CN110060121A - Method of Commodity Recommendation, device and storage medium based on feature ordering - Google Patents
Method of Commodity Recommendation, device and storage medium based on feature ordering Download PDFInfo
- Publication number
- CN110060121A CN110060121A CN201910191547.7A CN201910191547A CN110060121A CN 110060121 A CN110060121 A CN 110060121A CN 201910191547 A CN201910191547 A CN 201910191547A CN 110060121 A CN110060121 A CN 110060121A
- Authority
- CN
- China
- Prior art keywords
- commodity
- list
- similar
- commercial product
- product recommending
- 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
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0603—Catalogue ordering
-
- 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)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of Method of Commodity Recommendation based on feature ordering, device and storage mediums.After obtaining benchmark image Hash codes corresponding to benchmark commodity, image hash code corresponding to other commodity is obtained from character code data library, calculating benchmark image hash code and image wish the Hamming distance between code, similar commodity group is formed according to the corresponding similar commodity of the acquisition from small to large of Hamming distance, and similar commodity are successively added in commercial product recommending list in order, commodity scoring list and similar commodity are equidistantly distributed in commercial product recommending list simultaneously, realize the commercial product recommending that commodity scoring combines exterior of commodity, greatly improve user experience, be conducive to improve the buying rate of user.
Description
Technical field
The present invention relates to data processing field, especially a kind of Method of Commodity Recommendation based on feature ordering, device and deposit
Storage media.
Background technique
Currently, shopping platform is done shopping in more and more consumers selection line with the development of e-commerce.In order to
The buying rate of consumer is improved, each electric business platform usually shows the interested Recommendations of consumer in some pages, therefore
The method of commercial product recommending more can be more effective to promoting to do shopping close to the interest of consumer.In currently existing scheme, mainly according to collaboration
Filtering recommendation algorithms carry out commercial product recommending, although this method can generate according to the similitude and prediction scoring of commodity and recommend quotient
A point list is judged, but collaborative filtering relies primarily on the verbal description to item property, for the commodity based on appearance,
The item property of verbal description can not intuitively embody the interest of consumer, therefore for the commodity based on appearance,
Collaborative Filtering Recommendation Algorithm is difficult to carry out valuable commercial product recommending.
Summary of the invention
For overcome the deficiencies in the prior art, the purpose of the present invention is to provide a kind of commercial product recommendings based on feature ordering
Method, apparatus and storage medium combine exterior of commodity to generate commercial product recommending list in actual use, improve recommended commodity
Reference value.
Technical solution used by the present invention solves the problems, such as it is: in a first aspect, the present invention provides one kind to be based on feature
The Method of Commodity Recommendation of sequence, comprising the following steps:
Client reads benchmark commodity, and the benchmark image Hash of the benchmark commodity is obtained from character code data library
Code;
The client obtains the benchmark image Hash codes from the character code data library and all images are breathed out
Hamming distance between uncommon code successively obtains similar quotient corresponding to image hash code according to the Hamming distance from small to large
Product generate similar commodity group;
The similar commodity group is added in commercial product recommending list by the client.
Further, described image Hash codes are as obtained by the extracted characteristics of image of iterative quantization convolutional neural networks, institute
It states and is provided with spatial pyramid pond layer between the convolutional layer of convolutional neural networks and full articulamentum.
Further, the similar commodity in the similar commodity group are according to its image hash code and the benchmark image Hash codes
Between Hamming distance sort from small to large.
Further, further includes: the client obtains user and has purchased inventory, be included within user purchased it is similar in inventory
Commodity are deleted from similar commodity group.
Further, it is described the similar commodity group is added in commercial product recommending list specifically includes the following steps:
The client obtains commodity and scores list, and by the commodity score the commodity in list according to scoring from height to
Low to be successively added in the commercial product recommending list, the commodity scoring list is the column obtained according to Collaborative Filtering Recommendation Algorithm
Table;
Whether the similar commodity that the client successively detects in the similar commodity group are included in the commercial product recommending column
In table, if including, next similar commodity are detected, the similar commodity are otherwise added to commercial product recommending list
In.
Further, the similar commodity are added in the commercial product recommending list equal intervals.
Second aspect, the present invention provides the devices for recommending the commodity based on feature ordering, including following device:
Benchmark image Hash codes acquiring unit reads benchmark commodity for client, obtains from character code data library
The benchmark image Hash codes of the benchmark commodity;
Similar commodity group generation unit obtains the reference map from the character code data library for the client
As the Hamming distance between Hash codes and all image hash codes, image is successively obtained according to the Hamming distance from small to large
Similar commodity corresponding to Hash codes generate similar commodity group;
The similar commodity group is added to commercial product recommending list for the client by commercial product recommending list generation unit
In.
Further, further include following device:
Commodity recognition unit has been purchased, user has been obtained for the client and has purchased inventory, be included within user and purchased inventory
In similar commodity deleted from similar commodity group;
Commodity scoring list acquiring unit obtains commodity scoring list for the client, and the commodity is scored
Commodity in list are successively added in the commercial product recommending list from high to low according to scoring, and the commodity scoring list is root
The list obtained according to Collaborative Filtering Recommendation Algorithm;
Whether similar commodity adding unit successively detects the similar commodity in the similar commodity group for the client
It is included in the commercial product recommending list, if including, next similar commodity is detected, otherwise by the similar commodity
It is added in commercial product recommending list.
The third aspect, the present invention provides the commercial product recommending equipment based on feature ordering, including at least one control to handle
Device and memory for being communicated to connect at least one control processor;Memory, which is stored with, to be handled by least one control
The instruction that device executes, instruction is executed by least one control processor, so that at least one control processor is able to carry out as above
The Method of Commodity Recommendation based on feature ordering.
Fourth aspect, the present invention provides a kind of computer readable storage medium, computer-readable recording medium storage has
Computer executable instructions, computer executable instructions are for executing computer as described above based on the commodity of feature ordering
Recommended method.
5th aspect, the present invention also provides a kind of computer program product, the computer program product includes storage
Computer program on computer readable storage medium, the computer program include program instruction, when described program instructs
When being computer-executed, execute computer as described above based on the Method of Commodity Recommendation of feature ordering.
The one or more technical solutions provided in the embodiment of the present invention at least have the following beneficial effects: that the present invention adopts
With a kind of Method of Commodity Recommendation based on feature ordering, device and storage medium.Obtain reference map corresponding to benchmark commodity
As obtaining image hash code corresponding to other commodity, calculating benchmark image Hash from character code data library after Hash codes
Code and image wish the Hamming distance between code, form similar quotient according to the corresponding similar commodity of the acquisition from small to large of Hamming distance
Product group, and similar commodity are successively added in commercial product recommending list in order, for compared with the prior art, the present invention passes through
Hamming distance carries out feature ordering to the similarity of the appearance of commodity image, realizes the addition appearance phase in commercial product recommending list
As commodity.Meanwhile the application also equidistantly distributes commodity scoring list and similar commodity in commercial product recommending list,
User experience is greatly improved, the buying rate for improving user is conducive to.
Detailed description of the invention
The invention will be further described with example with reference to the accompanying drawing.
Fig. 1 is the flow chart for the Method of Commodity Recommendation based on feature ordering that the embodiment of the present invention one provides;
Fig. 2 is in the Method of Commodity Recommendation based on feature ordering of the offer of the embodiment of the present invention one by the similar commodity group
The flow chart being added in commercial product recommending list;
Fig. 3 is neural network used in the Method of Commodity Recommendation based on feature ordering of the offer of the embodiment of the present invention one
Structure chart;
Fig. 4 is the entire protocol figure for the Method of Commodity Recommendation based on feature ordering that the embodiment of the present invention one provides;
Fig. 5 is the schematic device of the device for recommending the commodity provided by Embodiment 2 of the present invention based on feature ordering;
Fig. 6 is the structural schematic diagram for the commercial product recommending equipment based on feature ordering that the embodiment of the present invention three provides.
Specific embodiment
Currently, shopping platform is done shopping in more and more consumers selection line with the development of e-commerce.In order to
The buying rate of consumer is improved, each electric business platform usually shows the interested Recommendations of consumer in some pages, therefore
The method of commercial product recommending more can be more effective to promoting to do shopping close to the interest of consumer.In currently existing scheme, mainly according to collaboration
Filtering recommendation algorithms carry out commercial product recommending, although this method can generate according to the similitude and prediction scoring of commodity and recommend quotient
A point list is judged, but collaborative filtering relies primarily on the verbal description to item property, for the commodity based on appearance,
The item property of verbal description can not intuitively embody the interest of consumer, therefore for the commodity based on appearance,
Collaborative Filtering Recommendation Algorithm is difficult to carry out valuable commercial product recommending.
Based on this, present invention employs Method of Commodity Recommendation, device and storage mediums based on feature ordering.Obtain benchmark
After benchmark image Hash codes corresponding to commodity, image Hash corresponding to other commodity is obtained from character code data library
Code, calculating benchmark image hash code and image wish the Hamming distance between code, corresponding according to the acquisition from small to large of Hamming distance
Similar commodity form similar commodity group, and similar commodity are successively added in commercial product recommending list in order, to compared with existing
For having technology, the present invention carries out feature ordering by similarity of the Hamming distance to the appearance of commodity image, realizes in quotient
The similar commodity of appearance are added in product recommendation list.Meanwhile the application also carries out commodity scoring list and similar commodity equidistantly
From being distributed in commercial product recommending list, user experience is greatly improved, is conducive to the buying rate for improving user.
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that described herein, specific examples are only used to explain the present invention, not
For limiting the present invention.
It should be noted that each feature in the embodiment of the present invention can be combined with each other, in this hair if do not conflicted
Within bright protection scope.In addition, though having carried out functional module division in schematic device, shows patrol in flow charts
Sequence is collected, but in some cases, it can be shown in the sequence execution in the module division being different from device or flow chart
The step of out or describing.
Referring to Fig.1, the present invention provides the Method of Commodity Recommendation based on feature ordering, comprising the following steps:
Step S100, client read benchmark commodity, the benchmark of the benchmark commodity are obtained from character code data library
Image hash code;
Step S200, the client obtain the benchmark image Hash codes and are owned from the character code data library
Image hash code between Hamming distance, according to the Hamming distance from small to large successively obtain image hash code corresponding to
Similar commodity generate similar commodity group;
The similar commodity group is added in commercial product recommending list by step S300, the client.
Wherein, in the present embodiment, the benchmark commodity are the commodity for generating recommendation list, can be user and have purchased
Any commodity in list are also possible to commodity of the user just in browsing pages.Preferably worked as according to client in the present embodiment
Preceding locating page situation determines, when the current page of client is the commodity page, using the commodity in the commodity page as base
It is suitable according to the time buying have been purchased the commodity in list when the current page of user terminal is the noncommodity page by quasi- commodity by user
Sequence is arranged as benchmark commodity.
Wherein, in the present embodiment, the character code data library can be the update when adding each commodity, can also be with
It is to regularly update, is preferably regularly updated in the present embodiment, since the addition of commodity and deletion frequency are very fast in electric business platform, frequently
Operation will increase server burden, therefore regularly updates and can more reasonably utilize server resource, such as update daily once.
Wherein, in the present embodiment, the similarity between exterior of commodity can be expressed using any type of feature coding,
Hash codes are preferably used in the present embodiment, the feature extracted by convolutional neural networks algorithm can preferably express commodity
External appearance characteristic, and external appearance characteristic is iterated after quantization is converted into Hash codes, the Hamming distance between different Hash codes is got over
Closely, it is meant that similar feature is more, also means that the appearance similarity degree of two pieces commodity is higher.
Wherein, in the present embodiment, the quantity that similar commodity obtain can be any amount, be also possible in the client
In advance preset quantity preferably presets the quantity of acquired similar commodity in the present embodiment, i.e., in the client
Set the quantity of commodity in similar commodity group.Since the commodity in electric business platform are more, not to the commodity number in similar commodity group
Amount is defined, and the burden that will lead to calculating is larger, and the commodity similitude excessive for Hamming distance is also poor, and do not have compared with
Good point of reference, therefore the present embodiment commodity amount in pre-set similar commodity group in the client are enough obtaining
Commodity when stop calculate, be more advantageous to save server resource.
With reference to Fig. 3, further, in another embodiment of the present invention, described image Hash codes are rolled up by iterative quantization
It is arranged obtained by the extracted characteristics of image of product neural network, between the convolutional layer of the convolutional neural networks and full articulamentum free
Between pyramid pond layer.
Wherein, in the present embodiment, it can be extracted by any type of feature extraction network for the characteristics of image of coding,
Preferred convolutional neural networks in the present embodiment, since convolutional neural networks have, level is deep, the high feature of dimension, extracted spy
Sign more usually can accurately describe the feature of figure.For example, preferably using convolutional Neural as shown in Figure 2 in the present embodiment
Network, including 9 layers of convolutional layer, 2 layers of full articulamentum and 1 layer softmax layers, wherein the convolution kernel size of convolutional layer is 3 × 3, step
A length of 1, the convolution kernel of full articulamentum is 2 × 2, step-length 2.
Wherein, in the present embodiment, convolutional Neural network can be used only and carries out image characteristics extraction, it can also be in convolution
Spatial pyramid pond layer is added between layer and maximum pond layer.Since the image of different commodity is of different sizes, and full articulamentum
The figure of input fixed size is usually required, spatial pyramid pond layer can be defeated after obtaining different size of input picture
The image of fixed size out, it is advantageously ensured that successfully extracting feature.In the present embodiment 3 layers in the layer of preferable space pyramid pond
Dimensional parameter is respectively 8 × 8,4 × 4 and 2 × 2, in the case where saving as M × N × 3, passes through spatial pyramid in input picture
Output after the layer of pond is 84 × 512 dimensions, that is, the characteristic dimension being input in full articulamentum is 4096 dimensions.
Wherein, in the present embodiment, the conversion of Hash codes can be any way, preferred iterative quantization in the present embodiment.
Iterative quantization can constantly approach the quantization error minimum value between characteristics of image and the Hash codes for setting bit number, to realize
The minimum of quantization error, the Hash codes obtained are more accurate.
Further, in another embodiment of the present invention, the similar commodity in the similar commodity group are according to its image
Hamming distance between Hash codes and the benchmark image Hash codes sorts from small to large.
Wherein, in the present embodiment, acquired similar commodity can be arranged according to random order, such as commodity price,
Commodity purchasing amount etc. preferably sorts according to the Hamming distance between image hash code, preferably body from small to large in the present embodiment
The correlation of existing exterior of commodity.
It further, in another embodiment of the present invention, further include that the client obtains user and purchased inventory, it will
It is deleted including having purchased the similar commodity in inventory from similar commodity group in user.
Wherein, in the present embodiment, any commodity can be recommended, preferably user is not bought in the present embodiment
Commodity are recommended, and are conducive to improve user experience.Preferably, basis has one by one purchased similar commodity with user in the present embodiment
Commodity in inventory are compared, to judge whether non-purchased goods, avoid entire service while be compared, cause to take
Device burden of being engaged in is larger.
Further, in another embodiment of the present invention, described that the similar commodity group is added to commercial product recommending column
In table specifically includes the following steps:
Step S310, the client obtain commodity and score list, and by the commodity score the commodity in list according to
Scoring is successively added to from high to low in the commercial product recommending list, and the commodity scoring list is to be calculated according to collaborative filtering recommending
The list that method obtains;
Whether step S320, the similar commodity that the client successively detects in the similar commodity group are included in the quotient
In product recommendation list, if including, next similar commodity are detected, the similar commodity are otherwise added to commodity and are pushed away
It recommends in list.
Wherein, in the present embodiment, the commodity in similar commodity group can be used only as the quotient in commercial product recommending list
Product, can also be in conjunction with the recommendation list of other forms.Preferably commodity are obtained according to Collaborative Filtering Recommendation Algorithm in the present embodiment to comment
Point list, it is preferable that the commodity for generating commodity scoring list be purchased inventory and the inventory that scored in commodity, according to
Family buying habit and scoring habit obtain out the prediction scoring of commodity, to obtain commodity scoring list.
Wherein, in the present embodiment, since first the commodity in commodity scoring list being added in commercial product recommending list, because
This preferably judges whether similar commodity to be included in commercial product recommending list in order to avoid repeating commodity in the present embodiment
In, if similar commodity in commercial product recommending list, are not repeated to add, continue to sentence to the similar commodity of next sequence
It is disconnected.
Further, in another embodiment of the present invention, the similar commodity are in medium of the commercial product recommending list
Every addition.
Wherein, in the present embodiment, similar commodity can be added by arbitrary form, preferably added at equal intervals in the present embodiment
Add, i.e., in commercial product recommending list existing commodity sequence moderate distance insertion, such as existing commodity 1, commodity 2 and commodity 3, point
Similar commodity 1, similar commodity 2 and similar commodity 3 are not added, be arranged in order after addition as commodity 1, similar commodity 1, commodity
2, similar commodity 2, commodity 3, similar commodity 3.It is capable of providing better user experience.
With reference to Fig. 4, in addition, another embodiment of the invention additionally provides the Method of Commodity Recommendation based on feature ordering,
The following steps are included:
Step S4100, client reads benchmark commodity, and the benchmark of the benchmark commodity is obtained from character code data library
Image hash code;
Step S4210, the described client obtains the benchmark image Hash codes and institute from the character code data library
Hamming distance between some image hash codes successively obtains corresponding to image hash code from small to large according to the Hamming distance
Similar commodity, generate similar commodity group;
Step S4220, the client obtains user and has purchased inventory, is included within user and has purchased similar commodity in inventory
It is deleted from similar commodity group;
Step S4310, the described client obtains commodity scoring list according to Collaborative Filtering Recommendation Algorithm;
Step S4320, the commodity in commodity scoring list are successively added to the quotient according to scoring from high to low
In product recommendation list;
Step S4400, detect whether similar commodity have been included in the commercial product recommending list one by one, if so, executing step
Rapid S4520, it is no to then follow the steps S4510;
Step S4510, the similar commodity are added at equal intervals in commercial product recommending list;
Step S4520, similar commodity are not added in commercial product recommending list.
Referring to Fig. 5, the second embodiment of the present invention additionally provides the device for recommending the commodity based on feature ordering, is based at this
In the device for recommending the commodity 5000 of feature ordering, including but not limited to: benchmark image Hash codes acquiring unit 5100, similar commodity
Group generation unit 5200, commercial product recommending list generation unit 5300.
Wherein, benchmark image Hash codes acquiring unit 5100 reads benchmark commodity for client, from character code data
The benchmark image Hash codes of the benchmark commodity are obtained in library;
Similar commodity group generation unit 5200 obtains the base from the character code data library for the client
Hamming distance between quasi- image hash code and all image hash codes successively obtains from small to large according to the Hamming distance
Similar commodity corresponding to image hash code generate similar commodity group;
The similar commodity group is added to commercial product recommending for the client by commercial product recommending list generation unit 5300
In list.
Further, in another embodiment of the invention, further include but be not limited to: having purchased commodity recognition unit 5210, quotient
Judge point list acquiring unit 5310 and similar commodity adding unit 5320;
Wherein, it has purchased commodity recognition unit 5210 and has purchased inventory for client acquisition user, be included within user
The similar commodity purchased in inventory are deleted from similar commodity group;
Commodity score list acquiring unit 5310 for client acquisition commodity scoring list, and the commodity are commented
The commodity divided in list are successively added in the commercial product recommending list from high to low according to scoring, and the commodity scoring list is
The list obtained according to Collaborative Filtering Recommendation Algorithm;
Similar commodity adding unit 5320 successively detects the similar commodity in the similar commodity group for the client
Whether it is included in the commercial product recommending list, if including, next similar commodity is detected, it otherwise will be described similar
Commodity are added in commercial product recommending list.
It should be noted that by the device for recommending the commodity in this present embodiment based on feature ordering and above-mentioned based on spy
The Method of Commodity Recommendation of sign sequence is based on identical inventive concept, and therefore, the corresponding contents in embodiment of the method are equally applicable to
Present apparatus embodiment, and will not be described here in detail.
Referring to Fig. 6, the embodiment of the invention also provides the commercial product recommending equipment based on feature ordering, should be based on feature ordering
Commercial product recommending equipment 6000 can be any type of intelligent terminal, such as mobile phone, tablet computer, personal computer etc..
Specifically, being somebody's turn to do the commercial product recommending equipment 6000 based on feature ordering includes: one or more control processors 6001
With memory 6002, in Fig. 6 by taking a control processor 6001 as an example.
Control processor 6001 can be connected with memory 6002 by bus or other modes, by total in Fig. 6
For line connection.
Memory 6002 be used as a kind of non-transient computer readable storage medium, can be used for storing non-transient software program,
Non-transitory computer executable program and module, such as the commercial product recommending equipment based on feature ordering in the embodiment of the present invention
Corresponding program instruction/module, for example, benchmark image Hash codes acquiring unit 5100 shown in Fig. 5 and similar commodity group are raw
At unit 5200.Control processor 6001 by operation be stored in memory 6002 non-transient software program, instruction and
Module is realized thereby executing the various function application and data processing of the device for recommending the commodity 5000 based on feature ordering
The Method of Commodity Recommendation based on feature ordering of above method embodiment.
Memory 6002 may include storing program area and storage data area, wherein storing program area can store operation system
Application program required for system, at least one function;Storage data area can be stored according to the commercial product recommending dress based on feature ordering
It sets 5000 and uses created data etc..In addition, memory 6002 may include high-speed random access memory, can also wrap
Include non-transient memory, a for example, at least disk memory, flush memory device or other non-transient solid-state memories.?
In some embodiments, optional memory 6002 includes the memory remotely located relative to control processor 6001, these are remote
Journey memory can extremely be somebody's turn to do the commercial product recommending equipment 6000 based on feature ordering by network connection.The example of above-mentioned network includes
But be not limited to internet, intranet, local area network, mobile radio communication and combinations thereof.
One or more of modules are stored in the memory 6002, at by one or more of controls
When managing the execution of device 6001, the Method of Commodity Recommendation based on feature ordering in above method embodiment is executed, for example, more than executing
Method and step S310 to S320 in method and step S100 to S300 in Fig. 1 of description, Fig. 2 realizes the device in Fig. 5
The function of 5100-5300.
The embodiment of the invention also provides a kind of computer readable storage medium, the computer-readable recording medium storage
There are computer executable instructions, which is executed by one or more control processors, for example, by Fig. 6
A control processor 6001 execute, may make said one or multiple control processors 6001 to execute above method embodiment
In the Method of Commodity Recommendation based on feature ordering, for example, execute the method and step S100 to S300 in Fig. 1 described above,
Method and step S310 to S320 in Fig. 2 realizes the function of the device 5100-5300 in Fig. 5.
The apparatus embodiments described above are merely exemplary, wherein described, device can as illustrated by the separation member
It is physically separated with being or may not be, it can it is in one place, or may be distributed over multiple network dresses
It sets.Some or all of the modules therein can be selected to achieve the purpose of the solution of this embodiment according to the actual needs.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can borrow
Help software that the mode of general hardware platform is added to realize.It will be appreciated by those skilled in the art that realizing in above-described embodiment method
All or part of the process is relevant hardware can be instructed to complete by computer program, and the program can be stored in one
In computer-readable storage medium, the program is when being executed, it may include such as the process of the embodiment of the above method.Wherein, institute
The storage medium stated can be magnetic disk, CD, read-only memory (ReadOnly Memory, ROM) or random access memory
(Random Access Memory, RAM) etc..
It is to be illustrated to preferable implementation of the invention, but the invention is not limited to above-mentioned embodiment party above
Formula, those skilled in the art can also make various equivalent variations on the premise of without prejudice to spirit of the invention or replace
It changes, these equivalent deformations or replacement are all included in the scope defined by the claims of the present application.
Claims (10)
1. the Method of Commodity Recommendation based on feature ordering, which comprises the following steps:
Client reads benchmark commodity, and the benchmark image Hash codes of the benchmark commodity are obtained from character code data library;
The client obtains the benchmark image Hash codes and all image hash codes from the character code data library
Between Hamming distance, similar commodity corresponding to image hash code are successively obtained according to the Hamming distance from small to large, it is raw
At similar commodity group;
The similar commodity group is added in commercial product recommending list by the client.
2. the Method of Commodity Recommendation according to claim 1 based on feature ordering, it is characterised in that: described image Hash codes
As obtained by the extracted characteristics of image of iterative quantization convolutional neural networks, the convolutional layer of the convolutional neural networks and full connection
Spatial pyramid pond layer is provided between layer.
3. the Method of Commodity Recommendation according to claim 1 based on feature ordering, it is characterised in that: the similar commodity group
In similar commodity sorted from small to large according to the Hamming distance between its image hash code and the benchmark image Hash codes.
4. the Method of Commodity Recommendation according to claim 1 based on feature ordering, which is characterized in that further include: the visitor
Family end obtains user and has purchased inventory, is included within the similar commodity that user has purchased in inventory and deletes from similar commodity group.
5. the Method of Commodity Recommendation according to claim 1 based on feature ordering, which is characterized in that it is described will be described similar
Commodity group be added in commercial product recommending list specifically includes the following steps:
The client obtains commodity and scores list, and by the commodity score the commodity in list according to scoring from high to low according to
Secondary to be added in the commercial product recommending list, the commodity scoring list is the list obtained according to Collaborative Filtering Recommendation Algorithm;
Whether the similar commodity that the client successively detects in the similar commodity group are included in the commercial product recommending list,
If including detecting to next similar commodity, otherwise the similar commodity being added in commercial product recommending list.
6. the Method of Commodity Recommendation according to claim 5 based on feature ordering, it is characterised in that: the similar commodity exist
The existing commodity equal intervals of the commercial product recommending list add.
7. the device for recommending the commodity based on feature ordering, which is characterized in that including following device:
Benchmark image Hash codes acquiring unit reads benchmark commodity for client, from character code data library described in acquisition
The benchmark image Hash codes of benchmark commodity;
Similar commodity group generation unit obtains the benchmark image from the character code data library for the client and breathes out
Hamming distance between uncommon code and all image hash codes successively obtains image Hash according to the Hamming distance from small to large
Similar commodity corresponding to code, generate similar commodity group;
The similar commodity group is added in commercial product recommending list by commercial product recommending list generation unit for the client.
8. the device for recommending the commodity according to claim 7 based on feature ordering, which is characterized in that further include following dress
It sets:
Commodity recognition unit has been purchased, user has been obtained for the client and has purchased inventory, be included within user and purchased in inventory
Similar commodity are deleted from similar commodity group;
Commodity score list acquiring unit, for the client obtain commodity score list, and by the commodity score list
In commodity be successively added in the commercial product recommending list from high to low according to scoring, commodity scoring list is according to association
The list obtained with filtering recommendation algorithms;Similar commodity adding unit successively detects the similar commodity for the client
Whether the similar commodity in group are included in the commercial product recommending list, if including, detect to next similar commodity,
Otherwise the similar commodity are added in commercial product recommending list.
9. a kind of commercial product recommending equipment based on feature ordering, it is characterised in that: including at least one control processor and be used for
The memory communicated to connect at least one described control processor;The memory is stored with can be by least one described control
The instruction that processor processed executes, described instruction are executed by least one described control processor, so that at least one described control
Processor is able to carry out the Method of Commodity Recommendation as claimed in any one of claims 1 to 6 based on feature ordering.
10. a kind of computer readable storage medium, it is characterised in that: the computer-readable recording medium storage has computer can
It executes instruction, the computer executable instructions are as claimed in any one of claims 1 to 6 based on spy for executing computer
Levy the Method of Commodity Recommendation of sequence.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910191547.7A CN110060121A (en) | 2019-03-14 | 2019-03-14 | Method of Commodity Recommendation, device and storage medium based on feature ordering |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910191547.7A CN110060121A (en) | 2019-03-14 | 2019-03-14 | Method of Commodity Recommendation, device and storage medium based on feature ordering |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110060121A true CN110060121A (en) | 2019-07-26 |
Family
ID=67317029
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910191547.7A Pending CN110060121A (en) | 2019-03-14 | 2019-03-14 | Method of Commodity Recommendation, device and storage medium based on feature ordering |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110060121A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110659962A (en) * | 2019-09-20 | 2020-01-07 | 苏州大学 | Commodity information output method and related device |
CN110992141A (en) * | 2019-11-28 | 2020-04-10 | 上海识装信息科技有限公司 | Method and device for recommending commodities based on identification of commodity images and electronic equipment |
CN113377973A (en) * | 2021-06-10 | 2021-09-10 | 电子科技大学 | Article recommendation method based on countermeasures hash |
CN113763067A (en) * | 2020-06-19 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Article information pushing method, device, equipment and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101657588B1 (en) * | 2015-11-13 | 2016-09-19 | (주)제이케이소프트 | System for Providing Product Dealing Information based on Product Image using Web and Method therefor |
CN106202362A (en) * | 2016-07-07 | 2016-12-07 | Tcl集团股份有限公司 | Image recommendation method and image recommendation device |
CN109146640A (en) * | 2018-08-30 | 2019-01-04 | 湖北工业大学 | Product search method and system on a kind of line |
CN109165306A (en) * | 2018-08-09 | 2019-01-08 | 长沙理工大学 | Image search method based on the study of multitask Hash |
CN109447713A (en) * | 2018-10-31 | 2019-03-08 | 国家电网公司 | A kind of recommended method and device of knowledge based map |
-
2019
- 2019-03-14 CN CN201910191547.7A patent/CN110060121A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101657588B1 (en) * | 2015-11-13 | 2016-09-19 | (주)제이케이소프트 | System for Providing Product Dealing Information based on Product Image using Web and Method therefor |
CN106202362A (en) * | 2016-07-07 | 2016-12-07 | Tcl集团股份有限公司 | Image recommendation method and image recommendation device |
CN109165306A (en) * | 2018-08-09 | 2019-01-08 | 长沙理工大学 | Image search method based on the study of multitask Hash |
CN109146640A (en) * | 2018-08-30 | 2019-01-04 | 湖北工业大学 | Product search method and system on a kind of line |
CN109447713A (en) * | 2018-10-31 | 2019-03-08 | 国家电网公司 | A kind of recommended method and device of knowledge based map |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110659962A (en) * | 2019-09-20 | 2020-01-07 | 苏州大学 | Commodity information output method and related device |
CN110659962B (en) * | 2019-09-20 | 2022-03-15 | 苏州大学 | Commodity information output method and related device |
CN110992141A (en) * | 2019-11-28 | 2020-04-10 | 上海识装信息科技有限公司 | Method and device for recommending commodities based on identification of commodity images and electronic equipment |
CN113763067A (en) * | 2020-06-19 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Article information pushing method, device, equipment and storage medium |
CN113377973A (en) * | 2021-06-10 | 2021-09-10 | 电子科技大学 | Article recommendation method based on countermeasures hash |
CN113377973B (en) * | 2021-06-10 | 2022-06-14 | 电子科技大学 | Article recommendation method based on countermeasures hash |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110060121A (en) | Method of Commodity Recommendation, device and storage medium based on feature ordering | |
US9607010B1 (en) | Techniques for shape-based search of content | |
CN109166017A (en) | Method for pushing, device, computer equipment and storage medium based on reunion class | |
CN109508419A (en) | A kind of recommended method and system of knowledge based study | |
CN103092861B (en) | A kind of choosing method of commodity representative picture and system | |
WO2015148422A1 (en) | Recommendation system with dual collaborative filter usage matrix | |
CN103248705B (en) | Server, client and method for processing video frequency | |
CN105022811A (en) | Menu generating method and terminal | |
CN103377443A (en) | Online trade platform and processing method thereof | |
CN111695960A (en) | Object recommendation system, method, electronic device and storage medium | |
CN110674391B (en) | Product data pushing method and system based on big data and computer equipment | |
CN106708883A (en) | Recommendation method and apparatus | |
CN106354734A (en) | Method and device for providing business object information | |
CN105335368B (en) | A kind of product clustering method and device | |
CN108805598A (en) | Similarity information determines method, server and computer readable storage medium | |
CN109962975A (en) | Information-pushing method, device, electronic equipment and system based on object identification | |
CN104809637A (en) | Commodity recommending method and system realized by computer | |
CN104615639B (en) | A kind of method and apparatus for providing the presentation information of picture | |
CN112307352B (en) | Content recommendation method, system, device and storage medium | |
CN104615721A (en) | Method and system for recommending communities based on returned goods related information | |
JP2018101339A (en) | Estimation device, estimation method, and estimation program | |
CN115423555A (en) | Commodity recommendation method and device, electronic equipment and storage medium | |
CN111753180A (en) | Search method, search device, electronic equipment and computer storage medium | |
CN109919433A (en) | Order information generation method, device and computer equipment | |
CN105224588A (en) | According to method and the device of photo dominant hue generation background music |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190726 |