CN106204124A - Personalized commercial coupling commending system and method - Google Patents
Personalized commercial coupling commending system and method Download PDFInfo
- Publication number
- CN106204124A CN106204124A CN201610519895.9A CN201610519895A CN106204124A CN 106204124 A CN106204124 A CN 106204124A CN 201610519895 A CN201610519895 A CN 201610519895A CN 106204124 A CN106204124 A CN 106204124A
- Authority
- CN
- China
- Prior art keywords
- commodity
- data
- color
- user
- shape
- 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.)
- Pending
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
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0269—Targeted advertisements based on user profile or attribute
- G06Q30/0271—Personalized advertisement
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Finance (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of personalized commercial coupling commending system, including: data acquisition unit, it is used for gathering commodity data and user data；Data storage device, is used for storing described commodity data and user data；And goods matching device.Goods matching device includes: color subsystem, obtains the matching degree between commodity and user for color data based on commodity；Shape subsystem, obtains the matching degree between commodity and user for shapes based on commodity；And style subsystem, obtain the matching degree between commodity and user for styles based on commodity.According to said system, three dimensions such as colors based on commodity, shape, style set up model to commodity, can provide mating of user and commodity on this basis, and treatment effeciency is high and matching effect is good.
Description
Technical field
The present invention relates to commercial product recommending technology, particularly relate to personalized commercial coupling commending system and method.
Background technology
It is quite ripe that current ecommerce has developed, and a lot of users can be directly at online purchase clothes, footwear
The various dress ornament commodity such as cap, although shopping online is convenient, have the shortcoming that cannot on-the-spot try on.Business such as online purchase dress ornaments
The picture mainly issued by browsing businessman when of product judges whether to be suitable for oneself, then buys and tries on, inappropriate
Time must carry out handling return, owing to the personal attribute of commodity affects relatively big, cause return of goods rate higher.If can promote
Recommend the matching degree between the commodity of user and user, the probability that user successfully buys can be increased undoubtedly, reduce return of goods rate.
Ecommerce, nonstandard class commodity (product that i.e. the aesthetics design class such as clothing, shoes and hats, jewelry is strong) are because aesthstic data
Cannot standardized transmission, therefore can band incoming call business's the operation pressure increase, system two ends (user side, kinds of goods end) both sides had hundred million
Data volume wait the accurate coupling carrying out data in the way of operation, therefore, from aesthetic theory, set up brand-new data pair
Stream mode, reaches and carries out personalized recommendation for each different people, improves electricity business's efficiency of operation, it is provided that user completely newly optimizes
More diversified shopping guide experiences.
Summary of the invention
For the deficiencies in the prior art, it is an object of the invention to provide a kind of personalized commercial coupling commending system and side
Method, can be that individual provides the commercial product recommending of higher matching degree, commodity data to distribute function.
To achieve these goals, the technical scheme is that
A kind of personalized commercial coupling commending system, including:
Data acquisition unit, is used for gathering commodity data and user data, and described commodity data includes color data, shape
Shape data, sized data and image, described user data includes figure and features data, preference data and social mentality's data；
Data storage device, is used for storing described commodity data and user data；
Goods matching device, including:
Color subsystem, for setting up color model according to described commodity data, described color model is according to described commodity
Color image label corresponding to color data output, and calculate described based on described color image label and described user data
Matching degree between commodity and user；
Shape subsystem, for setting up shape according to described commodity data, described shape is according to this business described
The entirety of the shape data output commodity of product and/or the shape type of parts, and shape class based on described entirety and/or parts
Type and described user data calculate the matching degree between described commodity and user；
Style subsystem, for setting up Style Model according to described commodity data, described Style Model is according to described commodity
The genre labels that data output is corresponding, and calculate between described commodity and user according to described genre labels and described user data
Matching degree；And
Coupling recommending module, for the output of comprehensive described color subsystem, shape subsystem and style subsystem
Matching degree data are to obtain the final matching results between described commodity and user.
A kind of personalized commercial coupling recommendation method, including:
Gathering and store commodity data and user data, described commodity data includes color data, shape data, size
Data and image, described user data includes the bodily form/figure and features data, preference data and social mentality's data；
Setting up color model according to described commodity data, described color model determines color according to the color data of described commodity
The numerical range of color rule, and export the color image label of correspondence accordingly, and based on described color image label and described use
User data calculates the matching degree between described commodity and user；
Setting up shape according to described commodity data, described shape exports according to the shape data of these commodity described
The entirety of commodity and/or the shape type of parts, and shape type based on described entirety and/or parts and described user data
Calculate the matching degree between described commodity and user；
Setting up Style Model according to described commodity data, described Style Model is according to the wind of described commodity data output correspondence
Case marker label, and calculate the matching degree between described commodity and user according to described genre labels and described user data；And
Comprehensive matching degree data based on color, shape and style are to obtain final between described commodity and user
Join result.
According to the technical scheme of the present embodiment, three dimensions such as colors based on commodity, shape, style set up mould to commodity
Type, and further and user model carries out carrying out matching operation between matching operation or two commodity, the most permissible
Offer user is mated with commodity, and treatment effeciency is high and matching effect is good.
Accompanying drawing explanation
The schematic diagram of the personalized commercial coupling commending system that Fig. 1 provides for the embodiment of the present invention.
Fig. 2 is the schematic diagram of the color subsystem of the personalized commercial coupling commending system of Fig. 1.
Fig. 3 is the schematic diagram of the hue circle of the color subsystem employing of Fig. 2.
Fig. 4 is the schematic diagram of the shape subsystem of the personalized commercial coupling commending system of Fig. 1.
Fig. 5 is the schematic diagram of the style subsystem of the personalized commercial coupling commending system of Fig. 1.
Fig. 6 is the goods model schematic diagram of the personalized commercial coupling commending system foundation of Fig. 1.
Fig. 7 is the schematic diagram of the user model of the employing of the personalized commercial coupling recommendation of Fig. 1.
The commodity data coupling that Fig. 8 provides for embodiment 1 recommends method flow diagram.
The block diagram of the commodity data coupling commending system that Fig. 9 provides for embodiment 2.
The flow chart of the matching process of the user based on color that Figure 10 provides for embodiment 3 and commodity.
The block diagram of the matching system of the user based on color that Figure 11 provides for embodiment 4 and commodity.
The flow chart of the tie-in sale method based on color that Figure 12 provides for embodiment 5.
The block diagram of the tie-in sale system based on color that Figure 13 provides for embodiment 6.
The flow chart of the matching process of the user based on shape that Figure 14 provides for embodiment 7 and commodity.
The block diagram of the matching system of the user based on shape that Figure 15 provides for embodiment 8 and commodity.
The flow chart of the tie-in sale method based on shape that Figure 16 provides for embodiment 9.
The block diagram of the tie-in sale system based on shape that Figure 17 provides for embodiment 10.
Detailed description of the invention
Describe the preferred embodiment of the present invention below in conjunction with the accompanying drawings in detail.
Fig. 1 shows the example structure schematic diagram of personalized commercial of the present invention coupling commending system, as it is shown in figure 1, this
Bright embodiment provides a kind of personalized commercial coupling commending system, including: data acquisition unit 10, data storage device 20,
Join recommendation apparatus 30 and client 40.
Data acquisition unit 10 can be such as that server, computer, notebook computer, smart mobile phone, portable image are adopted
Collection instrument etc., it is also possible to be the data collecting system collectively formed by client device and server.Data acquisition unit is used for adopting
Collection commodity data.The commodity data gathered includes, but are not limited to the size of commodity, picture, description label, title, classification etc..
The commodity data that data acquisition unit 10 gathers can be stored in data storage device 20.Except data acquisition packaging
Putting the data of 10 collections, data storage device 20 can also store other data, such as, user data, user data include but
It is not limited to the stature type of user, figure and features data, commodity preference data, sex, age, occupation, psychological research questionnaire result
Deng.The bodily form/figure and features data include, but are not limited to: the shape of face and size, height, body weight, waistline, hip circumference, shoulder breadth, arm
Length, lower limb length, arm are thick, neck diameter, lower limb shape etc. all be related to the data of humanoid state description.Except the data of these static state,
The bodily form of user/figure and features data can also be expressed with threedimensional model.The threedimensional model of human body can be swept by three-dimensional camera
Retouch acquisition.Further, except user's bodily form/figure and features data can also include some dynamic datas, such as a people stance,
Fiddle with the custom etc. of hair.
The coupling recommended models of the coupling built-in commodity of recommendation apparatus 30, for realizing business according to commodity data and user data
Mating and mating between commodity with commodity between product with user, thus the commodity data realizing personalization distributes service,
Check for user 50 so that the data meeting demand are sent to client 40.
Above-mentioned coupling recommended models comprises the steps that color subsystem 31, shape subsystem 32 and style subsystem 33.
Wherein, color subsystem 31 is recommended for the coupling realizing commodity according to color, and shape subsystem 32 is for realizing according to shape
The coupling of commodity is recommended, and Style Model 33 is recommended for the coupling realizing commodity according to the style of product.The following specifically describes each
The details of model.
Refering to Fig. 2, color subsystem 31 includes: color parsing module 311, genre labels module 312 and user's coupling
Module 313 and tie-in sale module 314.
Color parsing module 311 is for resolving the color of commodity to obtain the color data of each dimension of commodity.Basis
Color data such as can include the form and aspect of each color, lightness, purity, shared area ratio etc..Further, color solution
Analysis module 311 also calculates additional color attribute value by predefined model on the basis of base color data.
Form and aspect, the appearance appellation of the most all kinds of colors, such as bright red, Prussian blue, lemon yellow etc..Form and aspect are the primary features of color,
It it is the standard the most accurately distinguishing various different colors.Color beyond any black-white-gray has the attribute of form and aspect, and form and aspect are also
It is made up of primary colors, a normal complexion secondary color.
The lightness of color refers to the light levels of color.Various colored objects just produce due to the difference of they reflection light quantities
The light and shade of color is strong and weak.The lightness of color has two kinds of situations: one is same form and aspect difference lightness；Two is that the difference of shades of colour is bright
Degree.
The purity of color refers to the purity level of color.It represents the ratio of contained colour content in color, and ratio is bigger,
Color is the purest, and ratio is the least, then the purity of color is the lowest.In addition to three primary colors, the purity of other all of colors all without
Higher than 100%.
Refering to Fig. 3, it is the schematic diagram of Mang Saier (Munsell) color system.Color is divided into eight form and aspect by this system:
Red, yellow, and green red, yellow is yellow, green, bluish-green, blue, royal purple, purple, purple.These 8 form and aspect constitute a complete ring, it is possible to understand that
It is, form and aspect, however it is not limited to 8 two adjacent colors to be mixed again, it is possible to obtain more form and aspect.For lightness,
Can be divided into 10 grades 1-10, wherein 10 represent that lightness is the highest, and 0 represents the lightness end.For purity, can also be divided into equally
10 grades 1-10, wherein 10 represent that purity is the highest, and 0 represents that purity is minimum.It is understood that lightness is not limited to purity
It is divided into 10 grades.By color system as shown in Figure 3, a color just can be mapped in a three-dimensional system of coordinate.
In general, in image processing system, color value be with a kind of colour system such as RGB, CMYK or other
The colour system of meaning is expressed, and according to these color value, can directly obtain the form and aspect of color, lightness and pure through conversion
Angle value.
In a specific embodiment, commodity include two or three color, color parsing module 311 also by this two
Plant or three kinds of colors are placed in a hue circle, it will be understood that each color is corresponding to a point in hue circle.Color resolves
Module 311 can calculate absolute value | θ | of the central angle of adjacent two color dot.
According to above-mentioned base color data and additional color attribute value, color parsing module 311 obtains described business
The color image label of product.Color image label comprises the steps that age label, sex label, dressing occasion label, esthetic attribute mark
Sign and emotion label, wherein:
Age label such as comprises the steps that child's color, young color, old color, middle age color etc., and it is used for describing certain color and is
The no people being suitable for age level corresponding to this label.
Sex label such as comprises the steps that male's color, women color etc., and it is used for describing certain color if appropriate for this label pair
The sex answered.
Dressing occasion label such as comprises the steps that household, party, office, outing etc., and it is used for whether describing certain color
It is suitable for the dressing occasion that this label is corresponding.
Esthetic attribute label such as comprises the steps that simplicity, deep, steady, active etc., and it is corresponding for describing certain color
The aesthetic method of people.
Emotion label such as comprises the steps that excitement, calm, warm etc., and it is for describing the emotion of people corresponding to certain color
Impression.
Furthermore, it is to be understood that the kind of color image label is not restricted to inventory enumerated above.
Specifically, when only including single color in commodity, color parsing module 311 based on following steps will realize from
Color data is to the parsing of color image label:
Lightness and purity are defined as color parsing module 311 multiple continuous print lightness interval respectively and purity is interval.
Such as, if whole lightness interval is 1-10 (value the lowest expression lightness is the lowest), then definable 1-3 is that low-key, 4-7 be low in being, 8-
10 is to a high-profile.Similar, it is also possible to by the height of purity, purity to be defined as low-key (1-3), middle tune (4-7), fresh tune (8-
10)。
On the other hand, color parsing module 311 can predefine or obtain in data base lightness interval and purity interval
Combination and shades of colour image label between mapping table.It is to say, when color data (form and aspect, lightness and purity) determines
, the color image label of correspondence just can be got according to mapping table.
It is understood that a color likely corresponds to multiple different types of color image label, even same
Individual tag types is also possibly corresponding to multiple label.The most same color can belong simultaneously to male's color and women color.
And when commodity include two or three color, exist as it has been described above, color parsing module 311 can calculate two colors
Absolute value | θ | of central angle in hue circle.Having following color relationship: when | θ |=0, these two kinds of colors belong to color of the same race；When
0 < during | θ |≤15 °, these two kinds of colors belong to adjacent colour；When 15 °, < during | θ |≤30 °, these two kinds of colors belong to Adjacent color；When 30 °
< during | θ |≤90 °, these two kinds of colors belong to middle dyeing；When 90 °, < during | θ |≤120 °, these two kinds of colors belong to contrast color；When 12 °
< during | θ |≤180 °, these two kinds of colors belong to complementary color.When two kinds of colors are collocated with each other, its central angle in hue circle | θ
| meeting this commodity color of strong influence is to the aesthetic method of people, and can affect the style of commodity.
Hue circle is segmented into multiple interval, it is assumed that the angle in each interval is 1 °, then interval quantity is 360, and
Each interval is divided into again multiple subinterval from the center of circle to external diameter, and each subinterval represents the difference of purity, if purity being divided into
10 subintervals, represent the purity level of 1-10 respectively.So, this hue circle is split as 3600 subintervals.
Based on the description above, angle in hue circle of two kinds of colors and purity can affect color to the impression of people,
And the style of commodity can be affected.Therefore, color parsing module 311 can be given in 3600 subintervals, any two subinterval
Combination and corresponding color image label between corresponding relation.This corresponding relation can be by manually based on empirical rule
Input or machine learning obtain.So, when including two kinds of colors in commodity, according to two kinds of these corresponding relations of color lookup,
The color image label of commodity can be acquired.
Additionally, when two kinds of colors are collocated with each other, except color itself, the area ratio of color is also to affect outward appearance impression
Key factor, therefore, in the corresponding relation of the combination in above-mentioned subinterval and color image label, it is also contemplated that color
The impact of area ratio, say, that identical color combination, but there is different area ratios, can have different
Color image label.For example, same blueness and white, the pattern of white on the blue degree of 90%, and on the white background of 90%
Blue pattern is exactly diverse to the impression of people.
Color parsing module 311 can also be given in 3600 subintervals shown in Fig. 3, the combination in any three subintervals with
The corresponding corresponding relation between color image label.So, when including three kinds of colors in commodity, according to three kinds of color lookup
This corresponding relation, can acquire the color image label of commodity.
When the color included in product is four kinds or above, color parsing module 311 can provide the row of color combination
Table, each project definition in list goes out the scope such as four kinds or above color, area ratio, and corresponding color
Image label.Four kinds are included in commodity or during above color, by the row of the color data of commodity with color combination when detecting
Table goes to compare, when finding the project of coupling, directly using the color image label that is given in list as the color of commodity
Image label.Use this kind of scheme, four kinds of colors can be avoided to combine the huge workload brought, but provide four kinds or with
The acquisition of color image label under upper color combination situations.
Genre labels module 312 obtains aesthetic, the artistic style class corresponding to commodity for the color data according to commodity
Type, such as Baroco style, Rococo style, punk, Broadway style, China's the Republic of China style, the style twenties, super existing
Real doctrine style, POP style etc..Each label also can have age and the region of correspondence.Therefore, genre labels is permissible
Filter according to age and/or region.
Specifically, genre labels module 312 can predefine the representative colors pattern of every kind of style, color mould herein
Formula refers to that the combination of the form and aspect of color, lightness, purity is interval, the area ratio etc. of each color when multiple color occurs jointly.When
When the color data of certain commodity meets the typical color pattern of certain style, genre labels module 312 can determine that these commodity have
There is the genre labels of correspondence.It is understood that same color mode is also possibly corresponding to multiple style, namely have multiple
Different genre labels.Such as, certain color may belong simultaneously to surrealism style and Rococo style.
User's matching module 313 is for realizing mating of commodity and user according to color data.Specifically, user mates mould
Block 313 can realize mating of color data and user based on three levels.
First, based on color itself, such as, user can set the commodity color oneself wanted, then user's matching module
The color of object that commodity color and user set can be compared by 313, just makees commodity further when color is consistent
Process, or directly transmit client 40.For example, when user wishes to choose the overcoat of a claret, then can claret-red
This condition of color can be used directly to the color data with commodity and carry out matching operation, and the commodity only met just appear in
Join in result.
Second, the color image label set based on user.This refers to be supplied to by conventional color image label
User selects, and user can select certain color image label, such as user can select " vivaciously " label, then can be by user
The color image label selected is compared with the color image label of commodity, when the color image label of commodity includes that user selects
Color image label time just commodity are further processed, or directly commodity data is sent to client 40.
3rd, based on the Matching Model between color image label and user.Concrete, user's matching module 313 is first
The type label of user can be obtained.User type label herein is for describing the personal attribute of user, such as age, duty
Industry, orientation of aesthetics etc..And one default between each concrete user type label and all of color image label
Degree of joining (this matching degree can by manual type preset or by machine learning by the way of obtain and be stored in data base
In), matching degree herein can be simple bi-values, i.e. mates and does not mates, it is also possible to being a concrete numerical value, represents
The height of matching degree.The matching degree preset according to this, when the color image label of commodity determines, user's matching module 313
Can be according to the matching degree between color image tag computation and active user.And calculated matching degree may be used for business
The functions such as product data are ranked up, screen and distribute.
Tie-in sale module 314 realizes being collocated with each other between commodity based on color data.A concrete embodiment party
In formula, it is referred to said process and first obtains the color image label of commodity, when the color image label of two commodity is identical
Or when being mutually matched, can arrange in pairs or groups between two commodity, otherwise cannot arrange in pairs or groups.
In a specific embodiment, above-mentioned based on color data realize being collocated with each other between commodity include with
Lower step: set up colour match rule, it is for the collocation rule of definition two kinds or above color, definition every kind in every rule
The color attribute (including form and aspect, lightness and purity) of color is interval, and every rule can have a corresponding color image mark
Sign.Collocation rule can include F-rule or negative sense rule, and F-rule refers to legal belong to outstanding collocation, and bears
Refer to that the non-optimal that belongs to meeting regulation is arranged in pairs or groups to rule.When two or more commodity is collocated with each other, by business to be arranged in pairs or groups
The color of product, compared with each rule in colour match rule, for F-rule, shows this two kinds of colors when can not find
Can arrange in pairs or groups regular time, then these two commodity can not be arranged in pairs or groups.All of grouping of commodities is performed this process, can by all not
The grouping of commodities that can arrange in pairs or groups filters away, the remaining grouping of commodities that can arrange in pairs or groups.Owing to each colour match rule has one
Corresponding color image label, therefore, during the collocation of commodity, the combination of each commodity also has a corresponding color
Image label.This color image label can be used to allow user screen commodity, or does matching operation with the type of user, when with
Just the grouping of commodities that can arrange in pairs or groups is distributed to user or client during the type matching of user.It is appreciated that when carrying out three
When individual or above commodity carry out collocation process, both all combinations of three commodity can be performed said process, it is also possible to be
The collocation first carrying out any two commodity processes, then the grouping of commodities that can arrange in pairs or groups and the 3rd commodity are carried out collocation process.Adopt
By latter approach, operand can be reduced, accelerate the speed that collocation processes.
Refering to Fig. 4, shape subsystem 32 includes: shape analysis module 321, user's matching module 322, tie-in sale module
323。
Shape analysis module 321 is for resolving the shape data of commodity to obtain the shape of commodity, the shape of commodity
Model at least includes the shape (flat shape) of commodity profile (three-dimensional shape) and each parts.Profile refers to outside the entirety of commodity
See type.Represent with alphabetic sort method, can be generally divided into A, H, X, T-shaped.
Wherein A type: jacket and overcoat are with not closing waist, the wide bottom, or closing waist, the wide bottom are basic feature.Jacket is typically takeed on
Portion is narrower or naked shoulder, and clothing pendulum is loose loose；Skirt and trousers are all characterized with the wealthy pendulum of tight waist.
H type: jacket and overcoat are with not closing waist, the narrow bottom as basic feature.Clothing body is straight-tube shape；Skirt and trousers also more than
Under wide straight-tube shape be characterized.
X-type: jacket and overcoat are with loose shoulder, wealthy pendulum, closing waist as basic feature；Skirt and trousers are also with the loosest, middle
Thin tightly it is characterized.
T-shaped: jacket, overcoat, one-piece dress etc. are with exaggeration shoulder, the contraction bottom as principal character.
The profile of commodity can obtain, directly at commodity data by the overall photo of commodity is carried out image recognition analysis
By manually directly inputting during input system, or directly provided by commodity data provider.
The shape of parts can be divided into two levels: big class and group.As shown in the table:
It is appreciated that the Shape Classification in above-mentioned form is only signal, is not limited to the present invention, other classification
Mode equally can apply in the embodiment of the present invention.
The shape of commodity component can be obtained by lower step: first, is identified the picture of commodity, to obtain each portion
The style of part, as a example by collar, its possible style includes: crew neck, little crew neck, offneck and V neck.On the other hand, shape
Parsing module 321 pre-builds the mapping table between the style of parts and component shape, as shown in the table:
It is understood that picture herein is not limited to the overall picture of commodity, it is also possible to be the component home of commodity
Picture, such as, collar detail view.
So, when being identified the style of certain parts by image recognition technology, can obtain according to this mapping table
Obtain the shape of these parts.All parts of commodity are performed above-mentioned flow process, the shape of commodity can be obtained.At a tool
In the embodiment of body, the element portion of the shape of acquisition is as follows:
| Parts | Style | Shape |
| Collar | Crew neck | Circular |
| Shoulder type | Slip shoulder | Circular |
| Sleeve type | Bubble fold is tucked inside the sleeve | Circular |
| Forward swing | Circular arc | Circular |
| Rear pendulum | Circular arc | Circular |
| Version type | Sheet is cut in shirtfront | Square |
Further, it is to be appreciated that first, above-mentioned example illustrates as a example by jacket, but similar flow process is permissible
Use to any commodity；Second, the shape of commodity component is also not necessarily limited to be obtained by aforesaid way, such as, and the shape of commodity
Shape model directly can be provided need not be resolved again by commodity data presentation mode.
Further, shape analysis module 321 can also carry out quantity statistics, such as showing at upper table to the shape of parts
In example, the circular number of times occurred is 5, and the number of times of square appearance is 1, so, if will appear from the most shape of number of times as this
The global shape of commodity, then the global shape of these commodity is also circular (i.e. the global shape stylistic category of commodity).These quantity
The result of statistics also constitutes a part for shape.
User's matching module 322 realizes mating between commodity with user based on above-mentioned shape.Specifically, user
Matching module 322 processes following two relation: the matching relationship (A) between clothing silhouette and body shape feature, garment dimension
And the matching relationship (B) between body shape feature.
For relation A, user's matching module 322 pre-builds the matching relationship between clothing silhouette and user's physical characteristic
Mapping table A, user's physical characteristic herein uses body model to express, and stature model can include, but are not limited to following number
According to: stature type, neck type, face shape, height, body weight, brachium, lower limb length etc..
The stature type of user such as comprises the steps that hourglass shape, del, equilateral triangle, rectangle etc..Concrete at one
In embodiment, corresponding to every kind of stature type, the figure and features data of user are divided into 12 subintervals.It is right that each subinterval has
The height answered and weight range.Each subinterval corresponds to every kind of profile, has a corresponding matching degree.Such as, matching degree
Can be: 3 evaluation ranks such as " can arrange in pairs or groups " (numeral 1), " typically " (numeral 2), " can not arrange in pairs or groups " (numeral 3).
In a specific embodiment, the part-structure of mapping table A is as follows:
According to mapping table A, when the physical characteristic data of user determine, and the profile of clothing (commodity) determines, it is possible to
Obtain the matching degree between commodity and active user.And other physical characteristic can process in the way of employing is similar.
Further, the stature model of user is also not necessarily limited to use above-mentioned data to express, it is also possible to use more smart
Accurate threedimensional model, now, mapping table A can store the matching degree of clothing silhouette and the characteristic range of user's stature model.Warp
Cross similar flow process, equally determine the matching degree between commodity and user.
For relation B, user's matching module 322 pre-builds the matching relationship between garment dimension and user's physical characteristic
Mapping table B, mapping table B correspond to every kind of multiple subinterval of stature type definition according to the physical characteristic data of human body, and each
Subinterval has different height and weight range, corresponding to storing itself and every kind of garment dimension in each subinterval mapping table B
Matching degree.
Garment dimension can use international size S, M, L, XL, XXL of standard etc. to represent, but except international chi
Outside Ma, it is also possible to use certain specific dimensions (such as shoulder breadth, chest measurement, waist loose, sleeve length, hip circumference etc.) on clothing to represent, so can
Make the matching degree between size and buman body type feature more accurate.
In a specific embodiment, the part-structure of mapping table B is as follows:
According to mapping table B, when the physical characteristic data of user and the size of clothing or other sizes determine, so that it may
To obtain the matching degree between clothing and active user accurately.
It addition, matching degree based on user's stature with clothing silhouette acquisition user with commodity is not limited to above this side
Formula, such as, except considering entirety from the angle of profile, it is also possible to is simultaneously based on profile and obtains overall mating with stature type
Degree and each parts based on clothing do with the local stature feature of user and mate.
Tie-in sale module 323 calculates coupling when being collocated with each other between the two for shapes based on two commodity
Degree.
In a specific embodiment, tie-in sale module 323 profile calculation based on two commodity is between the two
Matching degree when arranging in pairs or groups mutually.Specifically, having a default matching degree between every two kinds of profiles, matching degree can be: " permissible
Collocation " 3 evaluation ranks such as (numeral 1), " typically " (numeral 2), " can not arrange in pairs or groups " (numeral 3).Therefore, when the exterior feature of two commodity
After shape determines, the matching degree preset according to this, matching degree during two tie-in sales can be got.
In a specific embodiment, tie-in sale module 323 calculates both phase laps based on two commodity component
The matching degree of timing.Specifically, tie-in sale module 323 pre-defines the matching degree between two different parts.Therefore, when
When the parts of commodity determine, its matching degree each other also determines that.
In a specific embodiment, tie-in sale module 323 distribution of shapes based on two commodity component calculates
Matching degree when both are collocated with each other.Specifically, including the shape of each parts in the shape of commodity, first, commodity are taken
Join module 323 shape based on each parts to carry out quantity statistics and represent shape with the shape obtaining commodity.And the not similar shape of commodity
Matching degree between shape can pre-define, so according to the global shape of commodity component, it is possible to obtains based on global shape
Matching degree.
It is understood that above various modes can be used alone, it is also possible to be used in any combination.And in basis
Different tie-in sale scenes, uses one or more modes therein can obtain optimal matching effect.Such as, for clothes
For dress collocation, the collocation employing profiles based on commodity of jacket and lower clothing carry out mating and can have preferably arranging effect, interior
Dress carries out mating with exterior collocation employing shapes based on parts or parts and can have preferably arranging effect, and two kinds not
The collocation of generic commodity is used in shape based on parts and carries out coupling and can have preferably arranging effect.These are specific
The rule using which kind of collocation mode under scene can pre-define, and performs according to predefined rule when actually used
Can.It is to say, first tie-in sale module 323 determines collocation type, then obtain mate most with type of arranging in pairs or groups based on shape
The tie-in sale scheme of shape, last according to matching degree when arranging in pairs or groups between the program two commodity of calculating and real according to matching degree
Now to the filtration of commodity, screen, the operation such as sequence.
According to above-mentioned shape subsystem 32, shapes based on commodity mating between commodity with user can be realized, with
And arrange in pairs or groups between two commodity.
Refering to Fig. 5, style subsystem 33 includes: style parsing module 331, user's matching module 332 and tie-in sale
Module 333.
Style parsing module 331 obtains the Style Model of commodity according to commodity data for resolving.Style Model includes
The data of following dimension: material, pattern, color, shape, profile, technique, style, parts etc..It is appreciated that these dimensions
Data are in addition to material with pattern, and other data are the most retouched in shape subsystem 32 at aforesaid color subsystem 31
State.And material and pattern data can be to be obtained by image recognition technology, by being manually entered or directly by data providing
There is provided.
On the other hand, style parsing module 331 can provide the feature of every kind of style, and the feature of style can include following dimension
The data of degree: material, pattern, color, profile, technique, style, parts etc..Then by the Style Model of commodity and these features
Match, if the match is successful, then give these commodity one corresponding genre labels.
User's matching module 332 realizes mating of user and commodity for Style Model based on commodity.Specifically, for
Given user model and Style Model, user's matching module 332 can provide default matching degree, thus obtain user and business
Matching degree based on genre labels between product.And user model such as can include preference and the figure and features feature etc. of user.
In a specific embodiment, corresponding to every kind of material, user's matching module 332 provides what material was mated
User type, if active user's type is identical with the user type corresponding to this material, is then considered as active user and this commodity phase
Coupling mutually.For example, if the material of commodity is cotton material, the type of its corresponding people is natural type.If active user
For natural type, then with this goods matching.The type of people is based on user data such as preference data, occupation data, dress herein
The data such as occasion, figure and features data obtain.And specifically which kind of preference, occupation, occasion are based on empirical rule corresponding to natural type
Or machine learning obtains, and unrestricted.
In a specific embodiment, user's matching module 332 can obtain the stylistic category that user wishes to obtain,
Or the scene that commodity use；User is wished to the stylistic category obtained, can directly genre labels with commodity compare
To judge whether coupling, and user only being set to the situation using scene, user's matching module 332 can provide genre labels
With the matching relationship table of use scene, so, may determine that whether commodity mate with user according to this matching relationship table.And this
Other data that individual matching relationship table can extend with material, such as cloth, technique etc. are set up.
In a specific embodiment, corresponding to every kind of pattern, user's matching module 332 provides what pattern was mated
User type, if active user's type is identical with the user type corresponding to this pattern, is then considered as active user and this commodity phase
Coupling mutually.For example, if the pattern of commodity is the common pinaster pattern of 19th century Switzerland, the type of its corresponding people be from
So type.If active user is natural type, then with this goods matching.
Be appreciated that similar to material, pattern can also based on user preference and the scene being desirable for realize with
The coupling of user.
Tie-in sale module 333 realizes being collocated with each other between commodity for Style Model based on commodity.Specifically,
Can realize being collocated with each other based on material and/or pattern.Whether tie-in sale module 333 can be given, between every kind of pattern
Can arrange in pairs or groups.Therefore, after the material on two commodity and/or pattern determine, whether can take according to what this was preset
The relation joined, can realize the collocation between two commodity.
According to above-mentioned style subsystem 33, styles based on commodity mating between commodity with user can be realized, with
And arrange in pairs or groups between two commodity.
Refering to Fig. 5, it by mating the signal of the model of the commodity that commending system is set up according to above-mentioned personalized commercial
Figure.Goods model includes three submodels: color model, shape and Style Model.And each submodel includes time
Level submodel.Such as color model includes monochrome, double-colored, the trichroism and secondary submodel of polychrome, and shape includes profile and parts
Secondary submodel, style includes material and pattern secondary submodel.Current submodel together constitutes the complete picture of commodity
Commodity are expressed by picture from each dimension.It is understood that above-mentioned submodel is merely illustrative with secondary submodel,
It is not limited to the present invention.Anyon model, the combination of secondary submodel in goods model may serve to realize commodity
And the collocation between commodity, also can realize mating between particular commodity or goods of joint with user.And concrete commodity it
Between collocation logic be described in above-mentioned each subsystem.
Refering to Fig. 6, the schematic diagram of its user model used by above-mentioned personalized commercial coupling commending system.User
Model includes three submodels: bodily form characteristic model, user preference set and social mentality's model.Wherein physical characteristic model
The various physical characteristic data such as the definition height of user, body weight, shoulder enclose, waistline, lower limb height, hip circumference, chest measurement, face, brachium.Preference
Set the preference permanently or temporarily preserving user, such as color, style, dressing occasion etc..And social mentality's model is used
The personality at family, makings, hobby, occupation, sex, social mentality's analysis result etc..Carrying out user and when mating of commodity, commodity
The secondary submodel of model can carry out matching operation with one or more submodels of user model.And the basis of computing is big
Amount is preset or the rule of storage after machine learning.
Embodiment 1
Refering to Fig. 7, the flow chart of its commodity data coupling recommendation method provided for embodiment 1.As it is shown in fig. 7, the party
Method comprises the following steps:
Step S101, obtains the user model of active user.As it has been described above, include the number of multiple dimension in user model
According to, this needs user to provide.Specifically, it is provided that interface or interface allow user can input the various data of their own,
Such as physical characteristic data, preference data and social mentality's data.If the data required for user model are the most complete, and store
In data storage device 20, then directly can read the user model of active user from storage device 20.
Step S102, does commodity to be recommended and user model matching degree computing, and filters to be recommended according to matching degree
Commodity.The above-mentioned the most all commodity sold of commodity to be recommended, by each commodity according to the description above
Carry out matching operation between mode with user model to obtain whether these commodity mate with active user, and filter out all not
The commodity joined.
Step S103, the items list after filtering is sent to client to be shown.According to the method for the present embodiment,
User is after logging in certain shopping website or application program, as long as user is logged in, and needed for having carried out user model
Want the collection of data, then all commodity that this shopping website or application program are shown are all the commodity with current matching,
This undoubtedly can be the most user-friendly, improves shopping at network efficiency, decreases user and screens the time of commodity.
Embodiment 2
Refering to Fig. 8, the flow chart of its commodity data coupling recommendation method provided for embodiment 2.As shown in Figure 8, the party
Method comprises the following steps:
Step S201, obtains the commodity that user selects.Such as, at a shopping website or application program, it is provided that
User selects the interface of commodity, after user selects some commodity, can send a request to application server 30, will use
The commodity that family selects are included in this request.Correspondingly, application server 30 can receive this request, and obtains what user selected
Commodity.
Step S202, obtains the Recommendations of the goods matching selected with described user.
Step S203, does matching operation and according to matching degree filtered recommendation commodity by described Recommendations and user model.
From data base, obtain all alternative commodity, enter between each alternative commodity mode and user model according to the description above
Row matching operation is to obtain whether these commodity mate with active user, and filters out all unmatched commodity.
Step S204, the commercial product recommending list after filtering is sent to client to be shown.
According to the method for the present embodiment, in shopping website or application program, after user selects a basic item,
The coupling recommendation method that can use the embodiment of the present invention obtains the commodity that the commodity selected with user are mutually matched, and selects for user
Select, reduce user and arrange in pairs or groups time of commodity, improve the efficiency of shopping.
Embodiment 3
Refering to Fig. 9, embodiment 3 provides the matching process of a kind of user based on color and commodity, as it is shown in figure 9, the party
Method includes:
Step S301, obtains the color data of commodity, and described color data includes the form and aspect of commodity color, lightness and pure
Degree；
Step S302, initializes color computational model according to described color data.It is appreciated that color calculation mould herein
Type be above-mentioned from single color, double-colored, trichroism and polychrome to the mapping relations model of color image label, can mutually join
According to.
Step S303, described color computational model is calculated the color image mark of described commodity according to described color data
Sign.
Step S304, obtains the type label of active user, according to predefined color image label and type label it
Between matched data calculate the matching degree between active user and described commodity；
Step S305, filters described commodity data according to described matching degree.
According to the method for the present embodiment, by the value of color being converted to corresponding color image label, can locate efficiently
Reason color is mated with user's.
Embodiment 4
Refering to Figure 10, embodiment 4 provides the matching system of a kind of user based on color and commodity, as shown in Figure 10, is somebody's turn to do
System includes: color parsing module 41, user's matching module 42 and data match module 43.
Color parsing module 41 is for obtaining the color data of commodity, and described color data includes the color included by commodity
Form and aspect, lightness and purity, according to described color data initialize color computational model to calculate according to described color data
Obtain the color image label of described commodity；
User's matching module 42 is for obtaining the type label of active user, according to predefined color image label and class
Matched data between type label calculates the matching degree between active user and described commodity；
Data match module 43 is for filtering described commodity data according to described matching degree.
According to the system of the present embodiment, by the value of color being converted to corresponding color image label, can locate efficiently
Reason color is mated with user's.
Embodiment 5
Refering to Figure 11, embodiment 5 provides the matching method of a kind of commodity based on color, as shown in figure 11, the method bag
Include:
Step S501, obtains the color data of commodity, and described color data includes the form and aspect of the color included by commodity, bright
Degree and purity；
Step S502, initializes color computational model according to described color data；
Step S503, described color computational model is calculated the color image mark of described commodity according to described color data
Sign；
Step S504, it is judged that the color image label of described two kinds of commodity to be matched is the most identical or mates, if so,
Then return the result that described two kinds of commodity to be matched can be arranged in pairs or groups.
According to the method for the present embodiment, by the value of color being converted to corresponding color image label, can locate efficiently
Reason commodity and the collocation computing of commodity.
Embodiment 6
Refering to Figure 12, embodiment 6 provides a kind of tie-in sale system based on color, and as shown in figure 12, this system includes:
Color parsing module 61 and user's matching module 62.
Color parsing module 61 is for obtaining the color data of to be matched two kind commodity, and described color data includes commodity
The included form and aspect of color, lightness and purity, initializes color computational model with according to described according to described color data
Color data is calculated the color image label of every kind of commodity；
Tie-in sale module 62 the most identical for the color image label judging described two kinds of commodity to be matched or
Coupling, the most then return the result that described two kinds of commodity to be matched can be arranged in pairs or groups.
According to the method for the present embodiment, by the value of color being converted to corresponding color image label, can locate efficiently
Reason commodity and the collocation computing of commodity.
Embodiment 7
Refering to Figure 13, embodiment 7 provides district's method of completing the square of a kind of commodity based on color, as shown in figure 13, the method bag
Include:
Step S701, sets up matching relationship mapping table between commodity shape type and user's stature type, described mapping table
Figure and features data according to user correspond to every kind of multiple subinterval of stature type definition, and each subinterval has and at least partly do not weighs
Folded height and body weight are interval, store mating between its with each shape type corresponding to mapping table described in each subinterval
Degree；Step S702, resolves the image of commodity to obtain the shape type of commodity；Step S703, obtains the stature class of active user
Type and figure and features data；Step S704, described to obtain according to mapping table described in described stature type and figure and features data retrieval
Matching degree between commodity and described active user, and according to described matching degree, commodity are filtered.
According to the method for the present embodiment, by commodity shape type and stature type are realized between user and commodity
Join computing, there is the highest matching efficiency and effect.
Embodiment 8
Refering to Figure 14, embodiment 8 provides the matching system of a kind of user based on shape and commodity, as shown in figure 14, is somebody's turn to do
System includes: mapping table sets up module 81, shape analysis module 82, data acquisition module 83 and user's matching module 84.
Mapping table sets up module 81 for setting up matching relationship mapping table between commodity shape type and user's stature type,
Described mapping table according to the figure and features data of user corresponding to every kind of multiple subinterval of stature type definition, each subinterval have to
The nonoverlapping height of small part and body weight are interval, corresponding to mapping table described in each subinterval store itself and each shape type it
Between matching degree；
Shape analysis module 82 is for resolving the image of commodity to obtain the shape type of commodity；Data acquisition module 83 is used
In the stature type and the figure and features data that obtain active user；User's matching module 84 is for according to described stature type and body
Mapping table described in looks data retrieval is to obtain the matching degree between described commodity and described active user, according to described matching degree pair
Commodity filter.
According to the system of the present embodiment, by commodity shape type and stature type are realized between user and commodity
Join computing, there is the highest matching efficiency and effect.
Embodiment 9
Refering to Figure 15, embodiment 9 provides a kind of tie-in sale method based on shape, and as shown in figure 15, the method includes:
Step S901, sets up the mapping relations between collocation type and collocation model；Step S902, according to two kinds of commodity to be arranged in pairs or groups
Classification determines collocation type；Step S903, obtains the collocation model corresponding with described collocation type according to described mapping relations；Step
Rapid S904, calculates the matching degree between to be arranged in pairs or groups two kind commodity based on described collocation model and returns result of calculation.
According to the method for the present embodiment, shapes based on commodity realize the collocation between commodity, and treatment effeciency is high, and collocation
Effective.
Embodiment 10
Refering to Figure 16, embodiment 10 provides a kind of tie-in sale system based on shape, as shown in figure 16, this system bag
Include: mapping table set up module 1001, collocation determination type module 1002, collocation model determine module 1003, tie-in sale module
1004。
Mapping table sets up module 1001 for setting up the mapping relations between collocation type and collocation model；Collocation type is true
Cover half block 1002 is for determining collocation type according to the classification of two kinds of commodity to be arranged in pairs or groups；Collocation model determine module 1003 for
The collocation model corresponding with described collocation type is obtained according to described mapping relations；Tie-in sale module 1004 is for based on described
Collocation model calculates the matching degree between to be arranged in pairs or groups two kind commodity and returns result of calculation.
According to the system of the present embodiment, shapes based on commodity realize the collocation between commodity, and treatment effeciency is high, and collocation
Effective.
The foregoing is only embodiments of the invention, not thereby limit the scope of the claims of the present invention, every utilize this
Equivalent structure or equivalence flow process that bright description and accompanying drawing content are made convert, or are directly or indirectly used in other relevant skills
Art field, is the most in like manner included in the scope of patent protection of the present invention.
Claims (10)
1. a personalized commercial coupling commending system, it is characterised in that including:
Data acquisition unit, is used for gathering commodity data and user data, and described commodity data includes color data, shape number
According to, sized data and image, described user data includes the bodily form/figure and features data, preference data and social mentality's data；
Data storage device, is used for storing described commodity data and user data；
Goods matching device, including:
Color subsystem, for setting up color model according to described commodity data, described color model is according to the face of described commodity
The numerical range that chromatic number is regular according to determining color, and export the color image label of correspondence accordingly, and based on described color image
Label and described user data set up search engine, to obtain the matching degree between described commodity and user；
Shape subsystem, for setting up shape according to described commodity data, described shape is according to outside described commodity
The graphic data output entirety of commodity and/or the shape type of parts, and based on described entirety and/or the shape type of parts and institute
State user data and calculate the matching degree between described commodity and user；
Style subsystem, for setting up Style Model according to described commodity data, described Style Model is according to described commodity data
The genre labels that output is corresponding, and according to described genre labels and described user data calculate between described commodity and user
Degree of joining；And
Coupling recommending module, for the coupling of the output of comprehensive described color subsystem, shape subsystem and style subsystem
Degrees of data is to obtain the final matching results between described commodity and user.
2. personalized commercial coupling commending system as claimed in claim 1, it is characterised in that described color subsystem includes:
Color parsing module, obtains the form and aspect of corresponding color, lightness and purity for the color data parsing obtaining commodity,
Color computational model is initialized to be calculated the color image label of described commodity according to analysis result；
User's matching module, for obtaining the type label of active user, according to predefined color according to described user data
Matched data between image label and type label calculates the matching degree between active user and described commodity.
3. personalized commercial coupling commending system as claimed in claim 2, it is characterised in that described color parsing module is by bright
Spend and be respectively defined as multiple continuous print lightness interval and purity interval with purity, obtain the color data place lightness of described commodity
Interval and purity are interval, and according to reflecting between combination and color image label between predefined lightness interval and purity district
Penetrate the color image label of commodity described in Relation acquisition.
4. personalized commercial coupling commending system as claimed in claim 2, it is characterised in that described color parsing module is also used
In hue circle is divided into the most adjacent multiple subinterval, and each subinterval is divided into multiple secondary subinterval group；Build
The color combination in vertical any two or three secondary subintervals and the mapping table of color image label；When described number of colours being detected
According to when including two or three color, from described mapping table, search the color image label of correspondence.
5. personalized commercial coupling commending system as claimed in claim 2, it is characterised in that described color parsing module is also used
In: set up color Assembly Listing, the corresponding combination including at least four color of each project in this list and correspondence
Color image label；When detecting that described color data includes four kinds or during above color, by the number of colours of described commodity
According to each comparison of item in described color Assembly Listing, when some in color data and the described list of described commodity
During project coupling, set the color image label corresponding with this project to described commodity.
6. personalized commercial coupling commending system as claimed in claim 1, it is characterised in that described shape subsystem includes:
Mapping table sets up module, is used for setting up matching relationship mapping table between commodity shape and user's figure and features model, described
Matching degree between mapping table figure and features data storage commodity shape and figure and features model；
Shape analysis module, for resolving the image of commodity to obtain the shape type of commodity；
User's matching module, for obtaining the stature type of active user, according to described stature type according to described user data
And mapping table described in figure and features data retrieval is to obtain the matching degree between described commodity and described active user.
7. personalized commercial coupling commending system as claimed in claim 6, it is characterised in that described shape analysis module resolves
The image of described commodity, to obtain the shape of each parts of described commodity, adds up the shape of all parts to obtain described commodity
Representing the shape shape type as described commodity, described shape type includes two-dimensional shapes type and/or 3D shape type.
8. personalized commercial coupling commending system as claimed in claim 6, it is characterised in that described mapping table also stores commodity
Matching degree between size and each subinterval of particular elements；
Described user's matching module is additionally operable to: obtain the size of this particular elements of described commodity, according to the chi of this particular elements
The described mapping table of very little lookup obtains the matching degree between size and active user.
9. personalized commercial coupling commending system as claimed in claim 1, it is characterised in that described style subsystem: set up
Matching relationship between genre labels and user preferences and figure and features feature；Obtain commodity according to described matching relationship to use with current
Matching degree between family.
10. a personalized commercial coupling recommendation method, it is characterised in that including:
Gathering and store commodity data and user data, described commodity data includes color data, shape data, sized data
And image, described user data includes figure and features, preference data and social mentality's data；
Set up color model according to described commodity data, according to the color data of described commodity, described color model determines that color is advised
Numerical range then, and export the color image label of correspondence accordingly, and based on described color image label and described number of users
According to setting up search engine to obtain the matching degree between described commodity and user；
Setting up shape according to described commodity data, described shape exports commodity according to the shape data of these commodity described
Entirety and/or the shape type of parts, and shape type based on described entirety and/or parts and described user data calculate
Matching degree between described commodity and user；
Setting up Style Model according to described commodity data, described Style Model is according to the style mark of described commodity data output correspondence
Sign, and calculate the matching degree between described commodity and user according to described genre labels and described user data；And
Comprehensive matching degree data based on color, shape and style finally mate knot to obtain between described commodity and user
Really.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610519895.9A CN106204124A (en) | 2016-07-02 | 2016-07-02 | Personalized commercial coupling commending system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610519895.9A CN106204124A (en) | 2016-07-02 | 2016-07-02 | Personalized commercial coupling commending system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106204124A true CN106204124A (en) | 2016-12-07 |
Family
ID=57465488
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610519895.9A Pending CN106204124A (en) | 2016-07-02 | 2016-07-02 | Personalized commercial coupling commending system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106204124A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106354768A (en) * | 2016-08-18 | 2017-01-25 | 向莉妮 | Matching method for users and commodities and commodity matching recommendation method based on color |
| CN107194312A (en) * | 2017-04-12 | 2017-09-22 | 广东银禧科技股份有限公司 | A kind of model recommendation method based on 3D printing model content |
| CN107256513A (en) * | 2017-06-19 | 2017-10-17 | 数据堂（北京）科技股份有限公司 | Method and device is recommended in a kind of collocation of object |
| CN107424012A (en) * | 2017-07-31 | 2017-12-01 | 京东方科技集团股份有限公司 | A kind of intelligent shopping guide method, intelligent shopping guide equipment |
| CN107679960A (en) * | 2017-10-12 | 2018-02-09 | 东北大学 | A kind of recommendation method of the customized clothing based on image of clothing and label text bimodal content analysis |
| CN107908685A (en) * | 2017-10-31 | 2018-04-13 | 西安交通大学 | The retrieval of various visual angles commodity image and recognition methods based on transfer learning |
| CN108022155A (en) * | 2017-12-07 | 2018-05-11 | 北京小米移动软件有限公司 | Jewel accessory recommends method and device |
| CN113052662A (en) * | 2021-04-20 | 2021-06-29 | 中国工商银行股份有限公司 | Data processing method, apparatus, electronic device, medium, and program product |
| CN110021061B (en) * | 2018-01-08 | 2021-10-29 | Oppo广东移动通信有限公司 | Collocation model construction method, clothing recommendation method, device, medium and terminal |
| CN114969862A (en) * | 2022-06-27 | 2022-08-30 | 广东时谛智能科技有限公司 | Method, device, equipment and storage medium for generating shoe body model according to user information |
| CN116911959A (en) * | 2023-09-12 | 2023-10-20 | 南通尚轩金属制品有限公司 | Data processing method for building material non-standard part |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103544632A (en) * | 2013-07-22 | 2014-01-29 | 杭州师范大学 | Method and system for individually recommending network commodities |
| CN103810626A (en) * | 2012-11-13 | 2014-05-21 | 姚德明 | Method for online shopping and online shopping intermediary |
| CN105654330A (en) * | 2015-06-22 | 2016-06-08 | 向莉妮 | Individual commodity matching recommendation system and method thereof |
-
2016
- 2016-07-02 CN CN201610519895.9A patent/CN106204124A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103810626A (en) * | 2012-11-13 | 2014-05-21 | 姚德明 | Method for online shopping and online shopping intermediary |
| CN103544632A (en) * | 2013-07-22 | 2014-01-29 | 杭州师范大学 | Method and system for individually recommending network commodities |
| CN105654330A (en) * | 2015-06-22 | 2016-06-08 | 向莉妮 | Individual commodity matching recommendation system and method thereof |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106354768A (en) * | 2016-08-18 | 2017-01-25 | 向莉妮 | Matching method for users and commodities and commodity matching recommendation method based on color |
| CN106354768B (en) * | 2016-08-18 | 2020-03-17 | 向莉妮 | Color-based user and commodity matching method and commodity matching recommendation method |
| CN107194312A (en) * | 2017-04-12 | 2017-09-22 | 广东银禧科技股份有限公司 | A kind of model recommendation method based on 3D printing model content |
| CN107256513A (en) * | 2017-06-19 | 2017-10-17 | 数据堂（北京）科技股份有限公司 | Method and device is recommended in a kind of collocation of object |
| CN107424012A (en) * | 2017-07-31 | 2017-12-01 | 京东方科技集团股份有限公司 | A kind of intelligent shopping guide method, intelligent shopping guide equipment |
| CN107679960B (en) * | 2017-10-12 | 2020-06-09 | 东北大学 | Personalized clothing recommendation method based on clothing image and label text bimodal content analysis |
| CN107679960A (en) * | 2017-10-12 | 2018-02-09 | 东北大学 | A kind of recommendation method of the customized clothing based on image of clothing and label text bimodal content analysis |
| CN107908685A (en) * | 2017-10-31 | 2018-04-13 | 西安交通大学 | The retrieval of various visual angles commodity image and recognition methods based on transfer learning |
| CN108022155A (en) * | 2017-12-07 | 2018-05-11 | 北京小米移动软件有限公司 | Jewel accessory recommends method and device |
| CN110021061B (en) * | 2018-01-08 | 2021-10-29 | Oppo广东移动通信有限公司 | Collocation model construction method, clothing recommendation method, device, medium and terminal |
| CN113052662A (en) * | 2021-04-20 | 2021-06-29 | 中国工商银行股份有限公司 | Data processing method, apparatus, electronic device, medium, and program product |
| CN113052662B (en) * | 2021-04-20 | 2024-02-06 | 中国工商银行股份有限公司 | Data processing method, device, electronic equipment and medium |
| CN114969862A (en) * | 2022-06-27 | 2022-08-30 | 广东时谛智能科技有限公司 | Method, device, equipment and storage medium for generating shoe body model according to user information |
| CN116911959A (en) * | 2023-09-12 | 2023-10-20 | 南通尚轩金属制品有限公司 | Data processing method for building material non-standard part |
| CN116911959B (en) * | 2023-09-12 | 2023-12-19 | 南通尚轩金属制品有限公司 | Data processing method for building material non-standard part |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106204124A (en) | Personalized commercial coupling commending system and method | |
| CN106354768A (en) | Matching method for users and commodities and commodity matching recommendation method based on color | |
| JP7136156B2 (en) | Search support system, search support method and program | |
| US10242396B2 (en) | Automatic color palette based recommendations for affiliated colors | |
| CN108256975A (en) | Wearing for 3-D effect is provided for virtual fitting person take system and method based on artificial intelligence | |
| JP6199685B2 (en) | Fashion coordination support device and fashion coordination support system | |
| US20140089781A1 (en) | Color-based designs | |
| CN106462979A (en) | Fashion preference analysis | |
| CN103440587A (en) | Personal image designing and product recommendation method based on online shopping | |
| CN105527946A (en) | Rapid garment system and method based on industrial Internet | |
| CN105426462A (en) | Image searching method and device based on image element | |
| AU2008237335A1 (en) | Computer system for rule-based clothing matching and filtering considering fit rules and fashion rules | |
| CN111767817B (en) | Dress collocation method and device, electronic equipment and storage medium | |
| CN109360050A (en) | Personal care garment management and personalized collocation recommendation intelligence system based on perceptual demand | |
| CN105761127A (en) | Clothing style intelligent matching method based on sign information and system thereof | |
| KR20200042203A (en) | Outfit coordination system and method based on user input Images | |
| CN108920828B (en) | Clothing matching method and system | |
| CN204695311U (en) | Individual commodity recommendation system | |
| KR20200091593A (en) | The user fitting type automatic online fashion coordination matching method | |
| JP6178487B1 (en) | Coordinate system, coordinate device, and program | |
| US20150039994A1 (en) | Color-based designs | |
| CN105678575B (en) | Personalized recommendation method and system based on user attribute knowledge base | |
| CN104317866B (en) | A kind of ready-made clothes matching process and its electronic device | |
| CN105512341A (en) | Personalized recommendation method and system based on big data searching | |
| US20100151430A1 (en) | Identifying a body shape |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20161207 |
|
| RJ01 | Rejection of invention patent application after publication |