CN109800769A - Product classification control method, device, computer equipment and storage medium - Google Patents
Product classification control method, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN109800769A CN109800769A CN201811577748.2A CN201811577748A CN109800769A CN 109800769 A CN109800769 A CN 109800769A CN 201811577748 A CN201811577748 A CN 201811577748A CN 109800769 A CN109800769 A CN 109800769A
- Authority
- CN
- China
- Prior art keywords
- product
- information
- sorted
- target product
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000003860 storage Methods 0.000 title claims abstract description 13
- 238000013507 mapping Methods 0.000 claims abstract description 24
- 230000015654 memory Effects 0.000 claims description 17
- 238000012545 processing Methods 0.000 claims description 15
- 238000006116 polymerization reaction Methods 0.000 claims description 8
- 238000004458 analytical method Methods 0.000 claims description 7
- 238000003062 neural network model Methods 0.000 claims description 4
- 239000000047 product Substances 0.000 description 426
- 238000010586 diagram Methods 0.000 description 18
- 240000003768 Solanum lycopersicum Species 0.000 description 15
- 235000007688 Lycopersicon esculentum Nutrition 0.000 description 14
- 244000061456 Solanum tuberosum Species 0.000 description 11
- 235000002595 Solanum tuberosum Nutrition 0.000 description 11
- 235000013580 sausages Nutrition 0.000 description 11
- 230000006870 function Effects 0.000 description 7
- 238000004519 manufacturing process Methods 0.000 description 7
- 238000007630 basic procedure Methods 0.000 description 6
- 238000004378 air conditioning Methods 0.000 description 5
- 235000002639 sodium chloride Nutrition 0.000 description 5
- 244000288157 Passiflora edulis Species 0.000 description 4
- 235000000370 Passiflora edulis Nutrition 0.000 description 4
- 238000013527 convolutional neural network Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000009885 systemic effect Effects 0.000 description 4
- FAPWRFPIFSIZLT-UHFFFAOYSA-M Sodium chloride Chemical group [Na+].[Cl-] FAPWRFPIFSIZLT-UHFFFAOYSA-M 0.000 description 3
- 235000013622 meat product Nutrition 0.000 description 3
- 238000003058 natural language processing Methods 0.000 description 3
- 150000003839 salts Chemical class 0.000 description 3
- 235000011511 Diospyros Nutrition 0.000 description 2
- 244000236655 Diospyros kaki Species 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 235000019692 hotdogs Nutrition 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 210000005036 nerve Anatomy 0.000 description 2
- 239000002356 single layer Substances 0.000 description 2
- 239000011780 sodium chloride Substances 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 244000099147 Ananas comosus Species 0.000 description 1
- 235000007119 Ananas comosus Nutrition 0.000 description 1
- 244000205754 Colocasia esculenta Species 0.000 description 1
- 235000006481 Colocasia esculenta Nutrition 0.000 description 1
- 241000196324 Embryophyta Species 0.000 description 1
- 244000061458 Solanum melongena Species 0.000 description 1
- 235000002597 Solanum melongena Nutrition 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 235000021028 berry Nutrition 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 239000006227 byproduct Substances 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 210000000936 intestine Anatomy 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the invention discloses a kind of product classification control method, device, computer equipment and storage medium, include the following steps: the name information for obtaining product to be sorted；The target product information of the target product to match with name information is searched according to preset matching strategy, wherein target product information includes the tag along sort information of the affiliated category of employment of target product；The tag along sort of product to be sorted is set according to tag along sort information, so that product to be sorted is matched with the affiliated category of employment mapping of target product.The tag along sort of product to be sorted is arranged by the tag along sort information according to existing target product for the embodiment of the present invention, it can be according to the corresponding relationship between existing product and industry, to derive industry label belonging to product to be sorted and be mapped, raising treats sort product and carries out the accuracy that classification labels, full-automatic classification labels mode, the workload manually to label can be effectively saved, treatment effeciency is high.
Description
Technical field
The present embodiments relate to product assortment processing technology field, especially a kind of product classification control method, device,
Computer equipment and storage medium.
Background technique
During life and work, people often touch some unknown new things, such as finger tip gyro, are
An a kind of bearing symmetrical structure, the nicknack that can be dallied on finger, before finger tip gyro emerges, people are not aware that
There is this toy of finger tip gyro, and after new things enter in people's lives, people need to divide unknown new things
Class is to judge industry belonging to new things or classification.
The classification work of existing new things is usually to be completed by manually, still, is manually labelled to new things
Mode is easy to appear the situation for leading to classification inaccuracy because of the influence of human subjective's factor, moreover, manually inferring new things
The heavy workload of industry or classification and treatment effeciency is not high increases new things and classifies the artificial and time cost to label.
Summary of the invention
The embodiment of the present invention provides a kind of according to existing product and the corresponding relationship of the industry deduction affiliated industry of new product
Product classification control method, device, computer equipment and the storage medium of label.
In order to solve the above technical problems, the technical solution that the embodiment of the invention uses is: providing a kind of production
Product classification control method, includes the following steps:
Obtain the name information of product to be sorted；
The target product information of the target product to match with the name information is searched according to preset matching strategy,
In, the target product information includes the tag along sort information of the affiliated category of employment of the target product；
The tag along sort of the product to be sorted is set according to the tag along sort information so that the product to be sorted with
The affiliated category of employment mapping matching of target product.
Optionally, the target that the target product to match with the name information is searched according to preset matching strategy
The step of product information, including such as following step:
The attribute that analysis generates the product to be sorted is carried out to the name information according to preset semantic understanding rule
Information；
It is searched in preset product database and the attribute information similarity highest and phase according to the matching strategy
It is higher than the target product information of preset matching threshold value like degree.
Optionally, the tag along sort that the product to be sorted is set according to the tag along sort information, so that described
Further include such as following step after the step of product to be sorted is matched with the affiliated category of employment mapping of the target product:
By the tag along sort information of the name information of the product to be sorted and the affiliated category of employment of the target product into
The product information of product to be sorted described in row combination producing；
The product information of the product to be sorted is stored into the product database.
Optionally, the target that the target product to match with the name information is searched according to preset matching strategy
The step of product information, including such as following step:
Polymerization is carried out to the name information and target product information according to preset clustering rule and generates clustering information；
The industry label of the clustering information mapped category of employment is obtained as the tag along sort information.
Optionally, the target that the target product to match with the name information is searched according to preset matching strategy
The step of product information, including such as following step:
Obtain the product image of the product to be sorted；
The target product information of the target product to match with the product image is searched according to the product image.
Optionally, the mesh that the target product to match with the product image is searched according to the product image
The step of marking product information, including such as following step:
The product image is input in preset product identification model, the product identification model is that training extremely restrains
Convolutional neural networks model；
Obtain the target product information of the target product of the product identification model output.
In order to solve the above technical problems, the embodiment of the present invention also provides a kind of product classification control device, comprising:
First obtains module, for obtaining the name information of product to be sorted；
First processing module, for searching the target product to match with the name information according to preset matching strategy
Target product information, wherein the target product information include the affiliated category of employment of the target product tag along sort letter
Breath；
First execution module, for the tag along sort of the product to be sorted to be arranged according to the tag along sort information, with
Match the product to be sorted with the affiliated category of employment mapping of the target product.
Optionally, further includes:
First processing submodule generates institute for carrying out analysis to the name information according to preset semantic understanding rule
State the attribute information of product to be sorted；
First implementation sub-module, for being searched in preset product database according to the matching strategy and the attribute
Information similarity highest and similarity are higher than the target product information of preset matching threshold value.
Optionally, further includes:
Second execution module, for by the name information of the product to be sorted and the affiliated category of employment of the target product
Tag along sort information be combined the product information for generating the product to be sorted；
Memory module, for storing the product information of the product to be sorted into the product database.
Optionally, further includes:
Second processing submodule, for being carried out according to preset clustering rule to the name information and target product information
Polymerization generates clustering information；
First acquisition submodule, for obtaining described in the industry label conduct of the clustering information mapped category of employment
Tag along sort information.
Optionally, further includes:
Second acquisition submodule, for obtaining the product image of the product to be sorted；
Second implementation sub-module, for searching the target to match with the product image according to the product image
The target product information of product.
Optionally, further includes:
Third implementation sub-module, for the product image to be input in preset product identification model, the product
Identification model is trained to convergent convolutional neural networks model；
Third acquisition submodule, the target product letter of the target product for obtaining the product identification model output
Breath.
In order to solve the above technical problems, the embodiment of the present invention also provides a kind of computer equipment, including memory and processing
Device is stored with computer-readable instruction in the memory, when the computer-readable instruction is executed by the processor, so that
The processor executes the step of the said goods classification control method.
In order to solve the above technical problems, the embodiment of the present invention also provides a kind of storage Jie for being stored with computer-readable instruction
Matter, when the computer-readable instruction is executed by one or more processors, so that one or more processors execute above-mentioned production
The step of product classification control method.
The embodiment of the present invention has the beneficial effect that by being searched and product to be sorted according to the name information of product to be sorted
The target product to match, and the target product information of target product is obtained, point including the affiliated category of employment of the target product
Then the tag along sort of product to be sorted is arranged in class label information according to the tag along sort information, so that product to be sorted
It is matched with target product affiliated category of employment mapping, i.e., by category of employment belonging to product assortment to be sorted to target product,
It can be according to the corresponding relationship between existing product and industry, to derive industry label belonging to product to be sorted and be reflected
It penetrates, improves and treat sort product and carry out the accuracy that labels of classification, full-automatic classification labels mode, can effectively save
The workload manually to label, treatment effeciency are high.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment
Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for
For those skilled in the art, without creative efforts, it can also be obtained according to these attached drawings other attached
Figure.
Fig. 1 is the basic procedure schematic diagram of product classification of embodiment of the present invention control method；
Fig. 2 is the flow diagram that the embodiment of the present invention searches target product information；
Fig. 3 is the flow diagram that the embodiment of the present invention stores the product to be sorted for being set tag along sort；
Fig. 4 is the flow diagram that the embodiment of the present invention obtains tag along sort information；
Fig. 5 is the flow diagram that the embodiment of the present invention obtains target product information；
Fig. 6 is the flow diagram that the embodiment of the present invention obtains target product information according to image；
Fig. 7 is product classification of embodiment of the present invention control device basic structure schematic diagram；
Fig. 8 is computer equipment of embodiment of the present invention basic structure block diagram.
Specific embodiment
In order to enable those skilled in the art to better understand the solution of the present invention, below in conjunction in the embodiment of the present invention
Attached drawing, technical scheme in the embodiment of the invention is clearly and completely described.
In some processes of the description in description and claims of this specification and above-mentioned attached drawing, contain according to
Multiple operations that particular order occurs, but it should be clearly understood that these operations can not be what appears in this article suitable according to its
Sequence is executed or is executed parallel, and serial number of operation such as 101,102 etc. is only used for distinguishing each different operation, serial number
It itself does not represent and any executes sequence.In addition, these processes may include more or fewer operations, and these operations can
To execute or execute parallel in order.It should be noted that the description such as " first " herein, " second ", is for distinguishing not
Same message, equipment, module etc., does not represent sequencing, does not also limit " first " and " second " and be different type.
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, those skilled in the art's every other implementation obtained without creative efforts
Example, shall fall within the protection scope of the present invention.
Embodiment 1
It is the basic procedure schematic diagram of the present embodiment product classification control method referring specifically to Fig. 1, Fig. 1.
As shown in Figure 1, a kind of product classification control method, includes the following steps:
S1100, the name information for obtaining product to be sorted；
Product to be sorted refers to the product of newly-increased and unknown type, and the name information of product to be sorted includes but is not limited to
Scientific name, alias and the other common first names of product to be sorted, such as: the scientific name of salt is sodium chloride, belongs in science and uses
Major name, and the alias that is often used includes refined salt, crude salt and dining table salt in living.When implementing, product to be sorted
Name information can be acquired by the information list of product to be sorted, such as product to be sorted is carried in information list
Name information and pictorial information etc., system can obtain the name information of product to be sorted according to the information list.
In one alternate embodiment, the name information of product to be sorted can be inputted by operator, with this hair
Bright product classification control method is applied to for PC (personal computer, personal computer) end, and system passes through the end PC
Display shows input frame to user, and user can input production to be sorted by the keyboard or touch screen at the end PC in input frame
The name information of product, such as during commodity are carried out classification and put by supermarket, supermarket newly into a collection of imported product, supermarket
Employee does not know the industry or classification of the imported product, so which kinds of goods does not know be placed into the imported product
In cabinet, at this point, the title of the imported product is input in system by the employee of supermarket, system can be obtained and be produced according to the import
The name information of product (product to be sorted) carries out classification to the imported product and labels.
S1200, the target product that the target product to match with the name information is searched according to preset matching strategy
Information, wherein the target product information includes the tag along sort information of the affiliated category of employment of the target product；
After the name information for obtaining product to be sorted, system is searched and the name information phase according to preset matching strategy
Matched target product, and the target product information of the target product is obtained, point including the affiliated category of employment of the target product
Class label information, wherein matching strategy is systemic presupposition for searching the mesh to match with the name information of product to be sorted
Product is marked, it, can be by the way that the title of the title of product to be sorted and target product be compared, to sentence when implementing
The similarity broken between product and target product to be sorted is preset when the similarity between product to be sorted and target product is higher than
Matching threshold (such as 95%) when, then judge product to be sorted and target product belong to same industry classification, when implementing, mesh
Mark product refers to that known or existing product, target product are provided with corresponding industry label, the sector label mapping
Category of employment belonging to target product, such as: target product is pineapple, and industry label mapped category of employment is tropical water
Fruit.
In one embodiment, system is provided with product database, and product database is used for several (examples of storage and management
As: 100,000,200,000 or 100 ten thousand) known to product and corresponding product product information, the product information of each product includes
The tag along sort information of the affiliated category of employment of the product, when system obtains the name information of product to be sorted, system traversal is produced
The target product that all products and lookup match for the name information in product database, when implementing, matching strategy
Working principle are as follows: in the name information and product database by calculating product to be sorted between the name information of each product
Hamming distance, and the target product information of the target product is obtained using the smallest product of Hamming distance as target product, including
The tag along sort information of the affiliated category of employment of the target product.
S1300, the tag along sort that the product to be sorted is set according to the tag along sort information, so that described to be sorted
Product is matched with the affiliated category of employment mapping of the target product.
After the tag along sort information for obtaining target product, product to be sorted is arranged according to the tag along sort information in system
Tag along sort, so that category of employment belonging to product mapping objects product to be sorted, realization is treated sort product and is divided
Class labels function.In one embodiment, then participle technique can be passed through by the name information of acquisition product to be sorted
Name information is handled, natural language similarity mode to target product similar with product to be sorted and mesh is passed through
The tag along sort information for marking the affiliated category of employment of product, is then arranged the industry mark of product to be sorted according to the tag along sort information
Sign the mapping matching for row label of going forward side by side.In one embodiment, by taking product to be sorted is sausage as an example, system is roasting by obtaining this
The name information of intestines, such as: entitled " sausage " of sausage, system are searched corresponding with " sausage " according to preset matching strategy
Target product, specifically, system searching to the target product to match with " sausage " includes ham sausage, bologna sausage and hot dog,
In, the tag along sort information of ham sausage, bologna sausage and the affiliated category of employment of hot dog is " meat products ", and system is further according to the contingency table
The industry label for signing information setting sausage is meat products, so that sausage is matched with " meat products " mapping, i.e., according to known to
The category of employment of product derives tag along sort belonging to product to be sorted, realize automation to unknown product to be sorted into
The capable function that labels, product classification treatment effeciency is high, saves the work that artificial judgment labels for increasing product to be sorted newly
Amount, product classification are more acurrate.
In one embodiment, matching strategy can also use the inference strategy of knowledge mapping, and knowledge mapping is substantially
Semantic network is a kind of data structure based on figure, is made of node (Point) (" entity ") and side (Edge) (" relationship ").
In knowledge mapping, each node is indicated present in real world " entity ", each edge " closing between entity and entity
System ".Knowledge mapping is the most effective representation of relationship.Generally, knowledge mapping is exactly all different types of letters
A relational network obtained from breath (Heterogeneous Information) links together.Knowledge mapping provide from
The angle of " relationship " goes the ability of problem analysis.It is to be sorted to derive the generic of product to be sorted by figure computational reasoning
The tag along sort of product is derived by the tag along sort of other known products in map, has certain AI reasoning energy
Power.For example: the relationship that we do not have evidence directly to indicate Mei Linda Gates and Seattle.However, we can pass through
Observe that " Mei Linda Gates-spouse-Bill Gates-chairman-Microsoft-is total comprising such paths in knowledge mapping
Portion is in-Seattle ", thus it is speculated that Seattle may be lived in by going out Melinda.Here it is one it is complete, go out result from relation derivation
Example.System can derive the tag along sort of product to be sorted by relation inference model, and the magnificent model of relationship includes being based on
Logic rules, knowledge based expression and it is based on deep learning, wherein the relation inference of logic-based rule includes but unlimited
In: Markov logical network (Markov Logic Network) model, based on the updated by probability of Bayesian network
(Probabilistic Relational Models) and FOIL (First Order based on statistical machine learning
Inductive Learner) algorithm.The relation inference of knowledge based expression includes but is not limited to: RESCAL tensor resolution model
(Tensor Factorization Model), SE (Structured Embedding) relation inference algorithm and TransE
(Translating Embedding) algorithm.Relation inference based on deep learning includes but is not limited to: single layer perceptron model
SLM (Single Layer Model), NTN nerve tensor model (Neural Tensor Networks) and DKRL
(Description-Embodied Knowledge Representation Learning) model.
The present embodiment by searching the target product to match with product to be sorted according to the name information of product to be sorted,
And the target product information of target product is obtained, and the tag along sort information including the affiliated category of employment of the target product, then root
The tag along sort of product to be sorted is set according to the tag along sort information, so that product to be sorted and the affiliated industry of target product
Classification mapping matching, i.e., by category of employment belonging to product assortment to be sorted to target product, can according to existing product and
Corresponding relationship between industry improves to derive industry label belonging to product to be sorted and be mapped to production to be sorted
Product carry out the accuracy that labels of classification, and full-automatic classification labels mode, can effectively save the work manually to label
Amount, treatment effeciency are high.
In one alternate embodiment, referring to Fig. 2, Fig. 2 is that one embodiment of the invention searches target product information
Flow diagram.
As shown in Fig. 2, step S1200 further includes such as following step:
S1210, the analysis generation product to be sorted is carried out to the name information according to preset semantic understanding rule
Attribute information；
After the name information for obtaining product to be sorted, system analyzes name information according to semantic understanding rule,
To obtain the attribute information of product to be sorted, semantic understanding rule is systemic presupposition for carrying out voice reason to name information
Solution, when implementing, can go to judge wait divide by NLP (Natural Language Processing, natural language processing)
The attribute information of class product, attribute information refer to the essential attribute for the article that the name information of product to be sorted is characterized, such as:
Although potato and potato title are different, potato and potato are the nicknames of potato, so, when the name of product to be sorted
When information is referred to as " potato ", system is analyzed to obtain by speech understanding to the name information, and " potato " actually referred to is horse
Bell potato, that is, the attribute information for generating product to be sorted is " potato ".
S1220, it is searched in preset product database according to the matching strategy with the attribute information similarity most
High and similarity is higher than the target product information of preset matching threshold value.
After the attribute information for obtaining product to be sorted, system is searched in product database and the category according to matching strategy
The target product of property information matches, and the target product information of target product is obtained, in one embodiment, with product to be sorted
For tomato, tomato is using mature succulence berry as the herbaceous plant of product in Solanaceae tomato genus, and system obtains western red
The name information " tomato " of persimmon, then obtains the attribute information of tomato by semantic understanding, what which was characterized
It is tomato, system finds the target product to match with the attribute information according to matching strategy in product database, such as
It includes eggplant, persimmon and tomato that target product is found in product database, wherein the similarity of tomato and the attribute information
Highest and be higher than preset matching threshold value, matching threshold is the comparison threshold value of systemic presupposition, such as the numerical value of matching threshold be 92%,
95% or 96%, then using tomato as the target product to match with product to be sorted, and then according to the tag along sort of tomato
Information is arranged the tag along sort of product to be sorted (tomato), so that tomato is belonged into category of employment described in tomato,
So that tomato is matched with tomato mapping.It should be pointed out that the specific value of matching threshold is not limited to above-mentioned numerical value, root
According to different application scenarios, other numerical value can also be arranged in matching threshold.
In another alternative embodiment, referring to Fig. 3, Fig. 3 is that one embodiment of the invention will be set contingency table
The basic procedure schematic diagram that the product to be sorted of label is stored.
As shown in figure 3, further including such as following step after step S1300:
S1400, by the tag along sort of the name information of the product to be sorted and the affiliated category of employment of the target product
Information is combined the product information for generating the product to be sorted；
After according to the tag along sort information of the affiliated category of employment of target product, the tag along sort of product to be sorted is set, it is
System believes the product of the name information of the product to be sorted and the tag along sort information combination producing product to be sorted of target product
Breath, specifically, the tag along sort of the name information of product to be sorted and product to be sorted can be associated to obtain to be sorted
The product information of product, so that the name information of product to be sorted maps the tag along sort of product to be sorted.
S1500, the product information of the product to be sorted is stored into the product database.
After the product information for generating product to be sorted, system stores the product information into product database, thus
Product to be sorted is converted into existing or known product by Unknown Product, thus the content of abundant product database, another party
Face can effectively avoid the later period again or repeat to treat the operation of sort product setting tag along sort.
In one alternate embodiment, referring to Fig. 4, Fig. 4 is that one embodiment of the invention obtains tag along sort information
Basic procedure schematic diagram.
As shown in figure 4, step S1200 includes such as following step:
S1230, polymerization generation cluster is carried out to the name information and target product information according to preset clustering rule
Information；
After the name information for obtaining product to be sorted, system matches according to matching strategy lookup with the name information
Then the target product information of target product carries out cluster generation to name information and target product information further according to clustering rule
Clustering information, wherein clustering rule be systemic presupposition for extract the common trait of name information and target product information with
Clustering information is obtained, certainly, polymeric rule can also be for carrying out polymerization summary according to name information and target product information
Clustering information is generated, which includes target producing line information mapped category of employment, i.e., poly-
Category information mapped category of employment is the upper level category of employment of target producing line information mapped category of employment, specifically,
By taking product to be sorted is electric blanket as an example, the name information of the product to be sorted is " electric blanket ", and system is looked into according to matching strategy
The target product information representation for looking for the target product to match with the name information is blanket, and system is according to clustering rule to the electricity
The clustering information that hot blanket and blanket carry out polymerization generation is bedding.
S1240, the industry label of the clustering information mapped category of employment is obtained as the tag along sort information.
After obtaining clustering information, system is using the industry label of the clustering information mapped category of employment as contingency table
Information is signed, and then the industry label of product to be sorted is set according to the tag along sort information, in one embodiment, with to be sorted
Product for for air-conditioning quilt, the name information of the product to be sorted is " air-conditioning quilt ", search and the name according to matching strategy by system
The target product information representation for the target product that information matches is referred to as quilt, system according to clustering rule to the air-conditioning quilt and by
The clustering information that son carries out polymerization generation is bedding, and the industry label of the clustering information mapped category of employment is on bed
Articles, the industry label that air-conditioning quilt is arranged in system is bedding, so that air-conditioning quilt is matched with bedding mapping.
In one alternate embodiment, the mesh to match with product to be sorted can also be searched by way of picture match
Product is marked, referring to Fig. 5, Fig. 5 is the basic procedure schematic diagram that one embodiment of the invention obtains target product information.
As shown in figure 5, step S1200 includes such as following step:
S1250, the product image for obtaining the product to be sorted；
Product image refers to the picture of product to be sorted, when implementing, is applied to product of the present invention classification control method
For on user terminal, user terminal include but is not limited to smart phone, plate and it is other have take pictures or camera function
Electronic equipment, system can acquire the product image of product to be sorted by the camera on user terminal.When implementing, with
For the product image for obtaining product to be sorted by way of shooting the video, system passes through the camera that is set on user terminal
It treats sort product to be shot to obtain target video, system can regard target by Video processing software (such as OpenCV)
Frequency is handled, and target video is split as several frame pictures, extracts picture figure from target video by timing acquiring mode
Picture.Such as a Target Photo was extracted in target video with 0.5 second one speed, then in obtained several Target Photos
In randomly select product image of the Target Photo as product to be sorted again；But it is not limited to this, according to specifically answering
With the difference of scene, the speed for acquiring picture image is able to carry out the adjustment of adaptability, and Adjustment principle is, system processing capacity
The stronger and tracking more high then acquisition time of accuracy requirement is shorter, is when reaching the Frequency Synchronization with picture pick-up device acquisition image
Only；Otherwise, then acquisition time interval is longer, but longest acquisition time interval must not exceed 1s.It is of course also possible to directly in target
Product image of the picture as product to be sorted is randomly selected in several frame pictures of video.
S1260, the target that the target product to match with the product image is searched according to the product image produce
Product information.
After the product image for obtaining product to be sorted, system can search and product phase to be sorted according to the product image
Matched target product, and then the target product information of the target product is obtained, when implementing, Fig. 6 please be participate in, Fig. 6 is this hair
The basic procedure schematic diagram of target product information is obtained in bright one embodiment according to image.
As shown in fig. 6, step S1260 further includes such as following step:
S1261, the product image is input in preset product identification model, the product identification model is trained
To convergent convolutional neural networks model；
Product identification model is the pre-set product to be sorted for identification of system and searches and product phase to be sorted
The tool for the target product matched, when implementing, LSTM network can be used, and (shot and long term remembers artificial nerve network model, Long
Short-Term Memory) it is used as neural network model.LSTM network is controlled discarding by " door " (gate) or increases letter
Breath, to realize the function of forgetting or memory." door " is a kind of structure for passing through header length, by a sigmoid (S
Sigmoid growth curve) function and dot product operation composition.The output valve of sigmoid function is lost completely in [0,1] section, 0 representative
It abandons, 1 representative passes through completely.Product to be sorted can be identified and match target product by training to have to convergent neural network model
Classifier, wherein product identification model includes above-mentioned neural network model, which includes N+1 knowledge
Other classifier, N are positive integer.
S1262, the target product information for obtaining the target product that the product identification model exports.
After the product image of product to be sorted is input in product identification model, product identification model can be according to the production
Product image searches the target product to match for product to be sorted, and the target product information of target product is exported,
Specifically, by the way that product image to be input in preset product identification model, product image is obtained in recognition classifier
Classification results, wherein classification results include corresponding multiple (such as 2, the 3 or 5) matching products of product image and
Confidence level (Confidence) with product.Wherein, the confidence level of matching product refers to product image by product identification model
After carrying out sifting sort, product image is classified into more than one matching product and obtains product to be sorted and matching production
The matched percentage value of condition.Due to finally obtaining the corresponding target product of product image as one kind, therefore need identical product figure
The confidence level of each matching product of picture is compared, for example, product to be sorted is " potato ", is classified into the confidence of potato
Degree is 0.95, and the confidence level for being classified into taro is 0.75.When the confidence level is greater than preset first threshold value, set described in confirmation
The matching product that reliability is characterized is the target product.
Preset first threshold value is traditionally arranged to be the numerical value between 0.9 to 1.It is greater than default first threshold by filtering out confidence level
The matching product of value is as final target product.For example, when preset first threshold value is 0.9, and product " egg to be sorted
Fruit ", the confidence level for being classified into passion fruit is 0.95, and the confidence level for being classified into egg is 0.56, due to 0.95 > 0.9, so
The target product to match with " passion fruit " is passion fruit, and system searches the product information conduct of passion fruit in product database
Target product information is exported.
In order to solve the above technical problems, the embodiment of the present invention also provides a kind of product classification control device.
It is the present embodiment product classification control device basic structure schematic diagram referring specifically to Fig. 7, Fig. 7.
As shown in fig. 7, a kind of product classification control device, comprising: first obtains module 2100, first processing module 2200
With the first execution module 2300, wherein the first acquisition module 2100 is used to obtain the name information of product to be sorted；First processing
The target product that module 2200 is used to search the target product to match with the name information according to preset matching strategy is believed
Breath, wherein the target product information includes the tag along sort information of the affiliated category of employment of the target product；First executes mould
Block 2300 is used to be arranged according to the tag along sort information tag along sort of the product to be sorted, so that the product to be sorted
It is matched with the affiliated category of employment mapping of the target product.
The present embodiment by searching the target product to match with product to be sorted according to the name information of product to be sorted,
And the target product information of target product is obtained, and the tag along sort information including the affiliated category of employment of the target product, then root
The tag along sort of product to be sorted is set according to the tag along sort information, so that product to be sorted and the affiliated industry of target product
Classification mapping matching, i.e., by category of employment belonging to product assortment to be sorted to target product, can according to existing product and
Corresponding relationship between industry improves to derive industry label belonging to product to be sorted and be mapped to production to be sorted
Product carry out the accuracy that labels of classification, and full-automatic classification labels mode, can effectively save the work manually to label
Amount, treatment effeciency are high.
In some embodiments, product classification control device further include: the first processing submodule and first executes submodule
Block, wherein the first processing submodule is used to carry out analysis to the name information according to preset semantic understanding rule to generate institute
State the attribute information of product to be sorted；First implementation sub-module is used for according to the matching strategy in preset product database
It searches with the attribute information similarity highest and similarity is higher than the target product information of preset matching threshold value.
In some embodiments, product classification control device further include: the second execution module and memory module, wherein
Second execution module is used for the contingency table of the name information of the product to be sorted and the affiliated category of employment of the target product
Label information is combined the product information for generating the product to be sorted；Memory module is used for the product of the product to be sorted
Information is stored into the product database.
In some embodiments, product classification control device further include: second processing submodule and first obtains submodule
Block, wherein second processing submodule is used to carry out the name information and target product information according to preset clustering rule
Polymerization generates clustering information；First acquisition submodule is used to obtain the industry label of the clustering information mapped category of employment
As the tag along sort information.
In some embodiments, product classification control device further include: the second acquisition submodule and second executes submodule
Block, wherein the second acquisition submodule is used to obtain the product image of the product to be sorted；Second implementation sub-module is used for basis
The product image searches the target product information of the target product to match with the product image.
In some embodiments, product classification control device further include: third implementation sub-module and third obtain submodule
Block, wherein third implementation sub-module for the product image to be input in preset product identification model, know by the product
Other model is trained to convergent convolutional neural networks model；Third acquisition submodule is defeated for obtaining the product identification model
The target product information of the target product out.
About the device in above-described embodiment, wherein modules execute the concrete mode of operation in related this method
Embodiment in be described in detail, no detailed explanation will be given here.
In order to solve the above technical problems, the embodiment of the present invention also provides computer equipment.It is this referring specifically to Fig. 8, Fig. 8
Embodiment computer equipment basic structure block diagram.
As shown in figure 8, the schematic diagram of internal structure of computer equipment.As shown in figure 8, the computer equipment includes passing through to be
Processor, non-volatile memory medium, memory and the network interface of bus of uniting connection.Wherein, the computer equipment is non-easy
The property lost storage medium is stored with operating system, database and computer-readable instruction, can be stored with control information sequence in database
Column, when which is executed by processor, may make processor to realize a kind of product classification control method.The calculating
The processor of machine equipment supports the operation of entire computer equipment for providing calculating and control ability.The computer equipment
It can be stored with computer-readable instruction in memory, when which is executed by processor, processor may make to hold
A kind of product classification control method of row.The network interface of the computer equipment is used for and terminal connection communication.Those skilled in the art
Member is it is appreciated that structure shown in figure, and only the block diagram of part-structure relevant to application scheme, is not constituted to this
The restriction for the computer equipment that application scheme is applied thereon, specific computer equipment may include more than as shown in the figure
Or less component, perhaps combine certain components or with different component layouts.
Processor obtains module 2100, first processing module 2200 and for executing in Fig. 7 first in present embodiment
One execution module 2300, program code and Various types of data needed for memory is stored with the above-mentioned module of execution.Network interface is used for
To the data transmission between user terminal or server.Memory in present embodiment is stored in product classification control device
Program code needed for executing all submodules and data, server is capable of the program code of invoking server and data execute institute
There is the function of submodule.
Computer by searching the target product to match with product to be sorted according to the name information of product to be sorted, and
Obtain the target product information of target product, the tag along sort information including the affiliated category of employment of the target product, then basis
The tag along sort of product to be sorted is arranged in the tag along sort information, so that product to be sorted and the affiliated industry class of target product
It Ying She not match, i.e., it, can be according to existing product and row by category of employment belonging to product assortment to be sorted to target product
Corresponding relationship between industry, to derive industry label belonging to product to be sorted and be mapped, sort product is treated in raising
Carrying out the accuracy that labels of classification, full-automatic classification labels mode, the workload manually to label can be effectively saved,
Treatment effeciency is high.
The present invention also provides a kind of storage mediums for being stored with computer-readable instruction, and the computer-readable instruction is by one
When a or multiple processors execute, so that one or more processors execute product classification controlling party described in any of the above-described embodiment
The step of method.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, which can be stored in computer-readable storage medium
In, the program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, storage medium above-mentioned can be magnetic
The non-volatile memory mediums such as dish, CD, read-only memory (Read-Only Memory, ROM) or random storage memory
Body (Random Access Memory, RAM) etc..
It should be understood that although each step in the flow chart of attached drawing is successively shown according to the instruction of arrow,
These steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly stating otherwise herein, these steps
Execution there is no stringent sequences to limit, can execute in the other order.Moreover, at least one in the flow chart of attached drawing
Part steps may include that perhaps these sub-steps of multiple stages or stage are not necessarily in synchronization to multiple sub-steps
Completion is executed, but can be executed at different times, execution sequence, which is also not necessarily, successively to be carried out, but can be with other
At least part of the sub-step or stage of step or other steps executes in turn or alternately.
The above is only some embodiments of the invention, it is noted that for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (10)
1. a kind of product classification control method, which is characterized in that include the following steps:
Obtain the name information of product to be sorted；
The target product information of the target product to match with the name information is searched according to preset matching strategy, wherein
The target product information includes the tag along sort information of the affiliated category of employment of the target product；
The tag along sort of the product to be sorted is set according to the tag along sort information so that the product to be sorted with it is described
The affiliated category of employment mapping matching of target product.
2. product classification control method according to claim 1, which is characterized in that described to be looked into according to preset matching strategy
The step of looking for the target product information of the target product to match with the name information, including such as following step:
The attribute information that analysis generates the product to be sorted is carried out to the name information according to preset semantic understanding rule；
It is searched in preset product database according to the matching strategy and the attribute information similarity highest and similarity
Higher than the target product information of preset matching threshold value.
3. product classification control method according to claim 2, which is characterized in that described according to the tag along sort information
The tag along sort of the product to be sorted is set, so that the product to be sorted and the affiliated category of employment of the target product map
Further include such as following step after the step of matching:
The tag along sort information of the name information of the product to be sorted and the affiliated category of employment of the target product is subjected to group
Symphysis at the product to be sorted product information；
The product information of the product to be sorted is stored into the product database.
4. product classification control method according to claim 1, which is characterized in that described to be looked into according to preset matching strategy
The step of looking for the target product information of the target product to match with the name information, including such as following step:
Polymerization is carried out to the name information and target product information according to preset clustering rule and generates clustering information；
The industry label of the clustering information mapped category of employment is obtained as the tag along sort information.
5. product classification control method according to claim 1, which is characterized in that described to be looked into according to preset matching strategy
The step of looking for the target product information of the target product to match with the name information, including such as following step:
Obtain the product image of the product to be sorted；
The target product information of the target product to match with the product image is searched according to the product image.
6. product classification control method according to claim 5, which is characterized in that described to be searched according to the product image
The step of target product information of the target product to match with the product image, including such as following step:
The product image is input in preset product identification model, the product identification model is trained to convergent volume
Product neural network model；
Obtain the target product information of the target product of the product identification model output.
7. a kind of product classification control device characterized by comprising
First obtains module, for obtaining the name information of product to be sorted；
First processing module, for searching the mesh of the target product to match with the name information according to preset matching strategy
Mark product information, wherein the target product information includes the tag along sort information of the affiliated category of employment of the target product；
First execution module, for the tag along sort of the product to be sorted to be arranged according to the tag along sort information, so that institute
Product to be sorted is stated to match with the affiliated category of employment mapping of the target product.
8. product classification control device according to claim 7, which is characterized in that further include:
First processing submodule, for according to preset semantic understanding rule to the name information carry out analysis generate it is described to
The attribute information of sort product；
First implementation sub-module, for being searched in preset product database according to the matching strategy and the attribute information
Similarity highest and similarity are higher than the target product information of preset matching threshold value.
9. a kind of computer equipment, including memory and processor, it is stored with computer-readable instruction in the memory, it is described
When computer-readable instruction is executed by the processor, so that the processor executes such as any one of claims 1 to 6 right
It is required that the step of product classification control method.
10. a kind of storage medium for being stored with computer-readable instruction, the computer-readable instruction is handled by one or more
When device executes, so that one or more processors execute the product classification control as described in any one of claims 1 to 6 claim
The step of method processed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811577748.2A CN109800769B (en) | 2018-12-20 | 2018-12-20 | Product classification control method, device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811577748.2A CN109800769B (en) | 2018-12-20 | 2018-12-20 | Product classification control method, device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109800769A true CN109800769A (en) | 2019-05-24 |
| CN109800769B CN109800769B (en) | 2024-08-02 |
Family
ID=66557329
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811577748.2A Active CN109800769B (en) | 2018-12-20 | 2018-12-20 | Product classification control method, device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109800769B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110825773A (en) * | 2019-10-31 | 2020-02-21 | 北京东软望海科技有限公司 | Information verification method and device, computer equipment and storage medium |
| CN111626808A (en) * | 2020-02-26 | 2020-09-04 | 京东数字科技控股有限公司 | Data processing method and apparatus, storage medium, and electronic apparatus |
| CN111985901A (en) * | 2020-08-24 | 2020-11-24 | 北京思特奇信息技术股份有限公司 | Marketing product configuration method, device, equipment and storage medium in telecommunication industry |
| CN112052910A (en) * | 2020-09-21 | 2020-12-08 | 深圳海关动植物检验检疫技术中心 | Food safety classification method and device, computer equipment and storage medium |
| CN116089614A (en) * | 2023-01-12 | 2023-05-09 | 杭州瓴羊智能服务有限公司 | Text marking method and device |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105243389A (en) * | 2015-09-28 | 2016-01-13 | 北京橙鑫数据科技有限公司 | Industry classification tag determining method and apparatus for company name |
| CN107436875A (en) * | 2016-05-25 | 2017-12-05 | 华为技术有限公司 | File classification method and device |
| CN107742128A (en) * | 2017-10-20 | 2018-02-27 | 百度在线网络技术（北京）有限公司 | Method and apparatus for output information |
| US20180121533A1 (en) * | 2016-10-31 | 2018-05-03 | Wal-Mart Stores, Inc. | Systems, method, and non-transitory computer-readable storage media for multi-modal product classification |
| CN108280124A (en) * | 2017-12-11 | 2018-07-13 | 北京三快在线科技有限公司 | Product classification method and device, ranking list generation method and device, electronic equipment |
| CN108563722A (en) * | 2018-04-03 | 2018-09-21 | 有米科技股份有限公司 | Trade classification method, system, computer equipment and the storage medium of text message |
| CN108595418A (en) * | 2018-04-03 | 2018-09-28 | 上海透云物联网科技有限公司 | A kind of commodity classification method and system |
| CN108921206A (en) * | 2018-06-15 | 2018-11-30 | 北京金山云网络技术有限公司 | A kind of image classification method, device, electronic equipment and storage medium |
-
2018
- 2018-12-20 CN CN201811577748.2A patent/CN109800769B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105243389A (en) * | 2015-09-28 | 2016-01-13 | 北京橙鑫数据科技有限公司 | Industry classification tag determining method and apparatus for company name |
| CN107436875A (en) * | 2016-05-25 | 2017-12-05 | 华为技术有限公司 | File classification method and device |
| US20180121533A1 (en) * | 2016-10-31 | 2018-05-03 | Wal-Mart Stores, Inc. | Systems, method, and non-transitory computer-readable storage media for multi-modal product classification |
| CN107742128A (en) * | 2017-10-20 | 2018-02-27 | 百度在线网络技术（北京）有限公司 | Method and apparatus for output information |
| CN108280124A (en) * | 2017-12-11 | 2018-07-13 | 北京三快在线科技有限公司 | Product classification method and device, ranking list generation method and device, electronic equipment |
| CN108563722A (en) * | 2018-04-03 | 2018-09-21 | 有米科技股份有限公司 | Trade classification method, system, computer equipment and the storage medium of text message |
| CN108595418A (en) * | 2018-04-03 | 2018-09-28 | 上海透云物联网科技有限公司 | A kind of commodity classification method and system |
| CN108921206A (en) * | 2018-06-15 | 2018-11-30 | 北京金山云网络技术有限公司 | A kind of image classification method, device, electronic equipment and storage medium |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110825773A (en) * | 2019-10-31 | 2020-02-21 | 北京东软望海科技有限公司 | Information verification method and device, computer equipment and storage medium |
| CN110825773B (en) * | 2019-10-31 | 2022-09-27 | 望海康信(北京)科技股份公司 | Information verification method and device, computer equipment and storage medium |
| CN111626808A (en) * | 2020-02-26 | 2020-09-04 | 京东数字科技控股有限公司 | Data processing method and apparatus, storage medium, and electronic apparatus |
| CN111985901A (en) * | 2020-08-24 | 2020-11-24 | 北京思特奇信息技术股份有限公司 | Marketing product configuration method, device, equipment and storage medium in telecommunication industry |
| CN111985901B (en) * | 2020-08-24 | 2024-02-02 | 北京思特奇信息技术股份有限公司 | Marketing product configuration method, device, equipment and storage medium in telecom industry |
| CN112052910A (en) * | 2020-09-21 | 2020-12-08 | 深圳海关动植物检验检疫技术中心 | Food safety classification method and device, computer equipment and storage medium |
| CN112052910B (en) * | 2020-09-21 | 2024-05-10 | 深圳海关动植物检验检疫技术中心 | Food safety classification method, device, computer equipment and storage medium |
| CN116089614A (en) * | 2023-01-12 | 2023-05-09 | 杭州瓴羊智能服务有限公司 | Text marking method and device |
| CN116089614B (en) * | 2023-01-12 | 2023-11-21 | 瓴羊智能科技有限公司 | Text marking method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109800769B (en) | 2024-08-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109800769A (en) | Product classification control method, device, computer equipment and storage medium | |
| CN111368079B (en) | Text classification method, model training method, device and storage medium | |
| CN109215754A (en) | Medical record data processing method, device, computer equipment and storage medium | |
| CN109472024A (en) | A kind of file classification method based on bidirectional circulating attention neural network | |
| CN112257449B (en) | Named entity recognition method and device, computer equipment and storage medium | |
| CN109829478B (en) | Problem classification method and device based on variation self-encoder | |
| Febrian et al. | Facial expression recognition using bidirectional LSTM-CNN | |
| CN110705490B (en) | Visual emotion recognition method | |
| CN113220865B (en) | Text similar vocabulary retrieval method, system, medium and electronic equipment | |
| CN110245228A (en) | The method and apparatus for determining text categories | |
| CN112732872B (en) | Biomedical text-oriented multi-label classification method based on subject attention mechanism | |
| Liu et al. | Compact feature learning for multi-domain image classification | |
| US11887216B2 (en) | High resolution conditional face generation | |
| CN116091667B (en) | Character artistic image generation system based on AIGC technology | |
| Rao et al. | Facial expression recognition with multiscale graph convolutional networks | |
| CN110909174B (en) | Knowledge graph-based method for improving entity link in simple question answering | |
| Jayanthi et al. | Comparison of convolutional neural networks architectures for mango leaf classification | |
| CN109074363A (en) | Data query method, data query system determine method and apparatus | |
| Gonzalez-Soler et al. | Semi-synthetic data generation for tattoo segmentation | |
| Tiwari et al. | Learning semantic image attributes using image recognition and knowledge graph embeddings | |
| US20220100792A1 (en) | Method, device, and program for retrieving image data by using deep learning algorithm | |
| Choudhary et al. | Activity recognition system via unification of CNN and SVM in complex domain | |
| Richter et al. | Facial expression classification on web images | |
| CN111062207B (en) | Expression image processing method and device, computer storage medium and electronic equipment | |
| CN113343773A (en) | Facial expression recognition system based on shallow convolutional neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |