WO2022161231A1 - 对象识别码处理方法、对象发布方法、装置、设备及介质 - Google Patents

对象识别码处理方法、对象发布方法、装置、设备及介质 Download PDF

Info

Publication number
WO2022161231A1
WO2022161231A1 PCT/CN2022/072842 CN2022072842W WO2022161231A1 WO 2022161231 A1 WO2022161231 A1 WO 2022161231A1 CN 2022072842 W CN2022072842 W CN 2022072842W WO 2022161231 A1 WO2022161231 A1 WO 2022161231A1
Authority
WO
WIPO (PCT)
Prior art keywords
identification code
information
target market
category
library
Prior art date
Application number
PCT/CN2022/072842
Other languages
English (en)
French (fr)
Inventor
黄志文
黄宇超
Original Assignee
阿里巴巴集团控股有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 阿里巴巴集团控股有限公司 filed Critical 阿里巴巴集团控股有限公司
Publication of WO2022161231A1 publication Critical patent/WO2022161231A1/zh

Links

Images

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]
    • G06Q30/0641—Shopping interfaces
    • G06Q30/0643—Graphical representation of items or shoppers
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/279—Recognition of textual entities
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/40—Document-oriented image-based pattern recognition
    • G06V30/41—Analysis of document content
    • G06V30/413—Classification of content, e.g. text, photographs or tables
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition

Definitions

  • the embodiments of the present disclosure relate to the technical field of data processing, and in particular, to a method for processing an object identification code, a method for publishing an object, an apparatus, a device, and a medium.
  • the traditional way of publishing objects is: the merchant manually fills in all the object information one by one on the object publishing page, and clicks the publish button on the object publishing page to publish the object.
  • the traditional method of object publishing requires more time, the efficiency of object publishing is low, and the object publishing may fail due to improper filling.
  • the inventors of the present disclosure also found that the object information that the merchant fills in the object publishing page may not conform to the target market's definition of object information. If the object information is directly filled in the object publishing page, the object publishing may fail.
  • At least one embodiment of the present disclosure provides an object identification code processing method, an object publishing method, an apparatus, an electronic device, and a storage medium.
  • an embodiment of the present disclosure provides an object identification code processing method, including:
  • the association relationship between the object identification code, the aggregated information, and the second object information is stored.
  • the first object information includes an identifier header
  • the aggregating the object identification code with at least a part of identification codes in the identification code library of the target market based on the first object information includes:
  • the identification code title is truncated multiple times to obtain a plurality of subtitles
  • the object identifiers are aggregated with at least a portion of the identifiers in the target market's library of identifiers based on the plurality of subtitles.
  • the aggregating the object identifiers with at least a portion of the identifiers in the identifier library of the target market based on the plurality of subtitles includes:
  • the object identifiers are aggregated with identifiers corresponding to the matched key attributes.
  • the matching each of the subtitles with a key attribute of at least a portion of the identifiers in the identifier library of the target market comprises:
  • Each of the subtitles is matched with key attributes common to each of the plurality of aggregated sets.
  • the first object information includes a first identifier header
  • the aggregating the object identification code with at least a part of identification codes in the identification code library of the target market based on the first object information includes:
  • Similar distance aggregation is performed between the object identification code and at least a part of the identification code in the identification code library of the target market.
  • the determining a similarity distance between the second identification code title and the first identification code title comprises:
  • the second identification code title and the first identification code title are input into an entity alignment model, and the entity alignment model is used for alignment, and the similarity distance between the second identification code title and the first identification code title is output.
  • the entity alignment model is a model obtained by training a Siamese neural network based on the identification code title pair of the same object pair.
  • the first object information includes a first object category and a first object attribute; the mapping of the first object information to the second object information of the target market includes:
  • the first object attribute is mapped to the second object attribute of the target market.
  • the mapping of the first object category to the second object category of the target market includes:
  • the first object information further includes an object picture
  • the mapping of the first object category to the second object category of the target market includes:
  • the object picture is input into an object recognition model, the object in the object picture is recognized by the object recognition model, and the second object category of the target market is output.
  • mapping of the first object category to the second object category of the target market further includes:
  • the text information is input into the text model, the text information is analyzed by the text model, and the second object category of the target market is output.
  • the first object information further includes an identification code title
  • the mapping of the first object category to the second object category of the target market further includes:
  • the identification code title is input into the text model, the identification code title is subjected to text analysis by the text model, and the second object category of the target market is output.
  • the method further includes:
  • the second object category is revised; wherein, the first rule is a rule based on algorithm batch mining; and the second rule is a rule accumulated based on industry experts.
  • the mapping of the first object category to the second object category of the target market includes:
  • the first object category is mapped to the second object category of the target market based on the category mapping relationship between the artificially deposited channels.
  • the method further includes:
  • the second object category is revised.
  • an embodiment of the present disclosure also provides an object publishing method, including:
  • the object information in the object publishing page is published.
  • the method further includes:
  • the object identification code is processed based on the object identification code processing method provided by any embodiment of the first aspect, and the object identification code associated with the object identification code is obtained. Aggregate information and second object information.
  • an embodiment of the present disclosure further provides an object identification code processing device, including:
  • an acquisition unit for acquiring an object identification code and first object information associated with the object identification code
  • an aggregation unit configured to aggregate the object identification code and at least a part of the identification codes in the identification code library of the target market based on the first object information to obtain aggregated information
  • mapping unit configured to map the first object information to the second object information of the target market
  • the storage unit is configured to store the association relationship among the object identification code, the aggregated information, and the second object information in the identification code library of the target market.
  • an embodiment of the present disclosure further provides an object publishing apparatus, including:
  • the display unit is used to respond to the object publishing request and display the object publishing page
  • a search unit configured to search for the aggregated information and second object information associated with the object identification code from the identification code library of the target market based on the object identification code;
  • a backfilling unit configured to backfill the object publishing page with the second object information, and backfill the object information corresponding to at least a part of the identification code corresponding to the aggregated information in the object publishing page;
  • the publishing unit is configured to publish the object information in the object publishing page in response to the information confirmation operation.
  • an embodiment of the present disclosure further provides an electronic device, including: a processor and a memory; the processor is configured to execute the process described in any embodiment of the first aspect by invoking a program or an instruction stored in the memory. steps of the method.
  • an embodiment of the present disclosure further provides a non-transitory computer-readable storage medium for storing a program or an instruction, where the program or instruction causes a computer to execute the steps of the method according to any embodiment of the first aspect.
  • the object identification code is aggregated with at least a part of the identification codes in the identification code library of the target market through the object information of the original market associated with the object identification code, so that when an object is released, it can be
  • One object identification code is used to release multiple aggregated objects, and the object identification code is used to assist merchants to quickly release objects and improve the efficiency of object release; in addition, by mapping the object information of the original market associated with the object identification code to the object information of the target market, Then, the object identification code, aggregation information, and the association relationship between the object identification code and the object information of the target market are stored in the identification code library of the target market, so that when an object is released, the object can be found from the identification code library of the target market.
  • the object information of the identification code in the target market ensures the successful release of the object, and the object release will not fail because the object information of the object identification code in the original market cannot be directly applied to the target market.
  • FIG. 1 is an exemplary application scenario diagram provided by an embodiment of the present disclosure
  • FIG. 2 is a block diagram of an object identification code processing apparatus provided by an embodiment of the present disclosure
  • FIG. 3 is a flowchart of an object identification code processing method provided by an embodiment of the present disclosure.
  • FIG. 4 is a block diagram of an object publishing apparatus provided by an embodiment of the present disclosure.
  • FIG. 5 is a flowchart of an object publishing method provided by an embodiment of the present disclosure.
  • FIG. 6 is a block diagram of an electronic device provided by an embodiment of the present disclosure.
  • E-commerce uses the network to conduct transaction activities and related service activities in the form of electronic transactions, making all aspects of traditional business activities electronic and networked.
  • SKU Stock Keeping Unit
  • SPU Standard Product Unit
  • SKU is the unit of measurement of inventory in and out (buyers purchase, merchants purchase, supplier stocking, factory production are all based on SKU), and it is the most commonly used in clothing, footwear, daily chemicals and other commodities.
  • a SKU in textiles usually means: size, color, style.
  • SKU is the smallest unit of inventory that is physically inseparable, that is to say, for a commodity, the specific inventory of goods can be determined according to the SKU.
  • SPU is the smallest unit of product information aggregation. It is a set of reusable and easy-to-retrieve standardized information. The set describes the characteristics of a product. In short, a product can correspond to one SPU.
  • the characteristics of the product can be described by multiple "attribute
  • attribute value pair” is exactly the same commodity, which can be abstracted into an SPU.
  • Both SPU and SKU are sets of attribute values.
  • the SPU attribute is an attribute that does not affect inventory and price, and is also called a key attribute; while the SKU attribute is an attribute that affects inventory and price, so it can also be called as Sale properties.
  • One SPU corresponds to multiple SKUs, and the difference between these SKUs lies in the difference in sales attributes (such as colors, specifications, etc.).
  • the e-commerce service system When the e-commerce service system displays product information to the user on a specific page, it usually first displays the product information in units of SPU, and then, when the user clicks to enter the SPU details page, or the e-commerce service system receives After the "add to shopping cart” or "buy now” request, the e-commerce service system displays the information of the specific optional SKU list.
  • the merchant can publish a new object (item), so that the user can browse the information of the new object published by the merchant on the merchant's store page.
  • the objects released usually include two categories: first, physical goods, such as clothing, electrical appliances, wear and other tangible goods, usually the functions provided by Taobao, Pinduoduo, Jingdong and other e-commerce service systems; second, virtual goods, such as services category of goods, including takeaways, videos, music, content and other intangible goods. Whether it is a physical commodity or a virtual commodity, it is the product of labor produced for sale and the product of labor for exchange. Therefore, the published object can be understood as a product for sale.
  • the inventors of the present disclosure also found that the object information that the merchant fills in the object publishing page may not conform to the target market's definition of object information. If the object information is directly filled in the object publishing page, the object publishing may fail.
  • the embodiments of the present disclosure provide an object identification code processing method, an object publishing method, an apparatus, a device, and a medium.
  • the object identification code processing function and the object rapid release function provided by the embodiments of the present disclosure are embedded in the third-party platform.
  • the object identification code processing function and the object rapid release function provided by the embodiments of the present disclosure can be implemented as commercial services, and the merchant's self-built e-commerce website can also be used by calling the interface. Services provided by embodiments of the present disclosure.
  • the object identification code processing function and the object rapid release function provided by the embodiments of the present disclosure can be applied to application scenarios of cross-market or cross-channel distribution.
  • RT-Mart products can be quickly released to Tmall Supermarket saves time and cost for merchants to publish products.
  • a function of quickly publishing objects is provided. Specifically, by using the object information of the original market associated with the object identification code, the object identification code and at least a part of the identification code in the identification code library of the target market are combined.
  • Aggregation makes it easy to publish multiple aggregated objects based on one object identification code when publishing objects, and use the object identification code to assist merchants to quickly publish objects and improve the efficiency of object publishing;
  • the object information of the target market is mapped to the object information of the target market, and then the association relationship between the object identification code, the aggregation information, and the object identification code in the object information of the target market is stored in the identification code library of the target market, so that when the object is released, it can be easily Find the object information of the object identification code in the target market from the identification code database of the target market to ensure that the object is successfully released, and the object release will not fail because the object information of the object identification code in the original market cannot be directly applied to the target market.
  • FIG. 1 is an exemplary application scenario diagram provided by an embodiment of the present disclosure, and does not represent all scenarios.
  • a merchant can initiate an object publishing request through various client devices 101 , for example, the merchant inputs the URL of the object publishing page in the browser of the client device 101 and confirms the access URL to initiate an object publishing request.
  • the browser of the client device 101 may obtain resources for displaying the object publishing page from the server device 103 through the network 102, and then render the object publishing page by using the obtained resources.
  • the merchant can then enter the object identification code on the object publishing page, and the merchant can also scan the object identification code with a scanning device such as a code scanner or a smartphone, and the scanning device automatically fills the object identification code into the object publishing page without manual input.
  • the object identification code may be a representation identification code of multi-field attribute information of objects that can be quickly identified, such as barcodes, two-dimensional codes, and three-dimensional codes.
  • the client device 101 can send the object identification code to the server device 103, and then can obtain the object information associated with the object identification code sent by the server device 103.
  • the object information includes, for example, the identification code title, object category, object attributes and One or more of the object pictures, in which the object category can be understood as the classification of the object; the object attributes are represented in the form of key-value pairs (attribute item-attribute value), for example, the attribute item is color, and the attribute value is red.
  • the client device 101 may backfill the object information associated with the object identifier to the object publishing page. Merchants can modify the incorrect information in the backfill information; they can also click the release control on the object release page after confirming that the backfill information is correct. The client device 101 publishes the object information in the object publishing page in response to the click operation of the publishing control.
  • the client device 101 may be a mobile electronic device such as a smart phone, a notebook computer, a tablet computer, etc.; the client device 101 may also be a fixed electronic device such as a desktop computer and a smart TV.
  • the server device 103 After receiving the resource request for displaying the object publishing page, the server device 103 can send the resources required for displaying the object publishing page to the browser of the client device 101 through the network 102, so that the browser of the client device 101 can display Object publishing page.
  • the server device 103 receives the object identification code sent by the client device 101, and then searches the identification code library 104 for the object information associated with the object identification code, and sends the found object information to the client device 101, so that the client The device 101 may backfill the object information to the object publishing page.
  • the server device 103 may be understood as an e-commerce service system (hardware implementation) or a hardware device that supports the operation of the e-commerce service system (software implementation).
  • the server device 103 may be a single server or a server group, and the server group may be centralized or distributed.
  • the identification code library 104 is the identification code library of the target market, and the target market can be understood as the market for which the object is released.
  • the target market of the object information associated with an object identification code (referred to as the original market) may be different from the target market. Therefore, before an object identification code is stored, the object identification code needs to be processed by the server device 103.
  • the object identification code is associated with new object information, and the new object information is information for the target market, so that the object identification code can be stored in the warehouse.
  • the identification code library 104 can further store the association relationship between object identification codes and aggregated information.
  • the aggregated information is the information obtained by the aggregation of the object identification code and at least a part of the identification codes stored in the identification code library 104.
  • the server device 103 can combine the object identification code with the identification code library 104 of the target market. At least a part of the identification codes are aggregated to obtain aggregated information.
  • the server device 103 can also send the object information corresponding to at least a part of the identification code corresponding to the aggregated information to the client device 101, so that the client device 101 backfills the at least part of the identification code corresponding to the aggregated information in the object publishing page.
  • the client device 101 backfills the at least part of the identification code corresponding to the aggregated information in the object publishing page.
  • the identification code library 104 may not store the object identification code that the server device 103 is looking for. In this case, the server device 103 needs to process the object identification code to obtain the object identification code in the target market.
  • the object information and the aggregation information, and the association relationship between the object identification code, the object information, and the aggregation information is stored in the identification code library 104 for subsequent search and reuse.
  • FIG. 2 is an exemplary block diagram of an object identification code processing apparatus 200 according to an embodiment of the present disclosure.
  • the object identification code processing apparatus 200 may be implemented as a part of the server device 103 in FIG. 1 .
  • the object identification code processing apparatus 200 may be divided into multiple units, for example, including but not limited to: an acquisition unit 201 , an aggregation unit 202 , a mapping unit 203 and a storage unit 204 . The details of each unit are as follows:
  • the obtaining unit 201 is configured to obtain the object identification code, and may also obtain first object information associated with the object identification code.
  • the first object information can be understood as object information for the original market.
  • the object identification code and the first object information associated with the object identification code can be obtained by purchasing from a third-party company. Therefore, the obtaining unit 201 can obtain the object identification code and the first object associated with the object identification code in batches information.
  • the obtaining unit 201 may obtain the object identification code and the first object information associated with the object identification code from different channels, wherein different channels may be understood as different e-commerce platforms, such as domestic e-commerce service systems or Service system of foreign e-commerce.
  • the first object information may include: one or more of an identification code title, an object category, an object attribute, and an object picture, wherein the object category can be understood as the classification of the object in the original market; the object attribute is represented by a key-value pair ( property item - property value).
  • the object picture may be replaced with the storage address of the object picture.
  • the content included in the first object information obtained from the same channel has two definitions:
  • the object attributes of memory modules in the original market include but are not limited to:
  • the attribute item i is the brand; the attribute value of the attribute item i is one of Springfield, ADATA, American Merchant, Corsair, Samsung, GALAXY, etc.;
  • the attribute item ii is the memory capacity; the attribute value of the attribute item ii is one of 2G, 4G, 8G, 16G, etc.;
  • the attribute item iii is the transmission type; the attribute value of the attribute item iii is one of DDR, DDR2, DDR3, DDR4, etc.;
  • the attribute item iv is the memory frequency; the attribute value of the attribute item iv is one of 2333MHz, 3000MHz, 4000MHz, and 4800MHz.
  • the title of the identification code of the memory stick is: New Samsung 128G 2S4RX4 PC4-2666V-RB2 DDR4 2666V ECC REG memory stick;
  • the object attributes of the memory stick in the original market include: the attribute item is the brand, and the corresponding attribute value is Samsung; the attribute item is the memory capacity, and the corresponding attribute value is 128G; the attribute item is the transmission type, and the corresponding attribute value is DDR4; the attribute item is Condition, the corresponding attribute value is brand new; the attribute item is after-sales service, the corresponding attribute value is the store's three guarantees; the attribute item is the applicable object, and the corresponding attribute value is the server.
  • the aggregation unit 202 is configured to aggregate the object identification code and at least a part of the identification codes in the identification code library of the target market to obtain aggregated information. In some embodiments, the aggregating unit 202 aggregates the object identification code with at least a portion of the identification code in the identification code library of the target market based on the first object information.
  • the aggregating unit 202 may truncate the identification code title included in the first object information for multiple times to obtain multiple subtitles.
  • the aggregation unit 202 starts to calculate the length from the first word of the title of the identification code, and when the calculated length reaches the preset truncation length, truncation is performed to obtain a subtitle.
  • the aggregating unit 202 may truncate the identifier header multiple times, and the preset truncation lengths referenced during each truncation are different.
  • the preset truncation length may be set according to actual needs. This embodiment does not limit the specific value of the preset truncation length.
  • the preset truncation length may be 3 words, 5 words or 7 words.
  • the length is calculated from the first word of the title of the identification code, because the information aggregated by the barcode is usually contained within a length of the first word of the title of the identification code.
  • barcode 06923410719178 is "Philips HTB5570D/93 home cinema system 3D 5.1 channels 1000W", which is truncated with the preset truncation length of 5 words, and the subtitle is "Philips HTB5570D/93 home cinema system”.
  • the aggregating unit 202 may aggregate the object identifiers with at least a portion of the identifiers in the target market's identifier library based on the plurality of subtitles. In some embodiments, the aggregating unit 202 may match each subtitle to key attributes of at least a portion of the identifiers in the target market's library of identifiers. In some embodiments, the aggregating unit 202 can pre-aggregate at least a part of the identification codes in the identification code library of the target market to obtain multiple aggregation sets; and then determine the key attributes common to each aggregation set; Key attributes common to multiple aggregate sets are matched. In some embodiments, the aggregation unit 202 may aggregate the object identifiers with the identifiers corresponding to the matched key attributes.
  • the aggregating unit 202 may determine a similarity distance between the first identification code title included in the first object information and the second identification code title of at least a portion of the identification code in the target market's identification code library, respectively. In some embodiments, the aggregating unit 202 may input the first identification code title and a second identification code title into the entity alignment model, perform alignment through the entity alignment model, and output the difference between the first identification code title and the second identification code title similarity distance.
  • the entity alignment model is a model obtained by training a twin neural network based on an identification code title pair of an object pair of the same item, and an object of the same item can be understood as an object with the same key attribute. In some embodiments, the entity alignment model may be other neural network models.
  • the entity alignment model converts them into two 128-dimensional vectors, and then calculates these two The similarity distance of a 128-dimensional vector, the similarity distance can be calculated by using the equidistance formulas of the Euclidean distance and the cosine distance. These distance formulas are commonly used formulas and will not be repeated here.
  • the aggregating unit 202 may perform similarity distance aggregation between the object identifiers and at least a portion of the identifiers in the identifier library of the target market. In some embodiments, the aggregation unit 202 may perform similarity distance aggregation based on clustering algorithms such as DBSCAN algorithm, K-means algorithm, etc. These clustering algorithms are commonly used algorithms and will not be described again.
  • the aggregation unit 202 may fuse aggregation results obtained in different ways to obtain final aggregation information.
  • the mapping unit 203 is configured to map the first object information to the second object information of the target market. In some embodiments, the mapping unit 203 may map the first object category in the first object information to the second object category of the target market.
  • the mapping unit 203 will "Computer Hardware/Monitor/Computer Peripherals>>Memory" is mapped to "Computer/Computer Accessories>>Memory”.
  • the mapping unit 203 may input the first object category and the first object attribute into the text model, perform text analysis on the first object category and the first object attribute through the text model, and output the second object of the target market category.
  • the mapping unit 203 can also identify the text information in the object picture in the first object information.
  • the identification method is, for example, inputting the object picture into a picture recognition network, performing picture recognition through the picture recognition network, and outputting the text information in the object picture.
  • the text information, wherein the image recognition network can be obtained by pre-training a neural network, and the type of the neural network is not limited here.
  • the mapping unit 203 inputs the recognized text information into the text model, analyzes the text information through the text model, and outputs the second object category of the target market.
  • the mapping unit 203 may input the identification code title in the first object information into the text model, perform text analysis on the identification code title through the text model, and output the second object category of the target market.
  • the text model can be trained based on the input and output targets.
  • the mapping unit 203 may input the object picture in the first object information into the object recognition model, identify the object in the object picture through the object recognition model, and output the second object category of the target market. In some embodiments, the mapping unit 203 may obtain the object image based on the storage address of the object image in the first object information, and then input the object image into the object recognition model, identify the object in the object image through the object recognition model, and output the first image of the target market. Two object categories.
  • the object recognition model is a model obtained by pre-training a neural network, and the type of the neural network is not limited here.
  • the mapping unit 203 may fuse the second object categories obtained in different ways to obtain the final second object category, so as to avoid inaccurate or unmappable second object categories obtained by a single method. question.
  • the mapping unit 203 modifies the second object category based on the first rule and/or the second rule to improve the mapping accuracy ;
  • the first rule is a rule based on algorithm batch mining;
  • the second rule is a rule based on accumulation of industry experts.
  • the object categories of the target market are shown in Table 1 below:
  • the mapping unit 203 inputs the identification code title "8424 watermelon 3 yuan/catties" into the text model, performs text analysis on the identification code title through the text model, and outputs the second object category of the target market as "other watermelons". Based on the first rule, the mapping unit 203 can modify the second object category to be "8424 watermelon” by using the keyword "8424" appearing in the title of the identification code.
  • the second rule is based on the rules accumulated by industry experts. For example, through the accumulated experience of industry experts, under the category of liquor, if the degree is greater than 41 degrees, it is determined as the category of "high-level liquor”.
  • the mapping unit 203 inputs the identification code title "42-degree liquor” into the text model, performs text analysis on the identification code title through the text model, and outputs the second object category of the target market as "liquor".
  • the mapping unit 203 may modify the second object category to be "highly liquor” based on the second rule.
  • the mapping unit 203 may map the first object category to the second object category of the target market based on the manually precipitated category mapping relationship between channels.
  • the category mapping relationship between artificially deposited channels is shown in Table 2 below:
  • the mapping unit 203 can directly map the first object category to the second object category of the target market based on the manually deposited category mapping relationship between the channels. For example, directly map the object category of the first channel "coffee/oatmeal/brewed drinks>>beverages>>functional drinks" to the object category of the target market (for example, the third channel) "drinks>>sports energy drinks>>sports Beverages”, category mapping is faster and more efficient.
  • the mapping unit 203 after mapping the first object category to the second object category of the target market, the mapping unit 203 corrects the second object category based on the user feedback information to improve the mapping accuracy.
  • the user feedback information is: the mapping of the first object category "apple-flavored beverage” to the second object category "apple” is incorrect, and it should be mapped to "beverage”, and the mapping unit 203 may, based on the user feedback information, correct the The second object category "apple” is "beverage”.
  • mapping unit 203 may modify the training data and retrain the text model based on the user feedback information. For example, retrain the text model so that the text model can output "drink” based on the input "apple-flavored drink”.
  • the mapping unit 203 can supplement the previously missed expert experience, that is, improve the rules (ie, the second rules) based on the accumulation of industry experts. For example, user feedback needs to distinguish the origin of cherries, then the second rule can be added: if the origin of cherries is outside of China, it is considered “imported cherries”, and if it is made in China, it is considered “domestic cherries” .
  • the mapping unit 203 may map the first object attribute in the first object information to the second object attribute of the target market.
  • the attribute item of the first object attribute in the first object information is color, and the corresponding attribute value is red
  • the attribute item of the object attribute of the target market is Color, and the corresponding attribute value is Red; therefore, the mapping unit 203 will "The attribute item is color, and the corresponding attribute value is red" is mapped to "the attribute item is Color, and the corresponding attribute value is Red”.
  • the mapping unit 203 may perform attribute mapping in a manner similar to category mapping, which will not be repeated here in order to avoid repetition. In some embodiments, the mapping unit 203 may perform attribute correction in a manner similar to category correction, which will not be repeated here to avoid repetition.
  • the storage unit 204 is configured to store the association relationship among the object identification code, the aggregation information, and the second object information in the identification code library of the target market.
  • the aggregated information and the second object information corresponding to the object identification code can be searched from the identification code library later.
  • the market caused the object release to fail.
  • a plurality of aggregated objects can be released based on one object identification code, thereby improving the efficiency of object release.
  • each unit in the object identification code processing apparatus 200 is only a logical function division, and there may be other division methods in actual implementation, such as the acquisition unit 201, the aggregation unit 202, the mapping unit 203 and the storage unit
  • At least two units in 204 may be implemented as one unit; the acquiring unit 201 , the aggregation unit 202 , the mapping unit 203 or the storage unit 204 may also be divided into multiple subunits.
  • each unit or sub-unit can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art may use different methods for implementing the described functionality for each particular application.
  • FIG. 3 is a flowchart of an object identification code processing method provided by an embodiment of the present disclosure.
  • the execution body of the method may be the server device 103 in FIG. 1 or the object identification code processing apparatus 200 in FIG. 2 .
  • the process of the object identification code processing method is described with the server device as the execution subject.
  • the server device acquires the object identification code and the first object information associated with the object identification code.
  • the first object information may be understood as object information for the original market.
  • the first object information may include: one or more of an identification code title, an object category, an object attribute, and an object picture, wherein the object category can be understood as the classification of the object in the original market; the object attribute is represented by a key-value pair ( property item - property value).
  • the object picture may be replaced with the storage address of the object picture.
  • the server device aggregates the object identification code and at least a part of the identification codes in the identification code library of the target market to obtain aggregated information.
  • the first object information includes an identification code header.
  • the server device truncates the identification code title multiple times to obtain multiple subtitles; based on the multiple subtitles, aggregates the object identification code and at least a part of the identification codes in the identification code library of the target market.
  • the server device may match each subtitle with the key attributes of at least a part of the identifiers in the identifier library of the target market; and then aggregate the object identifiers with the identifiers corresponding to the matched key attributes.
  • the server device may pre-aggregate at least a part of the identification codes in the identification code library of the target market to obtain multiple aggregation sets; then determine the key attributes common to each aggregation set; and then associate each subtitle with the The key attributes shared by each of the multiple aggregation sets are matched; finally, the object identifiers are aggregated with the identifiers corresponding to the matched key attributes.
  • the first object information includes a first identifier header.
  • the server device can determine the similarity distance between the second identification code title and the first identification code title for the second identification code title of at least a part of the identification code in the identification code library of the target market; and then associate the object identification code with the target market. Similar distance aggregation is performed on at least a part of the identification codes in the identification code library.
  • the server device may input the second identification code title and the first identification code title into the entity alignment model, perform alignment through the entity alignment model, and output the second identification code title and the first identification code title.
  • the similarity distance of the object identification code and at least a part of the identification code in the identification code library of the target market are further aggregated by the similarity distance.
  • the entity alignment model is a model obtained by training a twin neural network based on the identification code title pair of the same object pair.
  • step 303 the server device maps the first object information to the second object information of the target market.
  • the first object information includes a first object category and a first object attribute.
  • the server device may map the first object category to the second object category of the target market; and map the first object attribute to the second object attribute of the target market.
  • the server device may input the first object category and the first object attribute into the text model, perform text analysis on the first object category and the first object attribute through the text model, and output the second object of the target market category.
  • the first object information further includes a picture of the object.
  • the server device can input the object picture into the object recognition model, identify the object in the object picture through the object recognition model, and output the second object category of the target market.
  • the server device can identify text information in the object image; and then input the text information into the text model, analyze the text information through the text model, and output the second object category of the target market.
  • the first object information further includes an identification code header.
  • the server device can input the identification code title into the text model, perform text analysis on the identification code title through the text model, and output the second object category of the target market.
  • the server device may modify the second object category based on the first rule and/or the second rule;
  • the rules are based on algorithm batch mining rules; the second rule is based on rules accumulated by industry experts.
  • the server device may map the first object category to the second object category of the target market based on the manually precipitated category mapping relationship between channels.
  • the server device may modify the second object category based on user feedback information.
  • the server device may perform attribute mapping in a manner similar to category mapping, which is not repeated here in order to avoid repetition. In some embodiments, the server device may perform attribute correction in a manner similar to category correction, which will not be repeated here in order to avoid repetition.
  • the server device stores the association relationship among the object identification code, the aggregation information, and the second object information in the identification code library of the target market.
  • the aggregated information and the second object information corresponding to the object identification code can be searched from the identification code library later.
  • the market caused the object release to fail.
  • a plurality of aggregated objects can be released based on one object identification code, thereby improving the efficiency of object release.
  • FIG. 4 is an exemplary block diagram of an object publishing apparatus 400 provided by an embodiment of the present disclosure.
  • the object publishing apparatus 400 may be implemented as part of the client device 101 in FIG. 1 .
  • the object publishing apparatus 400 may be divided into a plurality of units, for example, including but not limited to: a displaying unit 401 , an obtaining unit 402 , a searching unit 403 , a backfilling unit 404 and a publishing unit 405 .
  • the details of each unit are as follows:
  • the display unit 401 is used to display the object release page in response to the object release request.
  • the presentation unit 401 may acquire resources for displaying the object publishing page from the server device, and then render the object publishing page by using the acquired resources.
  • the merchant may then enter the object identifier in the object posting page.
  • Merchants can also use code scanning guns, smart phones and other scanning devices to scan the object identification code, and the scanning device automatically fills the object identification code into the object release page without manual input.
  • the object identification code may be a representation identification code of multi-field attribute information of objects that can be quickly identified, such as barcodes, two-dimensional codes, and three-dimensional codes.
  • the obtaining unit 402 is configured to obtain the object identification code.
  • the obtaining unit 402 may receive the object identification code input in the object publishing page.
  • the merchant can manually input the object identification code on the object publishing page, and then the obtaining unit 402 can receive the object identification code input on the object publishing page.
  • the obtaining unit 402 may obtain the object identification code sent by the code scanning device.
  • the merchant can scan the object identification code with a code scanning device such as a code scanning gun and a smart phone, and the code scanning device automatically fills the object identification code into the object publishing page without manual input.
  • the search unit 403 is configured to search, based on the object identification code, from the identification code library of the target market, the aggregated information and the second object information associated with the object identification code.
  • the aggregated information is aggregated information obtained by the server device associating the first object information of the original market based on the object identification code, and by aggregating the object identification code and at least a part of the identification codes in the identification code library of the target market.
  • the second object information is the second object information obtained by the server device mapping the first object information to the target market.
  • the search unit 403 may directly access the identification code library of the target market through the network, and then search for the aggregated information and second object information associated with the object identification code. In some embodiments, the search unit 403 may send a query request carrying the object identification code to the server device, and the server device accesses the identification code library to search, and returns the search result to the search unit 403 .
  • the search unit 403 if it does not find the aggregated information and the second object information associated with the object identification code, it sends an object identification code processing request to the server device, and the server device processes the object identification code to obtain The aggregated information associated with the object identification code and the second object information.
  • the backfilling unit 404 is configured to backfill the object publishing page with the second object information, and backfill the object information corresponding to at least a part of the identification code corresponding to the aggregated information in the object publishing page.
  • the backfilling unit 404 may backfill the second object information into the next-level page of the object publishing page, and correspondingly, backfill at least a part of the identification code corresponding to the aggregated information in the next-level page of the object publishing page. Corresponding object information.
  • the publishing unit 405 is configured to publish a plurality of object information in the object publishing page in response to the information confirmation operation.
  • a plurality of aggregated objects can be released based on one object identification code, thereby improving the efficiency of object release.
  • the object information of the object identification code in the target market can be found from the identification code library of the target market, so as to ensure the successful release of the object, and the object information of the object identification code in the original market cannot be directly applied to the target market. As a result, the object release fails.
  • each unit in the object publishing apparatus 400 is only a logical function division, and there may be other division methods in actual implementation, for example, multiple units may be implemented as one unit; one unit may also be divided for multiple subunits. It can be understood that each unit or sub-unit can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art may use different methods for implementing the described functionality for each particular application.
  • FIG. 5 is a flowchart of an object publishing method provided by an embodiment of the present disclosure, and the execution body of the method may be the client device 101 in FIG. 1 .
  • the process of the object publishing method is described with the client device as the execution subject.
  • the client device responds to the object publishing request and displays the object publishing page.
  • the client device may obtain resources for displaying the object publishing page from the server device, and then render the object publishing page using the obtained resources.
  • the merchant may then enter the object identifier in the object posting page.
  • Merchants can also use code scanning guns, smart phones and other scanning devices to scan the object identification code, and the scanning device automatically fills the object identification code into the object release page without manual input.
  • the object identification code may be a representation identification code of multi-field attribute information of objects that can be quickly identified, such as barcodes, two-dimensional codes, and three-dimensional codes.
  • the client device obtains the object identification code.
  • the client device may receive the object identification code entered in the object posting page.
  • the merchant can manually input the object identification code on the object publishing page, and then the client device can receive the object identification code input on the object publishing page.
  • the client device can obtain the object identification code sent by the code scanning device.
  • the merchant can scan the object identification code with a code scanning device such as a code scanning gun and a smart phone, and the code scanning device automatically fills the object identification code into the object publishing page without manual input.
  • the client device searches for the aggregated information and second object information associated with the object identification code from the identification code library of the target market.
  • the aggregated information is aggregated information obtained by the server device associating the first object information of the original market based on the object identification code, and by aggregating the object identification code and at least a part of the identification codes in the identification code library of the target market.
  • the second object information is the second object information obtained by the server device mapping the first object information to the target market.
  • the client device can directly access the identification code library of the target market through the network, and then look up the aggregated information and the second object information associated with the object identification code.
  • the client device may send a query request carrying the object identification code to the server device, and the server device accesses the identification code library to search, and returns the search result to the client device.
  • the client device if it does not find the aggregated information and second object information associated with the object identification code, it sends an object identification code processing request to the server device, and the server device processes the object identification code to obtain The aggregated information associated with the object identification code and the second object information.
  • step 504 the client device backfills the object publishing page with the second object information, and backfills the object publishing page with object information corresponding to at least a part of the identification code corresponding to the aggregated information.
  • the client device may backfill the second object information into the next-level page of the object publishing page, and correspondingly, backfill at least a part of the identification code corresponding to the aggregated information in the lower-level page of the object publishing page. Corresponding object information.
  • the client device responds to the information confirmation operation, and publishes a plurality of object information in the object publishing page.
  • a plurality of aggregated objects can be released based on one object identification code, thereby improving the efficiency of object release.
  • the object information of the object identification code in the target market can be found from the identification code library of the target market, so as to ensure the successful release of the object, and the object information of the object identification code in the original market cannot be directly applied to the target market. As a result, the object release fails.
  • FIG. 6 is a schematic structural diagram of an electronic device provided by an embodiment of the present disclosure.
  • the electronic device includes: at least one processor 601 , at least one memory 602 and at least one communication interface 603 .
  • the various components in the electronic device are coupled together by a bus system 604 .
  • the communication interface 603 is used for information transmission with external devices. Understandably, the bus system 604 is used to implement the connection communication between these components.
  • the bus system 604 also includes a power bus, a control bus and a status signal bus. However, for clarity of illustration, the various buses are labeled as bus system 604 in FIG. 6 .
  • the memory 602 in this embodiment may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memory.
  • memory 602 stores the following elements, executable units or data structures, or subsets thereof, or extended sets of them: operating systems and applications.
  • the operating system includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., for implementing various basic tasks and processing hardware-based tasks.
  • Applications including various applications, such as media players (Media Player), browsers (Browser), etc., are used to implement various application tasks.
  • a program for implementing the object identification code processing method or the object publishing method provided by the embodiments of the present disclosure may be included in an application program.
  • the processor 601 calls the program or instruction stored in the memory 602, specifically, the program or instruction stored in the application program, and the processor 601 is configured to execute the object identification code processing provided by the embodiment of the present disclosure The steps of each embodiment of the method or object publishing method.
  • the object identification code processing method or the object publishing method provided by the embodiments of the present disclosure may be applied to the processor 601 or implemented by the processor 601 .
  • the processor 601 may be an integrated circuit chip with signal processing capability. In the implementation process, each step of the above-mentioned method may be completed by an integrated logic circuit of hardware in the processor 601 or an instruction in the form of software.
  • the above-mentioned processor 601 can be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a ready-made programmable gate array (Field Programmable Gate Array, FPGA) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.
  • a general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
  • the steps of the object identification code processing method or the object publishing method provided by the embodiments of the present disclosure may be directly embodied as executed by a hardware decoding processor, or executed by a combination of hardware and software units in the decoding processor.
  • the software unit may be located in random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory, registers and other storage media mature in the art.
  • the storage medium is located in the memory 602, and the processor 601 reads the information in the memory 602, and completes the steps of the method in combination with its hardware.
  • Embodiments of the present disclosure also provide a non-transitory computer-readable storage medium, where the non-transitory computer-readable storage medium stores programs or instructions, the programs or instructions cause a computer to execute, for example, a method for processing an object identification code or a method for issuing an object
  • the steps of each embodiment are not repeated here in order to avoid repeated description.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • General Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • General Engineering & Computer Science (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Health & Medical Sciences (AREA)
  • Multimedia (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

本公开实施例涉及一种对象识别码处理方法、对象发布方法、装置、设备及介质。本公开的至少一个实施例中，通过对象识别码关联的原市场的对象信息，将对象识别码与目标市场的识别码库中至少一部分识别码进行聚合，便于在发布对象时，可以基于一个对象识别码而发布多个聚合的对象，提高对象发布效率；另外，通过将对象识别码关联的原市场的对象信息映射为目标市场的对象信息，进而在目标市场的识别码库中存储对象识别码、聚合信息、对象识别码在目标市场的对象信息之间的关联关系，便于在发布对象时，可从目标市场的识别码库中查找到对象识别码在目标市场的对象信息，确保成功发布对象。

Description

对象识别码处理方法、对象发布方法、装置、设备及介质
本公开要求2021年01月27日递交的申请号为202110113486.X、发明名称为“对象识别码处理方法、对象发布方法、装置、设备及介质”中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本公开实施例涉及数据处理技术领域，具体涉及一种对象识别码处理方法、对象发布方法、装置、设备及介质。
背景技术
传统的对象发布方式为：商家将全部对象信息人工方式逐个填写到对象发布页面中，点击对象发布页面中的发布按钮来发布对象。传统的对象发布方式需要的时间较多，对象发布效率较低，且可能因填写不当导致对象发布失败。
另外，本公开发明人还发现：商家在对象发布页面中填写的对象信息可能不符合目标市场对于对象信息的定义，若直接将对象信息填写到对象发布页面，可能导致对象发布失败。
上述对问题的发现过程的描述，仅用于辅助理解本公开的技术方案，并不代表承认上述内容是现有技术。
发明内容
为了解决现有技术存在的至少一个问题，本公开的至少一个实施例提供了一种对象识别码处理方法、对象发布方法、装置、电子设备及存储介质。
第一方面，本公开实施例提出一种对象识别码处理方法，包括：
获取对象识别码以及与所述对象识别码关联的第一对象信息；
基于所述第一对象信息，将所述对象识别码与目标市场的识别码库中至少一部分识别码进行聚合，得到聚合信息；
将所述第一对象信息映射为所述目标市场的第二对象信息；
在所述目标市场的识别码库中，存储所述对象识别码、所述聚合信息、所述第二对象信息之间的关联关系。
在一些实施例中，所述第一对象信息包括识别码标题；
所述基于所述第一对象信息，将所述对象识别码与目标市场的识别码库中至少一部分识别码进行聚合包括：
将所述识别码标题进行多次截断，得到多个子标题；
基于所述多个子标题，将所述对象识别码与目标市场的识别码库中至少一 部分识别码进行聚合。
在一些实施例中，所述基于所述多个子标题，将所述对象识别码与目标市场的识别码库中至少一部分识别码进行聚合包括：
将每个所述子标题与所述目标市场的识别码库中至少一部分识别码的关键属性进行匹配；
将所述对象识别码与匹配的关键属性所对应的识别码进行聚合。
在一些实施例中，所述将每个所述子标题与所述目标市场的识别码库中至少一部分识别码的关键属性进行匹配包括：
预先将所述目标市场的识别码库中至少一部分识别码进行聚合，得到多个聚合集；
确定每个所述聚合集所共有的关键属性；
将每个所述子标题与所述多个聚合集各自共有的关键属性进行匹配。
在一些实施例中，所述第一对象信息包括第一识别码标题；
所述基于所述第一对象信息，将所述对象识别码与目标市场的识别码库中至少一部分识别码进行聚合包括：
针对所述目标市场的识别码库中至少一部分识别码的第二识别码标题，确定该第二识别码标题与所述第一识别码标题之间的相似距离；
将所述对象识别码与目标市场的识别码库中至少一部分识别码进行相似距离聚合。
在一些实施例中，所述确定该第二识别码标题与所述第一识别码标题之间的相似距离包括：
将该第二识别码标题和所述第一识别码标题输入实体对齐模型，通过所述实体对齐模型进行对齐，输出该第二识别码标题与所述第一识别码标题之间的相似距离。
在一些实施例中，所述实体对齐模型为基于同款对象对的识别码标题对训练孪生神经网络得到的模型。
在一些实施例中，所述第一对象信息包括第一对象类目和第一对象属性；所述将所述第一对象信息映射为所述目标市场的第二对象信息包括：
将所述第一对象类目映射为所述目标市场的第二对象类目；
将所述第一对象属性映射为所述目标市场的第二对象属性。
在一些实施例中，所述将所述第一对象类目映射为所述目标市场的第二对象类目包括：
将所述第一对象类目和所述第一对象属性输入文本模型，通过所述文本模型对所述第一对象类目和所述第一对象属性进行文本分析，输出所述目标市场 的第二对象类目。
在一些实施例中，所述第一对象信息还包括对象图片；
所述将所述第一对象类目映射为所述目标市场的第二对象类目包括：
将所述对象图片输入对象识别模型，通过所述对象识别模型识别所述对象图片中的对象，输出所述目标市场的第二对象类目。
在一些实施例中，所述将所述第一对象类目映射为所述目标市场的第二对象类目还包括：
识别所述对象图片中的文本信息；
将所述文本信息输入所述文本模型，通过所述文本模型对所述文本信息进行分析，输出所述目标市场的第二对象类目。
在一些实施例中，所述第一对象信息还包括识别码标题；
所述将所述第一对象类目映射为所述目标市场的第二对象类目还包括：
将所述识别码标题输入所述文本模型，通过所述文本模型对所述识别码标题进行文本分析，输出所述目标市场的第二对象类目。
在一些实施例中，所述将所述第一对象类目映射为所述目标市场的第二对象类目后，所述方法还包括：
基于第一规则和/或第二规则，修正所述第二对象类目；其中，所述第一规则为基于算法批量挖掘的规则；所述第二规则为基于行业专家积累的规则。
在一些实施例中，所述将所述第一对象类目映射为所述目标市场的第二对象类目包括：
基于人工沉淀的渠道间的类目映射关系，将所述第一对象类目映射为所述目标市场的第二对象类目。
在一些实施例中，所述将所述第一对象类目映射为所述目标市场的第二对象类目后，所述方法还包括：
基于用户反馈信息，修正所述第二对象类目。
第二方面，本公开实施例还提出一种对象发布方法，包括：
响应对象发布请求，展示对象发布页面；
获取对象识别码；
基于所述对象识别码，从所述目标市场的识别码库中，查找所述对象识别码关联的聚合信息和第二对象信息；
将所述第二对象信息回填所述对象发布页面，并在所述对象发布页面中回填所述聚合信息对应的至少一部分识别码所对应的对象信息；
响应信息确认操作，将所述对象发布页面中对象信息进行发布。
在一些实施例中，所述方法还包括：
若未查找到所述对象识别码关联的聚合信息和第二对象信息，则基于第一方面任一实施例提供的对象识别码处理方法处理所述对象识别码，得到所述对象识别码关联的聚合信息和第二对象信息。
第三方面，本公开实施例还提出一种对象识别码处理装置，包括：
获取单元，用于获取对象识别码以及与所述对象识别码关联的第一对象信息；
聚合单元，用于基于所述第一对象信息，将所述对象识别码与目标市场的识别码库中至少一部分识别码进行聚合，得到聚合信息；
映射单元，用于将所述第一对象信息映射为所述目标市场的第二对象信息；
存储单元，用于在所述目标市场的识别码库中，存储所述对象识别码、所述聚合信息、所述第二对象信息之间的关联关系。
第四方面，本公开实施例还提出一种对象发布装置，包括：
展示单元，用于响应对象发布请求，展示对象发布页面；
获取单元，用于获取对象识别码；
查找单元，用于基于所述对象识别码，从所述目标市场的识别码库中，查找所述对象识别码关联的聚合信息和第二对象信息；
回填单元，用于将所述第二对象信息回填所述对象发布页面，并在所述对象发布页面中回填所述聚合信息对应的至少一部分识别码所对应的对象信息；
发布单元，用于响应信息确认操作，将所述对象发布页面中对象信息进行发布。
第五方面，本公开实施例还提出一种电子设备，包括：处理器和存储器；所述处理器通过调用所述存储器存储的程序或指令，用于执行如第一方面任一实施例所述方法的步骤。
第六方面，本公开实施例还提出一种非暂态计算机可读存储介质，用于存储程序或指令，所述程序或指令使计算机执行如第一方面任一实施例所述方法的步骤。
可见，本公开的至少一个实施例中，通过对象识别码关联的原市场的对象信息，将对象识别码与目标市场的识别码库中至少一部分识别码进行聚合，便于在发布对象时，可以基于一个对象识别码而发布多个聚合的对象，利用对象识别码，协助商家快速发布对象，提高对象发布效率；另外，通过将对象识别码关联的原市场的对象信息映射为目标市场的对象信息，进而在目标市场的识别码库中存储对象识别码、聚合信息、对象识别码在目标市场的对象信息之间的关联关系，便于在发布对象时，可从目标市场的识别码库中查找到对象识别码在目标市场的对象信息，确保成功发布对象，不会因为对象识别码在原市场 的对象信息无法直接应用于目标市场而导致对象发布失败。
附图说明
为了更清楚地说明本公开实施例的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本公开的一些实施例，对于本领域普通技术人员来讲，还可以根据这些附图获得其他的附图。
图1是本公开实施例提供的一种示例性应用场景图；
图2是本公开实施例提供的一种对象识别码处理装置的框图；
图3是本公开实施例提供的一种对象识别码处理方法的流程图；
图4是本公开实施例提供的一种对象发布装置的框图；
图5是本公开实施例提供的一种对象发布方法的流程图；
图6是本公开实施例提供的一种电子设备的框图。
具体实施方式
为了能够更清楚地理解本公开的上述目的、特征和优点，下面结合附图和实施例对本公开作进一步的详细说明。可以理解的是，所描述的实施例是本公开的一部分实施例，而不是全部的实施例。此处所描述的具体实施例仅仅用于解释本公开，而非对本公开的限定。基于所描述的本公开的实施例，本领域普通技术人员所获得的所有其他实施例，都属于本公开保护的范围。
需要说明的是，在本文中，诸如“第一”和“第二”等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来，而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。
随着网络技术的发展，电子商务(简称电商)应运而生，电商利用网络以电子交易方式进行交易活动和相关服务活动，使传统商业活动各环节电子化、网络化。针对电商而言，存货单元(Stock Keeping Unit，SKU)和标准产品单元(Standard Product Unit，SPU)是两个常用概念。
SKU是库存进出计量的单位(买家购买、商家进货、供应商备货、工厂生产都是依据SKU进行的)，在服装、鞋类、日化等商品中使用最多最普遍。例如，纺织品中一个SKU通常表示：规格、颜色、款式。SKU是物理上不可分割的最小存货单元，也就是说，一款商品，可以根据SKU来确定具体的货物存量。
例如，一件M码(四个尺码：S码、M码、L码、X码)的粉色(三种颜色：粉色、黄色、黑色)某品牌女士风衣，其中M码、粉色就是一组SKU的 组合，而该女士风衣一共有4×3＝12个SKU组合。由于SKU是物理上不可分割的最小存货单元，因此，单凭尺码或者颜色无法确认这款商品的库存情况。
SPU是商品信息聚合的最小单位，是一组可复用、易检索的标准化信息的集合，该集合描述了一个产品的特性，简言之，一款商品就可以对应一个SPU。
在商品信息电子化过程中，商品的特性可以由多个“属性|属性值对”进行描述。“属性|属性值对”完全相同的商品，可以抽象成为一个SPU。
SPU和SKU都是属性值的集合，其中，SPU属性是不会影响到库存和价格的属性，又称为关键属性；而SKU属性是会影响到库存和价格的属性，因此，又可以称为销售属性。一个SPU对应多个SKU，这些SKU之间的区分在于销售属性(例如颜色、规格等属性)的不同。
电商的服务系统在具体的页面中向用户进行商品信息的展示时，通常会首先以SPU为单位进行展示，然后，当用户点击进入到SPU的详情页面后，或者电商的服务系统接收到“加入购物车”或“立即购买”的请求后，电商的服务系统再展示出具体可选的SKU列表的信息。
商家可发布新对象(item)，进而使用户可以在商家的店铺页面中浏览商家发布的新对象的信息。发布的对象通常包括两大类：一，实物商品，例如服装，电器，穿戴等有形的商品，通常是淘宝，拼多多，京东等电商的服务系统提供的功能；二，虚拟商品，例如服务类的商品，包括外卖，视频，音乐，内容等无形的商品。无论是实物商品还是虚拟商品，均是为了出售而生产的劳动成果，是用于交换的劳动产品。因此，发布的对象可以理解为用于出售的产品。
发布对象的应用场景通常包括两种：一，商家在第三方平台入驻，例如淘宝，京东，亚马逊，拼多多等电商的服务系统；二，商家自建电商网站。这两种应用场景都有快速发布对象的需求。传统的对象发布方式为：商家将全部对象信息人工方式逐个填写到对象发布页面中，点击对象发布页面中的发布按钮来发布对象。传统的对象发布方式需要的时间较多，对象发布效率较低，且可能因填写不当导致对象发布失败。
另外，本公开发明人还发现：商家在对象发布页面中填写的对象信息可能不符合目标市场对于对象信息的定义，若直接将对象信息填写到对象发布页面，可能导致对象发布失败。
为此，本公开实施例提供一种对象识别码处理方法、对象发布方法、装置、设备及介质。针对商家在第三方平台入驻的应用场景，本公开实施例提供的对象识别码处理功能和对象快速发布功能嵌在第三方平台内。针对商家自建电商网站的应用场景，本公开实施例提供的对象识别码处理功能和对象快速发布功能可以实现为商业化的服务，通过接口调用的方式使商家自建电商网站也可使 用本公开实施例提供的服务。在一些实施例中，本公开实施例提供的对象识别码处理功能和对象快速发布功能可以应用到跨市场或跨渠道铺货的应用场景，例如，大润发的商品可以通过EAN编码的方式快速发布到天猫超市，节省商家发布商品的时间和成本。
本公开的至少一个实施例中，提供了一种快速发布对象的功能，具体地，通过对象识别码关联的原市场的对象信息，将对象识别码与目标市场的识别码库中至少一部分识别码进行聚合，便于在发布对象时，可以基于一个对象识别码而发布多个聚合的对象，利用对象识别码，协助商家快速发布对象，提高对象发布效率；另外，通过将对象识别码关联的原市场的对象信息映射为目标市场的对象信息，进而在目标市场的识别码库中存储对象识别码、聚合信息、对象识别码在目标市场的对象信息之间的关联关系，便于在发布对象时，可从目标市场的识别码库中查找到对象识别码在目标市场的对象信息，确保成功发布对象，不会因为对象识别码在原市场的对象信息无法直接应用于目标市场而导致对象发布失败。
图1为本公开实施例提供的一种示例性应用场景图，并不表示全部场景。如图1所示，商家可以通过多种客户端设备101发起对象发布请求，例如商家在客户端设备101的浏览器中输入对象发布页面的网址，并确认访问网址，实现发起对象发布请求。
客户端设备101的浏览器可通过网络102从服务端设备103获取显示对象发布页面的资源，进而利用获取的资源渲染出对象发布页面。商家进而可在对象发布页面中输入对象识别码，商家也可以使用扫码枪，智能手机等扫码设备扫描对象识别码，扫码设备自动将对象识别码填入对象发布页面，无需人工输入。在一些实施例中，对象识别码可以为条码、二维码、三维码等可快速识别的对象多字段属性信息的表征识别码。
客户端设备101可将对象识别码发送给服务端设备103，进而可获取服务端设备103发送的对象识别码所关联的对象信息，对象信息例如包括：识别码标题、对象类目、对象属性和对象图片中的一种或多种，其中，对象类目可以理解为对象的分类；对象属性以键值对(属性项-属性值)的方式表示，例如属性项为颜色，属性值为红色。
客户端设备101可将对象识别码所关联的对象信息回填到对象发布页面。商家可以修改回填信息中的错误信息；也可以在确认回填信息无误后，点击对象发布页面上的发布控件。客户端设备101响应发布控件的点击操作，将对象发布页面中的对象信息进行发布。
在一些实施例中，客户端设备101可以为智能手机、笔记本电脑、平板电 脑等可移动的电子设备；客户端设备101也可以为台式计算机、智能电视等固定的电子设备。
服务端设备103接收到用于显示对象发布页面的资源请求后，可将显示对象发布页面所需的资源通过网络102发送给客户端设备101的浏览器，以使客户端设备101的浏览器显示对象发布页面。
服务端设备103接收到客户端设备101发送的对象识别码，进而从识别码库104中查找对象识别码关联的对象信息，并将查找到的对象信息发送给客户端设备101，以使客户端设备101可将对象信息回填到对象发布页面。
在一些实施例中，服务端设备103可以理解为电商的服务系统(硬件实现)或者支持电商的服务系统(软件实现)运行的硬件设备。在一些实施例中，服务端设备103可以是单个服务器，也可以是服务器群组，服务器群组可以是集中式的，也可以是分布式的。
识别码库104为目标市场的识别码库，目标市场可以理解为对象发布所针对的市场。而一个对象识别码所关联的对象信息所针对的市场(简称原市场)可能与目标市场不同，因此，一个对象识别码在入库之前，需要由服务端设备103对该对象识别码进行处理，以使该对象识别码关联新的对象信息，这些新的对象信息是针对目标市场的信息，从而能够将该对象识别码入库。
识别码库104中除了可以存储对象识别码、对象信息之间的关联关系，进一步可以存储对象识别码、聚合信息之间的关联关系。其中，聚合信息为该对象识别码与识别码库104中已存储的至少一部分识别码进行聚合得到的信息，具体地，服务端设备103可将该对象识别码与目标市场的识别码库104中至少一部分识别码进行聚合，得到聚合信息。这样，服务端设备103可将聚合信息对应的至少一部分识别码所对应的对象信息也发送给客户端设备101，进而使客户端设备101在对象发布页面中回填聚合信息对应的至少一部分识别码所对应的对象信息，在发布对象时，可以基于一个对象识别码而发布多个聚合的对象，提高对象发布效率。
在一些实施例中，识别码库104可能没有存储服务端设备103所要查找的对象识别码，这时，服务端设备103需要对该对象识别码进行处理，得到该对象识别码在目标市场中的对象信息以及聚合信息，并将该对象识别码、对象信息、聚合信息之间的关联关系存储到识别码库104中，以后后续查找复用。
图2为本公开实施例提供的一种对象识别码处理装置200的示例性框图。在一些实施例中，对象识别码处理装置200可以实现为图1中的服务端设备103的一部分。如图2所示，对象识别码处理装置200可划分为多个单元，例如可包括但不限于：获取单元201、聚合单元202、映射单元203和存储单元204。 各单元具体说明如下：
获取单元201，用于获取对象识别码，还可以获取与对象识别码关联的第一对象信息。第一对象信息可以理解为针对原市场的对象信息。在一些实施例中，对象识别码以及与对象识别码关联的第一对象信息可通过向第三方公司购买得到，因此，获取单元201可以批量获取对象识别码以及与对象识别码关联的第一对象信息。在一些实施例中，获取单元201可从不同渠道获取对象识别码以及与对象识别码关联的第一对象信息，其中，不同渠道可以理解为不同的电商平台，例如国内电商的服务系统或者国外电商的服务系统。
第一对象信息可包括：识别码标题、对象类目、对象属性和对象图片中的一种或多种，其中，对象类目可以理解为对象在原市场中的分类；对象属性以键值对(属性项-属性值)的方式表示。在一些实施例中，对象图片可以替换为对象图片的存储地址。
在一些实施例中，同一渠道获取的第一对象信息所包括的内容有两种定义方式：
一种为该渠道的对象发布时的信息定义。例如：
内存条在原市场的对象类目：电脑硬件/显示器/电脑周边>>内存；
内存条在原市场的对象属性包括但不限于：
属性项i为品牌；属性项i的属性值则为金士顿，威刚，美商，海盗船，三星，影驰等中的一个；
属性项ii为内存容量；属性项ii的属性值则为2G，4G，8G，16G等中的一个；
属性项iii为传输类型；属性项iii的属性值则为DDR，DDR2，DDR3，DDR4等中的一个；
属性项iv为内存频率；属性项iv的属性值则为2333MHz，3000MHz，4000MHz，4800MHz等中的一个。
另一种为该渠道售卖的对象的信息定义。例如：
内存条的识别码标题为：全新三星128G 2S4RX4 PC4-2666V-RB2 DDR4 2666V ECC REG内存条；
内存条在原市场的对象属性包括：属性项为品牌，对应的属性值为三星；属性项为内存容量，对应的属性值为128G；属性项为传输类型，对应的属性值为DDR4；属性项为成色，对应的属性值为全新；属性项为售后服务，对应的属性值为店铺三包；属性项为适用对象，对应的属性值为服务器。
聚合单元202，用于将对象识别码与目标市场的识别码库中至少一部分识别码进行聚合，得到聚合信息。在一些实施例中，聚合单元202基于第一对象 信息，将对象识别码与目标市场的识别码库中至少一部分识别码进行聚合。
在一些实施例中，聚合单元202可将第一对象信息包括的识别码标题进行多次截断，得到多个子标题。聚合单元202从识别码标题的首个词开始计算长度，当计算长度达到预设截断长度，则进行截断，得到一个子标题。聚合单元202可对识别码标题进行多次截断，每次截断时参考的预设截断长度不同。预设截断长度可根据实际需要进行设置，本实施例不限定预设截断长度的具体取值，例如，预设截断长度可以为3个词、5个词或7个词。实施例中从识别码标题的首个词开始计算长度，是因为条码聚合的信息通常包含在识别码标题首个词开始的一段长度内。
例如，条码06923410719178的标题为“Philips HTB5570D/93 home cinema system 3D 5.1 channels 1000W”，以预设截断长度为5个词进行截断，得到子标题为“Philips HTB5570D/93 home cinema system”。
在一些实施例中，聚合单元202可基于多个子标题，将对象识别码与目标市场的识别码库中至少一部分识别码进行聚合。在一些实施例中，聚合单元202可将每个子标题与目标市场的识别码库中至少一部分识别码的关键属性进行匹配。在一些实施例中，聚合单元202可预先将目标市场的识别码库中至少一部分识别码进行聚合，得到多个聚合集；进而确定每个聚合集所共有的关键属性；从而将每个子标题与多个聚合集各自共有的关键属性进行匹配。在一些实施例中，聚合单元202可将对象识别码与匹配的关键属性所对应的识别码进行聚合。
例如，一个聚合集所共有的关键属性包括：品牌＝Philips和型号＝HTB5570D/93，该聚合集可命名为：Philips HTB5570D/93分组。前述子标题“Philips HTB5570D/93 home cinema system”可与这个集合集所共有的关键属性：品牌＝Philips和型号＝HTB5570D/93匹配上，因此条码06923410719178会聚合在Philips HTB5570D/93分组中。
在一些实施例中，聚合单元202可确定第一对象信息包括的第一识别码标题分别与目标市场的识别码库中至少一部分识别码的第二识别码标题之间的相似距离。在一些实施例中，聚合单元202可将第一识别码标题和一个第二识别码标题输入实体对齐模型，通过实体对齐模型进行对齐，输出第一识别码标题与该第二识别码标题之间的相似距离。在一些实施例中，实体对齐模型为基于同款对象对的识别码标题对训练孪生神经网络得到的模型，同款对象可以理解为具有相同关键属性的对象。在一些实施例中，实体对齐模型可以为其他神经网络模型，第一识别码标题和一个第二识别码标题输入实体对齐模型后，由实体对齐模型转换为两个128维向量，进而计算这两个128维向量的相似距离，计算相似距离可采用欧式(eculidean)距离、余弦(cosine)距离等距离公式， 这些距离公式均为常用公式，不再赘述。
在一些实施例中，聚合单元202可将对象识别码与目标市场的识别码库中至少一部分识别码进行相似距离聚合。在一些实施例中，聚合单元202可基于DBSCAN算法、K-means算法等聚类算法进行相似距离聚合，这些聚类算法均为常用算法，不再赘述。
在一些实施例中，聚合单元202可将采用不同方式得到的聚合结果进行融合，得到最终的聚合信息。
映射单元203，用于将第一对象信息映射为目标市场的第二对象信息。在一些实施例中，映射单元203可将第一对象信息中的第一对象类目映射为目标市场的第二对象类目。
例如，内存条的第一对象信息中的第一对象类目为电脑硬件/显示器/电脑周边>>内存；而目标市场的对象类目为电脑/电脑配件>>内存；因此，映射单元203将“电脑硬件/显示器/电脑周边>>内存”映射为“电脑/电脑配件>>内存”。
在一些实施例中，映射单元203可将第一对象类目和第一对象属性输入文本模型，通过文本模型对第一对象类目和第一对象属性进行文本分析，输出目标市场的第二对象类目。在一些实施例中，映射单元203还可识别第一对象信息中的对象图片中的文本信息，识别方式例如为将对象图片输入图片识别网络，通过图片识别网络进行图片识别，输出对象图片中的文本信息，其中，图片识别网络可预先训练神经网络得到，在此不限定神经网络的类型。映射单元203将识别的文本信息输入文本模型，通过文本模型对文本信息进行分析，输出目标市场的第二对象类目。在一些实施例中，映射单元203可将第一对象信息中的识别码标题输入文本模型，通过文本模型对识别码标题进行文本分析，输出目标市场的第二对象类目其中。其中，文本模型可以基于输入和输出的目标进行训练得到。
在一些实施例中，映射单元203可将第一对象信息中的对象图片输入对象识别模型，通过对象识别模型识别对象图片中的对象，输出目标市场的第二对象类目。在一些实施例中，映射单元203可基于第一对象信息中对象图片的存储地址获取对象图片，进而将对象图片输入对象识别模型，通过对象识别模型识别对象图片中的对象，输出目标市场的第二对象类目。其中，对象识别模型为预先训练神经网络得到的模型，在此不限定神经网络的类型。
在一些实施例中，映射单元203可将采用不同方式得到的第二对象类目进行融合，得到最终的第二对象类目，避免采用单一方式得到的第二对象类目不准确或无法映射的问题。
在一些实施例中，映射单元203将第一对象类目映射为目标市场的第二对 象类目后，基于第一规则和/或第二规则，修正第二对象类目，以提升映射准确性；其中，第一规则为基于算法批量挖掘的规则；第二规则为基于行业专家积累的规则。
例如，目标市场的对象类目如下表1所示：
表1 目标市场的对象类目
Figure PCTCN2022072842-appb-000001
例如：映射单元203将识别码标题“8424西瓜3元/斤”输入文本模型，通过文本模型对识别码标题进行文本分析，输出目标市场的第二对象类目为“其他西瓜”。映射单元203基于第一规则可通过识别码标题中出现的“8424”这个关键词，修正第二对象类目为“8424西瓜”。
又例如，第二规则为基于行业专家积累的规则，比如通过行业专家的积累经验，白酒类目下，如果度数大于41度，则确定为“高度白酒”这个类目。相应地，映射单元203将识别码标题“42度白酒”输入文本模型，通过文本模型对识别码标题进行文本分析，输出目标市场的第二对象类目为“白酒”。映射单元203基于第二规则可修正第二对象类目为“高度白酒”。
在一些实施例中，映射单元203可基于人工沉淀的渠道间的类目映射关系，将第一对象类目映射为目标市场的第二对象类目。例如，人工沉淀的渠道间的类目映射关系如下表2所示：
表2 人工沉淀的渠道间的类目映射关系
Figure PCTCN2022072842-appb-000002
因此，映射单元203可基于人工沉淀的渠道间的类目映射关系，直接将第一对象类目映射为目标市场的第二对象类目。例如，直接将第一渠道的对象类目“咖啡/麦片/冲饮>>饮料>>功能饮品”映射为目标市场(例如第三渠道)的对象类目“饮料>>运动机能饮料>>运动饮料”，类目映射更加快速高效。
在一些实施例中，映射单元203将第一对象类目映射为目标市场的第二对象类目后，基于用户反馈信息，修正第二对象类目，以提升映射准确性。
例如，用户反馈信息为：将第一对象类目“苹果味的饮料”映射为第二对象类目“苹果”有误，应当映射为“饮料”，映射单元203可基于该用户反馈信息，修正第二对象类目“苹果”为“饮料”。
进一步地，映射单元203可基于用户反馈信息，修改训练数据，重新训练文本模型。例如重新训练文本模型，以使文本模型基于输入“苹果味的饮料”，可输出“饮料”。
进一步地，映射单元203可基于用户反馈信息，补充之前漏掉的专家经验，也即完善基于行业专家积累的规则(即第二规则)。例如，用户反馈需要区分车厘子的产地，那么可以补充第二规则：如果车厘子的产地是中国以外的地区就算“进口车厘子”，如果是中国产的就算“国产车厘子”。
在一些实施例中，映射单元203可将第一对象信息中的第一对象属性映射为目标市场的第二对象属性。例如，第一对象信息中的第一对象属性的属性项为颜色，对应的属性值为红色；而目标市场的对象属性的属性项为Color，对应的属性值为Red；因此，映射单元203将“属性项为颜色，对应的属性值为红色”映射为“属性项为Color，对应的属性值为Red”。
在一些实施例中，映射单元203可采用类似于类目映射的方式进行属性映射，为避免重复，在此不再赘述。在一些实施例中，映射单元203可采用类似于类目修正的方式进行属性修正，为避免重复，在此不再赘述。
存储单元204，用于在目标市场的识别码库中，存储对象识别码、聚合信息、第二对象信息之间的关联关系。本实施例中，通过将对象识别码、聚合信息、第二对象信息之间的关联关系存储到识别码库中，后续可从识别码库中查找对象识别码对应的聚合信息和第二对象信息，便于直接将聚合信息对应的至少一部分识别码所对应的对象信息以及第二对象信息回填对象发布页面，确保成功发布对象，不会因为对象识别码在原市场的第一对象信息无法直接应用于目标市场而导致对象发布失败。另外，在发布对象时，可以基于一个对象识别码而发布多个聚合的对象，提高对象发布效率。
在一些实施例中，对象识别码处理装置200中各单元的划分仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如获取单元201、聚合单元202、映射单元203和存储单元204中的至少两个单元可以实现为一个单元；获取单元201、聚合单元202、映射单元203或存储单元204也可以划分为多个子单元。可以理解的是，各个单元或子单元能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。本领域技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能。
图3是本公开实施例提供的一种对象识别码处理方法的流程图，该方法的执行主体可以为图1中的服务端设备103，也可以为图2中的对象识别码处理装置200。为便于描述，以服务端设备为执行主体描述对象识别码处理方法的 流程。
如图3所示，在步骤301中，服务端设备获取对象识别码以及与对象识别码关联的第一对象信息。其中，第一对象信息可以理解为针对原市场的对象信息。第一对象信息可包括：识别码标题、对象类目、对象属性和对象图片中的一种或多种，其中，对象类目可以理解为对象在原市场中的分类；对象属性以键值对(属性项-属性值)的方式表示。在一些实施例中，对象图片可以替换为对象图片的存储地址。
在步骤302中，服务端设备基于第一对象信息，将对象识别码与目标市场的识别码库中至少一部分识别码进行聚合，得到聚合信息。
在一些实施例中，第一对象信息包括识别码标题。服务端设备将识别码标题进行多次截断，得到多个子标题；基于多个子标题，将对象识别码与目标市场的识别码库中至少一部分识别码进行聚合。
在一些实施例中，服务端设备可将每个子标题与目标市场的识别码库中至少一部分识别码的关键属性进行匹配；进而将对象识别码与匹配的关键属性所对应的识别码进行聚合。
在一些实施例中，服务端设备可预先将目标市场的识别码库中至少一部分识别码进行聚合，得到多个聚合集；进而确定每个聚合集所共有的关键属性；从而将每个子标题与多个聚合集各自共有的关键属性进行匹配；最终将对象识别码与匹配的关键属性所对应的识别码进行聚合。
在一些实施例中，第一对象信息包括第一识别码标题。服务端设备可针对目标市场的识别码库中至少一部分识别码的第二识别码标题，确定该第二识别码标题与第一识别码标题之间的相似距离；进而将对象识别码与目标市场的识别码库中至少一部分识别码进行相似距离聚合。
在一些实施例中，服务端设备可将该第二识别码标题和第一识别码标题输入实体对齐模型，通过实体对齐模型进行对齐，输出该第二识别码标题与第一识别码标题之间的相似距离；进而将对象识别码与目标市场的识别码库中至少一部分识别码进行相似距离聚合。其中，实体对齐模型为基于同款对象对的识别码标题对训练孪生神经网络得到的模型。
在步骤303中，服务端设备将第一对象信息映射为目标市场的第二对象信息。
在一些实施例中，第一对象信息包括第一对象类目和第一对象属性。服务端设备可将第一对象类目映射为目标市场的第二对象类目；并将第一对象属性映射为目标市场的第二对象属性。
在一些实施例中，服务端设备可将第一对象类目和第一对象属性输入文本 模型，通过文本模型对第一对象类目和第一对象属性进行文本分析，输出目标市场的第二对象类目。
在一些实施例中，第一对象信息还包括对象图片。服务端设备可将对象图片输入对象识别模型，通过对象识别模型识别对象图片中的对象，输出目标市场的第二对象类目。在一些实施例中，服务端设备可识别对象图片中的文本信息；进而将文本信息输入文本模型，通过文本模型对文本信息进行分析，输出目标市场的第二对象类目。
在一些实施例中，第一对象信息还包括识别码标题。服务端设备可将识别码标题输入文本模型，通过文本模型对识别码标题进行文本分析，输出目标市场的第二对象类目。
在一些实施例中，服务端设备将第一对象类目映射为目标市场的第二对象类目后，可基于第一规则和/或第二规则，修正第二对象类目；其中，第一规则为基于算法批量挖掘的规则；第二规则为基于行业专家积累的规则。
在一些实施例中，服务端设备可基于人工沉淀的渠道间的类目映射关系，将第一对象类目映射为目标市场的第二对象类目。
在一些实施例中，服务端设备将第一对象类目映射为目标市场的第二对象类目后，可基于用户反馈信息，修正第二对象类目。
在一些实施例中，服务端设备可采用类似于类目映射的方式进行属性映射，为避免重复，在此不再赘述。在一些实施例中，服务端设备可采用类似于类目修正的方式进行属性修正，为避免重复，在此不再赘述。
在步骤304中，服务端设备在目标市场的识别码库中，存储对象识别码、聚合信息、第二对象信息之间的关联关系。本实施例中，通过将对象识别码、聚合信息、第二对象信息之间的关联关系存储到识别码库中，后续可从识别码库中查找对象识别码对应的聚合信息和第二对象信息，便于直接将聚合信息对应的至少一部分识别码所对应的对象信息以及第二对象信息回填对象发布页面，确保成功发布对象，不会因为对象识别码在原市场的第一对象信息无法直接应用于目标市场而导致对象发布失败。另外，在发布对象时，可以基于一个对象识别码而发布多个聚合的对象，提高对象发布效率。
以上对象识别码处理方法的各实施例的细节可参见图2所示的对象识别码处理装置200的各实施例的细节，为避免重复，不再赘述。
需要说明的是，对于前述的各方法实施例，为了简单描述，故将其都表述为一系列的动作组合，但是本领域技术人员能够理解，本公开实施例并不受所描述的动作顺序的限制，因为依据本公开实施例，某些步骤可以采用其他顺序或者同时进行。另外，本领域技术人员能够理解，说明书中所描述的实施例均 属于可选实施例。
图4是本公开实施例提供的一种对象发布装置400的示例性框图。在一些实施例中，对象发布装置400可以实现为图1中的客户端设备101的一部分。如图4所示，对象发布装置400可划分为多个单元，例如可包括但不限于：展示单元401、获取单元402、查找单元403、回填单元404和发布单元405。各单元具体说明如下：
展示单元401，用于响应对象发布请求，展示对象发布页面。在一些实施例中，展示单元401可从服务端设备获取显示对象发布页面的资源，进而利用获取的资源渲染出对象发布页面。商家进而可在对象发布页面中输入对象识别码。商家也可以使用扫码枪，智能手机等扫码设备扫描对象识别码，扫码设备自动将对象识别码填入对象发布页面，无需人工输入。在一些实施例中，对象识别码可以为条码、二维码、三维码等可快速识别的对象多字段属性信息的表征识别码。
获取单元402，用于获取对象识别码。在一些实施例中，获取单元402可接收对象发布页面中输入的对象识别码。本实施例中，商家可在对象发布页面中手动输入对象识别码，进而获取单元402可接收对象发布页面中输入的对象识别码。在一些实施例中，获取单元402可获取扫码设备发送的对象识别码。本实施例中，商家可使用扫码枪，智能手机等扫码设备扫描对象识别码，扫码设备自动将对象识别码填入对象发布页面，无需人工输入。
查找单元403，用于基于对象识别码，从目标市场的识别码库中，查找对象识别码关联的聚合信息和第二对象信息。其中，聚合信息为服务端设备基于对象识别码关联原市场的第一对象信息，将对象识别码与目标市场的识别码库中至少一部分识别码进行聚合，得到的聚合信息。第二对象信息为服务端设备将第一对象信息映射到目标市场得到的第二对象信息。
在一些实施例中，查找单元403可通过网络直接访问目标市场的识别码库，进而查找对象识别码关联的聚合信息和第二对象信息。在一些实施例中，查找单元403可将携带对象识别码的查询请求发送给服务端设备，由服务端设备访问识别码库进行查找，并将查找结果返回给查找单元403。
在一些实施例中，查找单元403若未查找到对象识别码关联的聚合信息和第二对象信息，则向服务端设备发送对象识别码处理请求，由服务端设备对对象识别码进行处理，得到对象识别码关联的聚合信息和第二对象信息。
回填单元404，用于将第二对象信息回填对象发布页面，并在对象发布页面中回填聚合信息对应的至少一部分识别码所对应的对象信息。在一些实施例中，回填单元404可将第二对象信息回填到对象发布页面的下一级页面中，相 应地，在对象发布页面的下一级页面中回填聚合信息对应的至少一部分识别码所对应的对象信息。
发布单元405，用于响应信息确认操作，将对象发布页面中多个对象信息进行发布。本实施例中，在发布对象时，可以基于一个对象识别码而发布多个聚合的对象，提高对象发布效率。另外，在发布对象时，可从目标市场的识别码库中查找到对象识别码在目标市场的对象信息，确保成功发布对象，不会因为对象识别码在原市场的对象信息无法直接应用于目标市场而导致对象发布失败。
在一些实施例中，对象发布装置400中各单元的划分仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如可以将多个单元实现为一个单元；也可以将一个单元划分为多个子单元。可以理解的是，各个单元或子单元能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。本领域技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能。
图5是本公开实施例提供的一种对象发布方法的流程图，该方法的执行主体可以为图1中的客户端设备101。为便于描述，以客户端设备为执行主体描述对象发布方法的流程。
如图5所示，在步骤501中，客户端设备响应对象发布请求，展示对象发布页面。在一些实施例中，客户端设备可从服务端设备获取显示对象发布页面的资源，进而利用获取的资源渲染出对象发布页面。商家进而可在对象发布页面中输入对象识别码。商家也可以使用扫码枪，智能手机等扫码设备扫描对象识别码，扫码设备自动将对象识别码填入对象发布页面，无需人工输入。在一些实施例中，对象识别码可以为条码、二维码、三维码等可快速识别的对象多字段属性信息的表征识别码。
在步骤502中，客户端设备获取对象识别码。在一些实施例中，客户端设备可接收对象发布页面中输入的对象识别码。本实施例中，商家可在对象发布页面中手动输入对象识别码，进而客户端设备可接收对象发布页面中输入的对象识别码。在一些实施例中，客户端设备可获取扫码设备发送的对象识别码。本实施例中，商家可使用扫码枪，智能手机等扫码设备扫描对象识别码，扫码设备自动将对象识别码填入对象发布页面，无需人工输入。
在步骤503中，客户端设备基于对象识别码，从目标市场的识别码库中，查找对象识别码关联的聚合信息和第二对象信息。其中，聚合信息为服务端设备基于对象识别码关联原市场的第一对象信息，将对象识别码与目标市场的识别码库中至少一部分识别码进行聚合，得到的聚合信息。第二对象信息为服务 端设备将第一对象信息映射到目标市场得到的第二对象信息。
在一些实施例中，客户端设备可通过网络直接访问目标市场的识别码库，进而查找对象识别码关联的聚合信息和第二对象信息。在一些实施例中，客户端设备可将携带对象识别码的查询请求发送给服务端设备，由服务端设备访问识别码库进行查找，并将查找结果返回给客户端设备。
在一些实施例中，客户端设备若未查找到对象识别码关联的聚合信息和第二对象信息，则向服务端设备发送对象识别码处理请求，由服务端设备对对象识别码进行处理，得到对象识别码关联的聚合信息和第二对象信息。
在步骤504中，客户端设备将第二对象信息回填对象发布页面，并在对象发布页面中回填聚合信息对应的至少一部分识别码所对应的对象信息。
在一些实施例中，客户端设备可将第二对象信息回填到对象发布页面的下一级页面中，相应地，在对象发布页面的下一级页面中回填聚合信息对应的至少一部分识别码所对应的对象信息。
在步骤505中，客户端设备响应信息确认操作，将对象发布页面中多个对象信息进行发布。本实施例中，在发布对象时，可以基于一个对象识别码而发布多个聚合的对象，提高对象发布效率。另外，在发布对象时，可从目标市场的识别码库中查找到对象识别码在目标市场的对象信息，确保成功发布对象，不会因为对象识别码在原市场的对象信息无法直接应用于目标市场而导致对象发布失败。
需要说明的是，对于前述的各方法实施例，为了简单描述，故将其都表述为一系列的动作组合，但是本领域技术人员能够理解，本公开实施例并不受所描述的动作顺序的限制，因为依据本公开实施例，某些步骤可以采用其他顺序或者同时进行。另外，本领域技术人员能够理解，说明书中所描述的实施例均属于可选实施例。
图6是本公开实施例提供的一种电子设备的结构示意图。如图6所示，电子设备包括：至少一个处理器601、至少一个存储器602和至少一个通信接口603。电子设备中的各个组件通过总线系统604耦合在一起。通信接口603，用于与外部设备之间的信息传输。可理解地，总线系统604用于实现这些组件之间的连接通信。总线系统604除包括数据总线之外，还包括电源总线、控制总线和状态信号总线。但为了清楚说明起见，在图6中将各种总线都标为总线系统604。
可以理解，本实施例中的存储器602可以是易失性存储器或非易失性存储器，或可包括易失性和非易失性存储器两者。
在一些实施方式中，存储器602存储了如下的元素，可执行单元或者数据 结构，或者他们的子集，或者他们的扩展集：操作系统和应用程序。
其中，操作系统，包含各种系统程序，例如框架层、核心库层、驱动层等，用于实现各种基础任务以及处理基于硬件的任务。应用程序，包含各种应用程序，例如媒体播放器(Media Player)、浏览器(Browser)等，用于实现各种应用任务。实现本公开实施例提供的对象识别码处理方法或对象发布方法的程序可以包含在应用程序中。
在本公开实施例中，处理器601通过调用存储器602存储的程序或指令，具体的，可以是应用程序中存储的程序或指令，处理器601用于执行本公开实施例提供的对象识别码处理方法或对象发布方法各实施例的步骤。
本公开实施例提供的对象识别码处理方法或对象发布方法可以应用于处理器601中，或者由处理器601实现。处理器601可以是一种集成电路芯片，具有信号的处理能力。在实现过程中，上述方法的各步骤可以通过处理器601中的硬件的集成逻辑电路或者软件形式的指令完成。上述的处理器601可以是通用处理器、数字信号处理器(Digital Signal Processor，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现成可编程门阵列(Field Programmable Gate Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。
本公开实施例提供的对象识别码处理方法或对象发布方法的步骤可以直接体现为硬件译码处理器执行完成，或者用译码处理器中的硬件及软件单元组合执行完成。软件单元可以位于随机存储器，闪存、只读存储器，可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器602，处理器601读取存储器602中的信息，结合其硬件完成方法的步骤。
本公开实施例还提出一种非暂态计算机可读存储介质，所述非暂态计算机可读存储介质存储程序或指令，所述程序或指令使计算机执行如对象识别码处理方法或对象发布方法各实施例的步骤，为避免重复描述，在此不再赘述。
需要说明的是，在本文中，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者装置不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者装置所固有的要素。在没有更多限制的情况下，由语句“包括……”限定的要素，并不排除在包括该要素的过程、方法、物品或者装置中还存在另外的相同要素。
本领域的技术人员能够理解，尽管在此所述的一些实施例包括其它实施例 中所包括的某些特征而不是其它特征，但是不同实施例的特征的组合意味着处于本公开的范围之内并且形成不同的实施例。
本领域的技术人员能够理解，对各个实施例的描述都各有侧重，某个实施例中没有详述的部分，可以参见其他实施例的相关描述。
虽然结合附图描述了本公开的实施方式，但是本领域技术人员可以在不脱离本公开的精神和范围的情况下做出各种修改和变型，这样的修改和变型均落入由所附权利要求所限定的范围之内。

Claims (21)

  1. 一种对象识别码处理方法，包括：
    获取对象识别码以及与所述对象识别码关联的第一对象信息；
    基于所述第一对象信息，将所述对象识别码与目标市场的识别码库中至少一部分识别码进行聚合，得到聚合信息；
    将所述第一对象信息映射为所述目标市场的第二对象信息；
    在所述目标市场的识别码库中，存储所述对象识别码、所述聚合信息、所述第二对象信息之间的关联关系。
  2. 根据权利要求1所述的方法，其中，所述第一对象信息包括识别码标题；
    所述基于所述第一对象信息，将所述对象识别码与目标市场的识别码库中至少一部分识别码进行聚合包括：
    将所述识别码标题进行多次截断，得到多个子标题；
    基于所述多个子标题，将所述对象识别码与目标市场的识别码库中至少一部分识别码进行聚合。
  3. 根据权利要求2所述的方法，其中，所述基于所述多个子标题，将所述对象识别码与目标市场的识别码库中至少一部分识别码进行聚合包括：
    将每个所述子标题与所述目标市场的识别码库中至少一部分识别码的关键属性进行匹配；
    将所述对象识别码与匹配的关键属性所对应的识别码进行聚合。
  4. 根据权利要求3所述的方法，其中，所述将每个所述子标题与所述目标市场的识别码库中至少一部分识别码的关键属性进行匹配包括：
    预先将所述目标市场的识别码库中至少一部分识别码进行聚合，得到多个聚合集；
    确定每个所述聚合集所共有的关键属性；
    将每个所述子标题与所述多个聚合集各自共有的关键属性进行匹配。
  5. 根据权利要求1所述的方法，其中，所述第一对象信息包括第一识别码标题；
    所述基于所述第一对象信息，将所述对象识别码与目标市场的识别码库中至少一部分识别码进行聚合包括：
    针对所述目标市场的识别码库中至少一部分识别码的第二识别码标题，确定该第二识别码标题与所述第一识别码标题之间的相似距离；
    将所述对象识别码与目标市场的识别码库中至少一部分识别码进行相似距离聚合。
  6. 根据权利要求5所述的方法，其中，所述确定该第二识别码标题与所述第一识别码标题之间的相似距离包括：
    将该第二识别码标题和所述第一识别码标题输入实体对齐模型，通过所述实体对齐模型进行对齐，输出该第二识别码标题与所述第一识别码标题之间的相似距离。
  7. 根据权利要求6所述的方法，其中，所述实体对齐模型为基于同款对象对的识别码标题对训练孪生神经网络得到的模型。
  8. 根据权利要求1所述的方法，其中，所述第一对象信息包括第一对象类目和第一对象属性；所述将所述第一对象信息映射为所述目标市场的第二对象信息包括：
    将所述第一对象类目映射为所述目标市场的第二对象类目；
    将所述第一对象属性映射为所述目标市场的第二对象属性。
  9. 根据权利要求8所述的方法，其中，所述将所述第一对象类目映射为所述目标市场的第二对象类目包括：
    将所述第一对象类目和所述第一对象属性输入文本模型，通过所述文本模型对所述第一对象类目和所述第一对象属性进行文本分析，输出所述目标市场的第二对象类目。
  10. 根据权利要求9所述的方法，其中，所述第一对象信息还包括对象图片；
    所述将所述第一对象类目映射为所述目标市场的第二对象类目包括：
    将所述对象图片输入对象识别模型，通过所述对象识别模型识别所述对象图片中的对象，输出所述目标市场的第二对象类目。
  11. 根据权利要求10所述的方法，其中，所述将所述第一对象类目映射为所述目标市场的第二对象类目还包括：
    识别所述对象图片中的文本信息；
    将所述文本信息输入所述文本模型，通过所述文本模型对所述文本信息进行分析，输出所述目标市场的第二对象类目。
  12. 根据权利要求9所述的方法，其中，所述第一对象信息还包括识别码标题；
    所述将所述第一对象类目映射为所述目标市场的第二对象类目还包括：
    将所述识别码标题输入所述文本模型，通过所述文本模型对所述识别码标题进行文本分析，输出所述目标市场的第二对象类目。
  13. 根据权利要求8至12任一项所述的方法，其中，所述将所述第一对象类目映射为所述目标市场的第二对象类目后，所述方法还包括：
    基于第一规则和/或第二规则，修正所述第二对象类目；其中，所述第一规则为基于算法批量挖掘的规则；所述第二规则为基于行业专家积累的规则。
  14. 根据权利要求8所述的方法，其中，所述将所述第一对象类目映射为所述目标市场的第二对象类目包括：
    基于人工沉淀的渠道间的类目映射关系，将所述第一对象类目映射为所述目标市场的第二对象类目。
  15. 根据权利要求8所述的方法，其中，所述将所述第一对象类目映射为所述目标市场的第二对象类目后，所述方法还包括：
    基于用户反馈信息，修正所述第二对象类目。
  16. 一种对象发布方法，包括：
    响应对象发布请求，展示对象发布页面；
    获取对象识别码；
    基于所述对象识别码，从目标市场的识别码库中，查找所述对象识别码关联的聚合信息和第二对象信息；
    将所述第二对象信息回填所述对象发布页面，并在所述对象发布页面中回填所述聚合信息对应的至少一部分识别码所对应的对象信息；
    响应信息确认操作，将所述对象发布页面中对象信息进行发布。
  17. 根据权利要求16所述的方法，其中，所述方法还包括：
    若未查找到所述对象识别码关联的聚合信息和第二对象信息，则基于权利要求1至15任一项所述的对象识别码处理方法处理所述对象识别码，得到所述对象识别码关联的聚合信息和第二对象信息。
  18. 一种对象识别码处理装置，包括：
    获取单元，用于获取对象识别码以及与所述对象识别码关联的第一对象信息；
    聚合单元，用于基于所述第一对象信息，将所述对象识别码与目标市场的识别码库中至少一部分识别码进行聚合，得到聚合信息；
    映射单元，用于将所述第一对象信息映射为所述目标市场的第二对象信息；
    存储单元，用于在所述目标市场的识别码库中，存储所述对象识别码、所述聚合信息、所述第二对象信息之间的关联关系。
  19. 一种对象发布装置，包括：
    展示单元，用于响应对象发布请求，展示对象发布页面；
    获取单元，用于获取对象识别码；
    查找单元，用于基于所述对象识别码，从目标市场的识别码库中，查找所述对象识别码关联的聚合信息和第二对象信息；
    回填单元，用于将所述第二对象信息回填所述对象发布页面，并在所述对象发布页面中回填所述聚合信息对应的至少一部分识别码所对应的对象信息；
    发布单元，用于响应信息确认操作，将所述对象发布页面中对象信息进行发布。
  20. 一种电子设备，包括：处理器和存储器；
    所述处理器通过调用所述存储器存储的程序或指令，用于执行如权利要求1至17任一项所述方法的步骤。
  21. 一种非暂态计算机可读存储介质，其中，所述非暂态计算机可读存储介质 存储程序或指令，所述程序或指令使计算机执行如权利要求1至17任一项所述方法的步骤。
PCT/CN2022/072842 2021-01-27 2022-01-20 对象识别码处理方法、对象发布方法、装置、设备及介质 WO2022161231A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110113486.X 2021-01-27
CN202110113486.XA CN113298609A (zh) 2021-01-27 2021-01-27 对象识别码处理方法、对象发布方法、装置、设备及介质

Publications (1)

Publication Number Publication Date
WO2022161231A1 true WO2022161231A1 (zh) 2022-08-04

Family

ID=77318870

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/072842 WO2022161231A1 (zh) 2021-01-27 2022-01-20 对象识别码处理方法、对象发布方法、装置、设备及介质

Country Status (2)

Country Link
CN (1) CN113298609A (zh)
WO (1) WO2022161231A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113298609A (zh) * 2021-01-27 2021-08-24 阿里巴巴集团控股有限公司 对象识别码处理方法、对象发布方法、装置、设备及介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104657396A (zh) * 2013-11-25 2015-05-27 腾讯科技（深圳）有限公司 数据迁移方法及装置
WO2017162074A1 (zh) * 2016-03-25 2017-09-28 阿里巴巴集团控股有限公司 商品映射方法、装置及设备
US10163144B1 (en) * 2014-09-05 2018-12-25 Amazon Technologies, Inc. Extracting data from a catalog
CN109428904A (zh) * 2017-08-22 2019-03-05 阿里巴巴集团控股有限公司 业务对象的发布、获取方法、装置及系统
CN113298609A (zh) * 2021-01-27 2021-08-24 阿里巴巴集团控股有限公司 对象识别码处理方法、对象发布方法、装置、设备及介质

Family Cites Families (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8484097B1 (en) * 2011-03-31 2013-07-09 Amazon Technologies, Inc. Method, system, and computer readable medium for selection of catalog items for inclusion on a network page
CN103425658B (zh) * 2012-05-15 2017-06-09 阿里巴巴集团控股有限公司 商品信息的存储、搜索方法与移动终端和服务器
CN105005917A (zh) * 2015-07-07 2015-10-28 上海晶赞科技发展有限公司 一种通用的关联不同电商网站单品的方法
CN106919543A (zh) * 2015-12-24 2017-07-04 阿里巴巴集团控股有限公司 确定商品对象标题文本的方法及装置
CN106919619B (zh) * 2015-12-28 2021-09-07 阿里巴巴集团控股有限公司 一种商品聚类方法、装置及电子设备
CN107464162B (zh) * 2017-07-28 2022-12-30 腾讯科技（深圳）有限公司 商品关联方法、装置及计算机可读存储介质
CN107705066B (zh) * 2017-09-15 2022-01-07 唯品会(海南)电子商务有限公司 一种商品入库时信息录入方法及电子设备
CN108171586A (zh) * 2018-01-23 2018-06-15 北京值得买科技股份有限公司 一种商品聚类方法及装置
CN110245997A (zh) * 2018-03-08 2019-09-17 阿里巴巴集团控股有限公司 获取商品信息的方法、装置、系统、设备和存储介质
CN110688560A (zh) * 2018-06-19 2020-01-14 阿里巴巴集团控股有限公司 一种信息推荐的方法和装置
CN111178999A (zh) * 2018-11-09 2020-05-19 商派软件有限公司 一种基于已知商品获得商品信息的共享方法
CN110287419A (zh) * 2019-06-28 2019-09-27 重庆金窝窝网络科技有限公司 一种商品推荐方法、装置、服务器及存储介质
CN111210492A (zh) * 2019-12-30 2020-05-29 广州市创乐信息技术有限公司 一种图片处理的方法及装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104657396A (zh) * 2013-11-25 2015-05-27 腾讯科技（深圳）有限公司 数据迁移方法及装置
US10163144B1 (en) * 2014-09-05 2018-12-25 Amazon Technologies, Inc. Extracting data from a catalog
WO2017162074A1 (zh) * 2016-03-25 2017-09-28 阿里巴巴集团控股有限公司 商品映射方法、装置及设备
CN109428904A (zh) * 2017-08-22 2019-03-05 阿里巴巴集团控股有限公司 业务对象的发布、获取方法、装置及系统
CN113298609A (zh) * 2021-01-27 2021-08-24 阿里巴巴集团控股有限公司 对象识别码处理方法、对象发布方法、装置、设备及介质

Also Published As

Publication number Publication date
CN113298609A (zh) 2021-08-24

Similar Documents

Publication Publication Date Title
US20230192947A1 (en) System and method allowing social fashion selection in an electronic marketplace
US9892447B2 (en) Performing image searches in a network-based publication system
WO2020024555A1 (zh) 一种商品信息推送方法、相关装置及存储介质
US10169803B2 (en) Color based social networking recommendations
TWI640878B (zh) Query word fusion method, product information publishing method, search method and system
US10210262B2 (en) Systems and methods to identify a filter set in a query comprised of keywords
US11392963B2 (en) Determining and using brand information in electronic commerce
CN103970761A (zh) 一种商品数据搜索方法及装置
US11734736B2 (en) Building containers of uncategorized items
US20140129373A1 (en) Item recommendations based on true fit determination
US11930067B2 (en) Commerce applications: data handshake between an on-line service and a third-party partner
US20130085900A1 (en) Enhancing the search experience in a networked publication system by improved search and listing process
US20150066674A1 (en) Systems and methods to identify and associate related items
US20240037545A1 (en) Systems and methods for associating a user's shopping experiences across multiple channels
KR20200041173A (ko) 다수의 오픈 마켓에의 상품 정보 일괄 등록 시스템 및 등록 방법
US20190205769A1 (en) Data Processing Method, Apparatus, Device and Computer Readable Storage Media
WO2022161231A1 (zh) 对象识别码处理方法、对象发布方法、装置、设备及介质
US20160078514A1 (en) Trade matching system of consumer and supplier and method thereof
US8315917B2 (en) Catalog generation based on divergent listings
CN114219589B (zh) 虚拟实体对象的生成和页面显示方法、装置和电子设备
US11734242B1 (en) Architecture for resolution of inconsistent item identifiers in a global catalog
CN111753181A (zh) 基于图像的搜索方法、装置、服务器、客户端及介质
CN110751496A (zh) 一种商品价格检测方法和装置
CA3054785A1 (en) Method for searching for electronic transaction certificate, and electronic transaction terminal

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22745106

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 22745106

Country of ref document: EP

Kind code of ref document: A1