThe similarity value of i.e. above-mentioned user characteristics and the said goods feature.
For the similarity value of obtained above-mentioned user characteristics and the said goods feature, when bigger, the expression of similarity value
The user is interested in the product.Therefore can be judged whether to need to recommend the production to the user according to the size of similarity value
Product, therefore corresponding preset value can be set, judging unit 40 judges whether above-mentioned similarity value is greater than preset value.
When above-mentioned similarity value is greater than preset value, execution unit 50 then recommends the product to the user, to realize
The much information and product characteristic information for making full use of user, accurately predict user to the interest level of product, for
Reasonable product is recommended at family, improves user experience.
Products Show device in the present embodiment, the generation method of first rating matrix, comprising:
All products are arranged in the preference of multiple dimensions according to user and correspond to for the user data for obtaining some user
Grading parameters construct to obtain first rating matrix.
The generating mode of above-mentioned first rating matrix is specific, including will first obtain the user data of some user, according to
Evaluation, the quantity of purchase product, the number of browsing product and the time for browsing product etc. of all products in family after a purchase
The preference of multiple dimensions is arranged corresponding grading parameters and constructs to obtain above-mentioned first rating matrix.For example, user by certain
Evaluation after a product purchase, which can be subdivided into, the differences etc. such as enjoys a lot, likes, generally liking, disliking and disliking very much
The evaluation of grade, for user to the different grades of evaluation of some product, corresponding grading parameters can be set gradually as 5,4,
3, the different grading parameters such as 2 and 1 are realized and grading parameters are arranged to the evaluation of some product according to user.For buying the production
The quantity of product, can according to purchase more than four times, purchase three times, the different purchase such as purchase is secondary, purchase is primary and does not buy
Number is bought, corresponding grading parameters are set gradually as the different grading parameters such as 5,4,3,2 and 1, is realized according to buying the production
Grading parameters are arranged in the quantity of product.Similarly, for information such as times for browsing the number of the product and browsing the product,
Different grading parameters to be arranged using the above method, details are not described herein.For user is arranged under above-mentioned multiple dimensions
Scoring vector is obtained to the grading parameters of some product, grading parameters of the user to all products are constructed to obtain above-mentioned first and are commented
Sub-matrix.
Products Show device in the present embodiment, the generation method of second rating matrix, comprising:
All users are arranged in the preference of multiple dimensions according to product and correspond to for the product data for obtaining some product
Grading parameters construct to obtain second rating matrix.
The generating mode of above-mentioned second rating matrix is specific, including will first obtain the product data of some product, according to this
Income situation that product is applicable in all users, the appropriate setting pair for multiple dimensions such as being applicable in the age and be applicable in occupation
The grading parameters answered construct to obtain above-mentioned second rating matrix.The selling price of some product is different, the income feelings being applicable in
The user of condition will be different, such as the price of some product is very high, and for the user of the different levels of consumption, corresponding grading parameters will
Difference, the user strong for consuming capacity set 1 for grading parameters of the product to user, the user weak for consuming capacity,
0 is set by grading parameters of the product to user.The applicable age for some product is 20 years old to 30 years old, for being in 20 years old
To the user of 30 years old age bracket, then 1 is set by grading parameters of the product to the user, for being not at 20 years old to 30 years old age
The user of section, then set 0 for grading parameters of the product to the user.Similarly, it for being applicable in the information such as occupation, can use
The above method is arranged different grading parameters, and details are not described herein.For product is arranged under above-mentioned multiple dimensions to some
The grading parameters of user obtain scoring vector, and grading parameters of the product to all users are constructed to obtain above-mentioned second scoring square
Battle array.
Products Show device referring to Fig. 4, in another embodiment, further includes:
Expansion unit 201, for carrying out data extending processing respectively to first rating matrix and the second rating matrix.
It is preset above-mentioned first rating matrix to be input to for above-mentioned first rating matrix and the second rating matrix
It carries out calculating in first matrix of depths decomposition model and above-mentioned second rating matrix is input to preset second matrix of depths
Before being calculated in decomposition model, expansion unit 201 carries out above-mentioned first rating matrix and the second rating matrix respectively
Data extending processing, technology transform size increase training set size, its object is to avoid the generation of over-fitting, to improve
First matrix of depths decomposition model and the accuracy of the second matrix of depths decomposition model prediction.For example, for the first scoring square
Battle array, the concrete mode for carrying out data extending processing is that extraction section matrix expands as first expanded in the first rating matrix
Matrix, then the first technology transform is input in the first matrix of depths decomposition model together with the first rating matrix and is calculated；
The method for carrying out data extending processing to the second rating matrix is identical with the above method, and details are not described herein.
Products Show device in the present embodiment, second computing unit 30 are specifically used for according to formula
The similarity value of the user characteristics and the product feature is calculated, wherein piFor the user characteristics, qjFor the product
Feature,For the similarity value.
For the user characteristics and product feature of input, it is special to need to carry out to be calculated above-mentioned user according to preset formula
The similarity value for the said goods feature of seeking peace, above-mentioned preset formula areWherein piFor above-mentioned user characteristics, qjFor
The said goods feature,For the similarity value of above-mentioned user characteristics and the said goods feature, above-mentioned similarity value is used to indicate
Family is to the interested degree of product, when similarity value is bigger, illustrates that user is higher to the interested degree of product.
Referring to Fig. 5, Products Show device in the present embodiment, the execution unit 50 includes:
First matching module 51, for the similarity value to be matched with preset recommendation grade table, the recommendation
Table of grading includes the corresponding relationship of different similarity value ranges and recommendation grade；
Second matching module 52, for exporting recommendation grade according to matching result；
Recommending module 53, for recommending the product to the user according to the recommendation grade.
The similarity value of above-mentioned user characteristics and the said goods feature for output, the first matching module 51 can also incite somebody to action
User matches the interested similarity value of product with preset recommendation grade table, above-mentioned
Recommendation grade table includes the corresponding relationship of different similarity value ranges and recommendation grade, whereinValue it is bigger, table
Show that user is higher to the interested degree of the product；Second matching module 52 is according to corresponding recommendation of matching result output etc.
Grade, wherein recommend can for it is strongly recommended that and recommend etc. different recommendation grades, recommending module 53 according to above-mentioned recommendation grade to
The user recommends the product, realizes more humane for the reasonable product of user's recommendation.
Referring to Fig. 6, a kind of computer equipment is also provided in the embodiment of the present invention, which can be server,
Its internal structure can be as shown in Figure 6.The computer equipment includes processor, the memory, network connected by system bus
Interface and database.Wherein, the processor of the Computer Design is for providing calculating and control ability.The computer equipment is deposited
Reservoir includes non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system, computer program
And database.The internal memory provides environment for the operation of operating system and computer program in non-volatile memory medium.It should
The database of computer equipment is for data such as pre-set product recommended methods.The network interface of the computer equipment is used for and outside
Terminal by network connection communication.To realize Products Show method when the computer program is executed by processor.
Above-mentioned processor executes the step of the said goods recommended method: obtaining the user data and some product of some user
Product data, and user is obtained to the first rating matrix of product and according to the said goods data according to above-mentioned user data
Product is obtained to the second rating matrix of user, above-mentioned user data is data of some user to all products, the said goods
Data are data of some product to all users；Above-mentioned first rating matrix is input to preset first matrix of depths to decompose
It carries out that user characteristics are calculated in model, and above-mentioned second rating matrix is input to preset second matrix of depths and is decomposed
It carries out that product feature is calculated in model；Above-mentioned user characteristics and product feature are input in preset formula and are calculated
Obtain the similarity value of above-mentioned user characteristics and the said goods feature；Judge whether above-mentioned similarity value is greater than preset value；If so,
Then recommend the product to the user.
Above-mentioned computer equipment, the generation method of above-mentioned first rating matrix, comprising: obtain the number of users of some user
According to corresponding grading parameters be arranged constructing to obtain above-mentioned first in the preference of multiple dimensions to all products according to user and comment
Sub-matrix.
In one embodiment, the generation method of above-mentioned second rating matrix, comprising: obtain the product number of some product
According to corresponding grading parameters be arranged constructing to obtain above-mentioned second in the appropriate of multiple dimensions to all users according to product and comment
Sub-matrix.
In one embodiment, above-mentioned that above-mentioned first rating matrix is input to preset first matrix of depths decomposition model
In carry out that user characteristics are calculated, and above-mentioned second rating matrix is input to preset second matrix of depths decomposition model
In carry out the step of product feature is calculated before, comprising: to above-mentioned first rating matrix and the second rating matrix respectively into
The processing of row data extending.
In one embodiment, above-mentioned above-mentioned user characteristics and product feature are input in preset formula calculates
The step of obtaining the similarity value of above-mentioned user characteristics and the said goods feature, comprising: according to formulaIt calculates
To the similarity value of above-mentioned user characteristics and the said goods feature, wherein piFor above-mentioned user characteristics, qjFor the said goods feature,For above-mentioned similarity value.
In one embodiment, above-mentioned the step of recommending the product to the user, comprising: by above-mentioned similarity value and preset
Recommendation grade table matched, above-mentioned recommendation grade table includes the corresponding relationship of different similarity value ranges and recommendation grade；
Recommendation grade is exported according to matching result；Recommend the product to the user according to above-mentioned recommendation grade.
In one embodiment, above-mentioned that above-mentioned first rating matrix and the second rating matrix are carried out at data extending respectively
The method of reason, comprising: carry out data in a manner of part matrix to extract in above-mentioned first rating matrix and the second rating matrix
Expand.
It will be understood by those skilled in the art that structure shown in Fig. 6, only part relevant to application scheme is tied
The block diagram of structure does not constitute the restriction for the computer equipment being applied thereon to application scheme.
One embodiment of the invention also provides a kind of computer readable storage medium, is stored thereon with computer program, calculates
Machine program realizes a kind of Products Show method when being executed by processor, specifically: obtain some user user data and some
The product data of product, and user is obtained to the first rating matrix of product and according to the said goods according to above-mentioned user data
Data obtain product to the second rating matrix of user, and above-mentioned user data is data of some user to all products, above-mentioned
Product data are data of some product to all users；Above-mentioned first rating matrix is input to preset first matrix of depths
It carries out that user characteristics are calculated in decomposition model, and above-mentioned second rating matrix is input to preset second matrix of depths
It carries out that product feature is calculated in decomposition model；Above-mentioned user characteristics and product feature are input in preset formula and are carried out
The similarity value of above-mentioned user characteristics and the said goods feature is calculated；Judge whether above-mentioned similarity value is greater than preset value；
If so, recommending the product to the user.
Above-mentioned computer readable storage medium, by the generation method of above-mentioned first rating matrix, comprising: obtain some user
User data, corresponding grading parameters are arranged in the preference of multiple dimensions to all products according to user and construct to obtain
State the first rating matrix.
In one embodiment, the generation method of above-mentioned second rating matrix, comprising: obtain the product number of some product
According to corresponding grading parameters be arranged constructing to obtain above-mentioned second in the appropriate of multiple dimensions to all users according to product and comment
Sub-matrix.
In one embodiment, above-mentioned that above-mentioned first rating matrix is input to preset first matrix of depths decomposition model
In carry out that user characteristics are calculated, and above-mentioned second rating matrix is input to preset second matrix of depths decomposition model
In carry out the step of product feature is calculated before, comprising: to above-mentioned first rating matrix and the second rating matrix respectively into
The processing of row data extending.
In one embodiment, above-mentioned above-mentioned user characteristics and product feature are input in preset formula calculates
The step of obtaining the similarity value of above-mentioned user characteristics and the said goods feature, comprising: according to formulaIt calculates
To the similarity value of above-mentioned user characteristics and the said goods feature, wherein piFor above-mentioned user characteristics, qjFor the said goods feature,For above-mentioned similarity value.
In one embodiment, above-mentioned the step of recommending the product to the user, comprising: by above-mentioned similarity value and preset
Recommendation grade table matched, above-mentioned recommendation grade table includes the corresponding relationship of different similarity value ranges and recommendation grade；
Recommendation grade is exported according to matching result；Recommend the product to the user according to above-mentioned recommendation grade.
In one embodiment, above-mentioned that above-mentioned first rating matrix and the second rating matrix are carried out at data extending respectively
The method of reason, comprising: carry out data in a manner of part matrix to extract in above-mentioned first rating matrix and the second rating matrix
Expand.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can store and a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
Any reference used in provided herein and embodiment to memory, storage, database or other media,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, mono- diversified forms of RAM can obtain,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double speed are according to rate SDRAM (SSRSDRAM), enhancing
Type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
In conclusion above-mentioned user characteristics that the Products Show method according to the present invention obtains and the said goods feature
Similarity value, it is bigger when similarity value, indicate that the user is interested in the product, when above-mentioned similarity value is greater than preset value
When, then recommend the product to the user, to realize the much information for making full use of user and product characteristic information, accurately
User is predicted to the interest level of product, recommends reasonable product for user, improves user experience.
The above description is only a preferred embodiment of the present invention, is not intended to limit the scope of the invention, all utilizations
Equivalent structure or equivalent flow shift made by description of the invention and accompanying drawing content is applied directly or indirectly in other correlations
Technical field, be included within the scope of the present invention.
Claims (10)
1. a kind of Products Show method characterized by comprising
The user data of some user and the product data of some product are obtained, and user is obtained to production according to the user data
First rating matrix of product and product is obtained to the second rating matrix of user, the user data according to the product data
It is some user to the data of all products, the product data are data of some product to all users；
First rating matrix is input to and carries out that user characteristics are calculated in preset first matrix of depths decomposition model,
And second rating matrix is input to and carries out that product feature is calculated in preset second matrix of depths decomposition model；
The user characteristics and product feature are input in preset formula and be calculated user characteristics and described
The similarity value of product feature；
Judge whether the similarity value is greater than preset value；
If so, recommending the product to the user.
2. Products Show method according to claim 1, which is characterized in that the generation method of first rating matrix,
Include:
Corresponding comment is arranged in the user data for obtaining some user, the preference according to user to all products in multiple dimensions
Parametric configuration is divided to obtain first rating matrix.
3. Products Show method according to claim 1, which is characterized in that the generation method of second rating matrix,
Include:
Corresponding comment is arranged in the appropriate of multiple dimensions to all users according to product in the product data for obtaining some product
Parametric configuration is divided to obtain second rating matrix.
4. Products Show method according to claim 1, which is characterized in that described to be input to first rating matrix
It carries out that user characteristics are calculated in preset first matrix of depths decomposition model, and second rating matrix is input to
Before carrying out the step of product feature is calculated in preset second matrix of depths decomposition model, comprising:
Data extending processing is carried out respectively to first rating matrix and the second rating matrix.
5. Products Show method according to claim 1, which is characterized in that described by the user characteristics and product feature
The step of being input to the similarity value for carrying out being calculated the user characteristics and the product feature in preset formula, packet
It includes:
According to formulaThe similarity value of the user characteristics and the product feature is calculated, wherein piFor institute
State user characteristics, qjFor the product feature,For the similarity value.
6. Products Show method according to claim 1, which is characterized in that the step for recommending the product to the user
Suddenly, comprising:
The similarity value is matched with preset recommendation grade table, the recommendation grade table includes different similarity value models
Enclose the corresponding relationship with recommendation grade；
Recommendation grade is exported according to matching result；
Recommend the product to the user according to the recommendation grade.
7. Products Show method according to claim 4, which is characterized in that described to first rating matrix and second
The method that rating matrix carries out data extending processing respectively, comprising:
Data extending is carried out in a manner of part matrix to extract in first rating matrix and the second rating matrix.
8. a kind of Products Show device characterized by comprising
Structural unit, for obtaining the user data of some user and the product data of some product, and according to the number of users
According to obtaining user to the first rating matrix of product and obtain product according to the product data and score square to the second of user
Battle array, the user data are data of some user to all products, and the product data are some product to all users'
Data；
First computing unit is carried out for first rating matrix to be input in preset first matrix of depths decomposition model
User characteristics are calculated, and second rating matrix is input in preset second matrix of depths decomposition model and is carried out
Product feature is calculated；
Institute is calculated for the user characteristics and product feature to be input in preset formula in second computing unit
State the similarity value of user characteristics and the product feature；
Judging unit, for judging whether the similarity value is greater than preset value；
Execution unit, for when the similarity value is greater than preset value, then recommending the product to the user.
9. a kind of computer equipment, including memory and processor, the memory are stored with computer program, feature exists
In the step of processor realizes any one of claims 1 to 7 the method when executing the computer program.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
The step of method described in any one of claims 1 to 7 is realized when being executed by processor.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810607862.9A CN109034941B (en) | 2018-06-13 | 2018-06-13 | Product recommendation method and device, computer equipment and storage medium |
| PCT/CN2018/108033 WO2019237569A1 (en) | 2018-06-13 | 2018-09-27 | Product recommendation method and apparatus, and computer device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810607862.9A CN109034941B (en) | 2018-06-13 | 2018-06-13 | Product recommendation method and device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109034941A true CN109034941A (en) | 2018-12-18 |
| CN109034941B CN109034941B (en) | 2023-03-31 |
Family
ID=64609539
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810607862.9A Active CN109034941B (en) | 2018-06-13 | 2018-06-13 | Product recommendation method and device, computer equipment and storage medium |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN109034941B (en) |
| WO (1) | WO2019237569A1 (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110083688A (en) * | 2019-05-10 | 2019-08-02 | 北京百度网讯科技有限公司 | Search result recalls method, apparatus, server and storage medium |
| CN110598040A (en) * | 2019-09-17 | 2019-12-20 | 上海喜马拉雅科技有限公司 | Album recalling method, device, equipment and storage medium |
| CN113344673A (en) * | 2021-06-28 | 2021-09-03 | 平安信托有限责任公司 | Product pushing method and device, electronic equipment and storage medium |
| CN115880076A (en) * | 2022-04-20 | 2023-03-31 | 北京中关村科金技术有限公司 | Trusted product recommendation method, device and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102426686A (en) * | 2011-09-29 | 2012-04-25 | 南京大学 | Internet information product recommending method based on matrix decomposition |
| CN103246672A (en) * | 2012-02-09 | 2013-08-14 | 中国科学技术大学 | Method and device for performing personalized recommendation on users |
| CN106570008A (en) * | 2015-10-09 | 2017-04-19 | 阿里巴巴集团控股有限公司 | Recommendation method and device |
| CN107563841A (en) * | 2017-08-03 | 2018-01-09 | 电子科技大学 | A kind of commending system decomposed that scored based on user |
| CN107578270A (en) * | 2017-08-03 | 2018-01-12 | ***股份有限公司 | A kind of construction method, device and the computing device of financial label |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663627A (en) * | 2012-04-26 | 2012-09-12 | 焦点科技股份有限公司 | Personalized recommendation method |
| CN104021163B (en) * | 2014-05-28 | 2017-10-24 | 深圳市盛讯达科技股份有限公司 | Products Show system and method |
| CN107292648A (en) * | 2016-04-01 | 2017-10-24 | 株式会社理光 | A kind of user behavior analysis method and device |
| CN107562795A (en) * | 2017-08-01 | 2018-01-09 | 广州市香港科大***研究院 | Recommendation method and device based on Heterogeneous Information network |
-
2018
- 2018-06-13 CN CN201810607862.9A patent/CN109034941B/en active Active
- 2018-09-27 WO PCT/CN2018/108033 patent/WO2019237569A1/en active Application Filing
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102426686A (en) * | 2011-09-29 | 2012-04-25 | 南京大学 | Internet information product recommending method based on matrix decomposition |
| CN103246672A (en) * | 2012-02-09 | 2013-08-14 | 中国科学技术大学 | Method and device for performing personalized recommendation on users |
| CN106570008A (en) * | 2015-10-09 | 2017-04-19 | 阿里巴巴集团控股有限公司 | Recommendation method and device |
| CN107563841A (en) * | 2017-08-03 | 2018-01-09 | 电子科技大学 | A kind of commending system decomposed that scored based on user |
| CN107578270A (en) * | 2017-08-03 | 2018-01-12 | ***股份有限公司 | A kind of construction method, device and the computing device of financial label |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110083688A (en) * | 2019-05-10 | 2019-08-02 | 北京百度网讯科技有限公司 | Search result recalls method, apparatus, server and storage medium |
| CN110083688B (en) * | 2019-05-10 | 2022-03-25 | 北京百度网讯科技有限公司 | Search result recall method, device, server and storage medium |
| CN110598040A (en) * | 2019-09-17 | 2019-12-20 | 上海喜马拉雅科技有限公司 | Album recalling method, device, equipment and storage medium |
| CN113344673A (en) * | 2021-06-28 | 2021-09-03 | 平安信托有限责任公司 | Product pushing method and device, electronic equipment and storage medium |
| CN115880076A (en) * | 2022-04-20 | 2023-03-31 | 北京中关村科金技术有限公司 | Trusted product recommendation method, device and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2019237569A1 (en) | 2019-12-19 |
| CN109034941B (en) | 2023-03-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109509054B (en) | Commodity recommendation method under mass data, electronic device and storage medium | |
| US10607243B2 (en) | User behavior analysis method and device as well as non-transitory computer-readable medium | |
| Özmen et al. | RMARS: robustification of multivariate adaptive regression spline under polyhedral uncertainty | |
| CN109034941A (en) | Products Show method, apparatus, computer equipment and storage medium | |
| Wang et al. | A semiparametric gaussian copula regression model for predicting financial risks from earnings calls | |
| Lesage et al. | Uncertainty analysis in LCA using precalculated aggregated datasets | |
| CN106251174A (en) | Information recommendation method and device | |
| CN107492008A (en) | Information recommendation method, device, server and computer-readable storage medium | |
| CN105469263A (en) | Commodity recommendation method and device | |
| CN109582876A (en) | Tourism industry user portrait building method, device and computer equipment | |
| CN104866474A (en) | Personalized data searching method and device | |
| CN111080398A (en) | Commodity recommendation method and device, computer equipment and storage medium | |
| CN105574025A (en) | Methods and devices for sorting score calculation and model building, and commodity recommendation system | |
| CN108205775A (en) | The recommendation method, apparatus and client of a kind of business object | |
| CN109145193A (en) | A kind of information-pushing method and system | |
| CN113191838B (en) | Shopping recommendation method and system based on heterogeneous graph neural network | |
| CN111460316B (en) | Knowledge system-oriented personalized recommendation method and computer storage medium | |
| CN104346428A (en) | Information processing apparatus, information processing method, and program | |
| CN104915440A (en) | Commodity de-duplication method and system | |
| Shah | Book recommendation system using item based collaborative filtering | |
| US20200074324A1 (en) | Noise contrastive estimation for collaborative filtering | |
| CN111161009B (en) | Information pushing method, device, computer equipment and storage medium | |
| CN115641179A (en) | Information pushing method and device and electronic equipment | |
| CN113763031A (en) | Commodity recommendation method and device, electronic equipment and storage medium | |
| CN110889748B (en) | Store platform product recommendation method, store platform product recommendation device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |