Label management method and device, computer equipment and storage medium
By using BI business system and Bloom filter in the e-commerce system, product information is converted into binary objects and a relationship table between product labels and binary objects is generated, which solves the problem of product labels and detailed storage space requirements, and achieves efficient product query and storage space saving effect.
Patent Information
- Application Number
- CN202411773011.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-04
- Publication Date
- 2025-04-15
AI Technical Summary
In the e-commerce industry, although the separate storage of product labels and product details reduces space usage, there is still a large demand for data storage space.
By obtaining the target product set, input it into the BI business system for data processing tasks, generating a BI product label table, and using the Bloom filter to convert product information into binary objects, generating a relationship table of target product labels and binary objects, so as to quickly match and report the corresponding tags in response to product query instructions.
This method effectively saves storage space, improves product query efficiency, and reduces system resource consumption.
Smart Images
Figure CN119830938A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of computer technology, and in particular to a tag management method, apparatus, computer equipment and storage medium. Background Art
[0002] In the e-commerce industry, it is often necessary to label products, which are generally stored in the system in the form of product labels + product details. In this way, each product label will carry product details, which will take up a lot of space resources.
[0003] In the prior art, product labels and product details are stored separately, and a correspondence between product labels and product details is established to facilitate subsequent calls to the product details corresponding to any one or several product labels. However, the data storage space required to store product labels and product details in this way is still very large. Summary of the invention
[0004] Based on this, it is necessary to provide a label management method, device, computer equipment and storage medium that can save storage space in response to the above technical problems.
[0005] In order to solve the above technical problems, in a first aspect, a tag management method is provided, the method comprising:
[0006] Obtain a target commodity set, which includes multiple target commodities, input the multiple target commodity information into the BI business system to perform a data processing task, and generate a BI commodity label table; the BI commodity label table includes multiple target commodity information and multiple target commodity labels corresponding to the multiple target commodity information;
[0007] The BI product label table is processed through the Bloom filter, which converts multiple target product information into binary objects and generates a relationship table between target product labels and binary objects;
[0008] In response to receiving a product query instruction sent by the system, determining whether the product corresponding to the product query instruction matches any target product in the target product set;
[0009] If the product corresponding to the product query instruction matches the target product in the target product set, the target product label and the binary object relationship table corresponding to the target product matching the product corresponding to the product query instruction are reported to the system.
[0010] In one embodiment, before obtaining the target product set, the method further includes:
[0011] Get the label frame and label frame properties;
[0012] In response to receiving a supplementary product set, the supplementary product set includes a plurality of supplementary products, obtaining product attributes corresponding to the plurality of supplementary products;
[0013] Calculate the similarity between the product attributes of multiple supplementary products and the label frame attributes, and use the label frame whose similarity between the label frame attribute and the supplementary product attribute is greater than a threshold as the supplementary label frame;
[0014] Extracting supplementary product feature information, generating supplementary product tags based on the supplementary product feature information and a supplementary tag framework, associating the supplementary product tags with the supplementary products, and generating a supplementary product set.
[0015] In one embodiment, extracting the supplementary product feature information and generating a supplementary product tag according to the supplementary product feature information of the supplementary product includes:
[0016] Acquire image information, text information, and location information of the supplementary product, and perform multimodal fusion on the image information and text information of the supplementary product through modal joint learning to obtain feature information of the supplementary product, wherein the feature information of the supplementary product includes features of the supplementary product in multiple modalities;
[0017] Inputting the supplementary label frame into the deep learning model, and outputting feature information of the supplementary label frame;
[0018] The feature information of the supplementary product is matched with the feature information of the supplementary label frame. When the feature information of the supplementary product matches the feature information of the supplementary label frame, the feature information of the supplementary product is embedded in the supplementary label frame to form a supplementary product label.
[0019] In one embodiment, the method further comprises:
[0020] In response to the BI business system receiving the target product set, the BI business system obtains target product label attribute information corresponding to the target product set;
[0021] Determine whether the attribute information of the target product label information is consistent with the preset attribute information;
[0022] If the attribute information of the target product label information is consistent with the preset attribute information, the target product label attribute information is displayed in the BI product label table in the first field corresponding to the preset attribute information;
[0023] If the attribute information of the target product label information is inconsistent with the preset attribute information, a second field matching the target product label attribute information is queried from the preset field table, and the second field is displayed in the BI product label table.
[0024] In one embodiment, the method further comprises:
[0025] In response to receiving a product query instruction sent by the system, determining whether the product corresponding to the product query instruction matches any target product in the target product set;
[0026] If the product corresponding to the product query instruction does not match any target product in the target product set;
[0027] Acquire product information of a product corresponding to the product query instruction from the supplementary product set, convert the product information of the product corresponding to the product query instruction acquired from the supplementary product set into a binary object, and obtain a supplementary object;
[0028] Importing the supplementary object into the target product tag and binary object relationship table;
[0029] The target product label and the binary object corresponding to the target product matching the product corresponding to the product query instruction are reported to the system.
[0030] In one of the embodiments, the BI system further includes a counting Bloom filter, and the method includes:
[0031] In response to reporting the target product label and the binary object relationship table to the system, calculating the number of products corresponding to each target product label reported to the system;
[0032] The counting Bloom filter obtains the number of commodities corresponding to each target commodity tag of the reporting system, and adds the value of the number of commodities corresponding to each target commodity tag of the reporting system to the initial count value of the counting Bloom filter to update the count value of the counting Bloom filter.
[0033] In one embodiment, the method further comprises:
[0034] Splitting the counting Bloom filter into a preset number of sub-Bloom filters;
[0035] Select any sub-Bloom filter from the multiple sub-Bloom filters as the target Bloom filter for the current application;
[0036] When the target Bloom filter satisfies the condition that causes the target Bloom filter to have a false alarm, the unapplied sub-Bloom filter is called as the target Bloom filter.
[0037] In order to solve the above technical problem, in a second aspect, a label management device is provided, the device comprising:
[0038] An acquisition module is used to acquire a target commodity set, where the target commodity set includes multiple target commodities;
[0039] The business module is used to input multiple target product information into the BI business system to perform a data processing task and generate a BI product label table; the BI product label table includes multiple target product information and multiple target product labels corresponding to the multiple target product information;
[0040] The conversion module is used to process the BI product label table through a Bloom filter. The Bloom filter converts multiple target product information into binary objects and generates a relationship table between target product labels and binary objects.
[0041] A judgment module, configured to judge, in response to receiving a product query instruction sent by the system, whether the product corresponding to the product query instruction matches any target product in the target product set;
[0042] The reporting module is used to match the product corresponding to the product query instruction with the target product in the target product set, and report the target product label and the binary object relationship table corresponding to the target product matching the product corresponding to the product query instruction to the system.
[0043] In order to solve the above technical problem, a third aspect provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program: the processor implements the steps of the above first aspect method when executing the computer program.
[0044] In order to solve the above technical problem, in a fourth aspect, the present application provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method described in the first aspect are implemented.
[0045] Different from the prior art, the present application provides a label management method, which obtains a target product set, which includes multiple target products, inputs multiple target product information into the BI business system to perform a data task, and generates a BI product label table; the BI product label table includes multiple target product information and multiple target product labels corresponding to the multiple target product information; the BI product label table is processed by a Bloom filter, the Bloom filter converts the multiple target product information into binary objects, and generates a target product label and binary object relationship table; in response to receiving a product query instruction sent by the system, it is determined whether the product corresponding to the product query instruction matches any target product in the target product set; if so, the target product label corresponding to the target product that matches the product corresponding to the product query instruction and the binary object relationship table are reported to the system. In this way, the use of the BI business system in combination with the Bloom filter to generate a target product label and a binary object relationship table can greatly save space resources. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] Figure 1 A flowchart of a tag management method in one embodiment;
[0047] Figure 2 A flowchart of a tag management method according to another embodiment;
[0048] Figure 3 is a structural block diagram of a label management device in one embodiment;
[0049] Figure 4 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION
[0050] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0051] In one embodiment, Figure 1 As shown, a tag management method is provided, and the method specifically includes the following steps:
[0052] Step 101, obtain a target product set, which includes multiple target products, input the multiple target product information into the BI business system to perform a data task, and generate a BI product label table; the BI product label table includes multiple target product information and multiple target product labels corresponding to the multiple target product information.
[0053] Specifically, the target product set can be obtained by packet capture, and the target product set can be saved in local memory or database for subsequent data processing. Each product in the target product set can be analyzed by data analysis tools such as Pandas to obtain product information of multiple target products in the target product set. After obtaining the target product set, the information of multiple target products in the target product set can be input into the BI business system to perform data tasks and generate a BI product label table. Among them, BI business system refers to a business intelligence system, which has powerful tools and technologies for collecting, analyzing and displaying data, assisting users in data analysis, revealing trends, correlations and patterns, and then supporting corporate decision-making and business optimization.
[0054] Specifically, after the BI business system receives the target product set, the BI business system obtains the attribute information of the target tag and determines whether the attribute information of the target tag is consistent with the preset attribute information; since the tag meanings of different product tags may be the same, a preset field table can be set, and the field table stores the fixed format of the fields of the product tags. When the attribute information of the target tag information is consistent with the preset attribute information, the target tag attribute information is displayed in the BI product tag table with the first field corresponding to the preset attribute information. Alternatively, if the attribute information of the target tag information is inconsistent with the preset attribute information, a second field matching the target tag attribute information is queried from the preset field table, and the second field is displayed in the BI product tag table.
[0055] In one embodiment, before obtaining the target product set, the method further includes: obtaining a label frame and label frame attributes; in response to receiving a supplementary product set, the supplementary product set includes multiple supplementary products, obtaining product attributes corresponding to the multiple supplementary products; calculating the similarity between the product attributes of the multiple supplementary products and the label frame attributes, and taking the label frame whose similarity between the label frame attributes and the supplementary product attributes is greater than a threshold as a supplementary label frame; extracting supplementary product feature information, generating a supplementary product tag based on the supplementary product feature information and the supplementary label frame of the supplementary product, associating the supplementary product tag with the supplementary product, and generating a supplementary product set.
[0056] Specifically, a label pool may be set up, the label pool including multiple label frames, and a label frame whose label frame attributes have a similarity greater than a threshold with the supplementary product attributes of the supplementary products in the supplementary product set may be selected from the label pool as a supplementary label frame, and feature information of the supplementary products may be extracted, and the feature information of the supplementary products may be embedded in the supplementary label frame to form a supplementary product label, and the obtained supplementary product label may be associated with the corresponding supplementary product to generate a supplementary product set.
[0057] Among them, the image information, text information, and location information of the supplementary products can be extracted, and the image information and text information of the supplementary products can be multimodally fused through modal joint learning to obtain the feature information of the supplementary products. Among them, the feature information of multiple supplementary products includes the target product features of multiple modalities; the supplementary label frame is input into the deep learning model, and the feature information of the supplementary label frame is output. In this embodiment, the features of the image, text, and location information of the supplementary products after multimodal fusion can be used as the input of the deep learning model. This method can use image, text, and location information at the same time to provide more comprehensive and accurate information, thereby improving the matching accuracy.
[0058] The feature information of the supplementary product is matched with the feature information of the supplementary label frame. Optionally, the similarity between the feature information of the supplementary product and the feature information of the label frame can be calculated to determine whether the feature information of the supplementary product matches the feature information of the supplementary label frame. Furthermore, feature vectors can be extracted from the feature information of the supplementary product and the feature information of the supplementary label frame respectively. For ease of explanation, the feature vector extracted from the feature information of the supplementary product can be referred to as the first feature vector. The feature vector extracted from the feature information of the supplementary label frame is referred to as the second feature vector.
[0059] The spatial distance between the first eigenvector and the second eigenvector may be calculated, and the calculated spatial distance may represent the degree of similarity between the first eigenvector and the second eigenvector.
[0060] In this embodiment, if the spatial distance between the first feature vector and the second feature vector is larger, it means that the feature information of the supplementary product and the feature information of the supplementary label frame are less similar. Conversely, if the spatial distance between the first feature vector and the second feature vector is smaller, it means that the feature information of the supplementary product and the feature information of the supplementary label frame are more similar. Optionally, the spatial distance can be Euclidean distance, Chebyshev distance, Minkowski distance, Manhattan distance or cosine distance, etc., which is not limited here.
[0061] In another embodiment, the similarity calculation module may be used to determine the similarity between the first feature vector and the second feature vector, that is, the first feature vector and the second feature vector may be input into the similarity calculation model to obtain the similarity between the first feature vector and the second feature vector.
[0062] When the feature information of the supplementary product matches the feature information of the supplementary label frame, the feature information of the supplementary product is embedded in the supplementary label frame to form a supplementary product label.
[0063] Step 102, the BI product label table is processed by a Bloom filter, the Bloom filter converts multiple target product information into binary objects, and generates a relationship table between target product labels and binary objects.
[0064] After obtaining the BI product information table, the target product information and multiple target product tags corresponding to the multiple target product information are obtained from the BI product information table, and the product information corresponding to the multiple target products is processed, and the product information originally in the form of a string is converted into a binary object, and the binary objects after the product information corresponding to the multiple target products is converted are associated with the product tags corresponding to the target products, and a target product tag and binary object relationship table is generated. The Bloom filter is composed of a binary array and a hash algorithm. The string entering the Bloom filter will be mapped to each binary array through the hash algorithm. It is precisely because of the use of a binary array that an element corresponds to an array position (bit) through a hash function. If a URL (64B) is actually stored, the space ratio is about 64*8:1, that is, a 512:1 bit array, which enables BloomFilter to save a lot of space. In terms of space efficiency, it is extremely efficient. It can be understood that if the product details need to be obtained later, it can be deserialized to convert the binary object into a string.
[0065] And BloomFilter (binary object generated by BloomFilter) does not need to be traversed one by one like a linked list query. Instead, it is like an array, similar to directly taking the index to find the required result. In terms of time and space efficiency, with a very low misrecognition rate, it is far superior to other algorithms in all aspects.
[0066] In a specific implementation, the multiple target product information can be converted into binary objects by using a Bloom filter. The Bloom filter here can be a Hive-based Bloom filter, which is used to determine whether an input string may exist in the Bloom filter.
[0067] Exemplarily, a Bloom filter index is created in Hive by using Hive's BloomFilter UDF function and create bloomfilter on syntax. After the index is created, the BLOOM_FILTER() function can be used to query whether a specified value exists, and all rows containing 'value' in the matching Bloom filter are returned, where the 'value' row includes the value to be queried.
[0068] In one embodiment, the system further includes a counting Bloom filter. In response to reporting the target product tag and the binary object relationship table to the system, the number of products corresponding to each target product tag reported to the system is calculated; the counting Bloom filter obtains the number of products corresponding to each target product tag reported to the system, and adds the number of products corresponding to each target product tag reported to the system to the initial count value of the counting Bloom filter to update the count value of the Bloom filter.
[0069] Among them, the counting Bloom filter is to add a counter to each bit on the basis of the Bloom filter, and the counter is represented by an n-bit BitSet. Each time an element is added to the BitSet of the counting Bloom filter, the counter of the corresponding digit is also incremented by 1. When the number of occurrences of a certain element is to be counted, the minimum value of the counters corresponding to all digits of the element after being mapped by the hash function is taken as the number of occurrences of the element in this counting Bloom filter. Therefore, according to the above principle, after the number of goods corresponding to each target product label of the reporting system is put into the counting Bloom filter, the number of reports corresponding to the number of goods corresponding to each target product label of the reporting system can be obtained according to the minimum value of the count values of all counters corresponding to the number of goods corresponding to the target product label of the reporting system in the counting Bloom filter. Among them, BitSet is equivalent to a bool array that can only store binary, that is, 0 and 1.
[0070] Bloom filter is a fast and efficient data structure used to determine whether an element exists in a set. However, due to its specific design, its fault tolerance and false alarm rate are limited. Bloom filter maps elements to different positions of the bit array through hash function. The default size of the bit array of Bloom filter is fixed. When the number of elements exceeds the capacity of the bit array, it may cause more hash collisions, thereby reducing fault tolerance. In order to solve this problem, a dynamic expansion mechanism can be implemented. For example, the counting Bloom filter can be split into a preset number of sub-Bloom filters; any sub-Bloom filter is selected from the multiple sub-Bloom filters as the target Bloom filter currently applied; when the target Bloom filter meets the conditions that cause the target Bloom filter to have a false alarm, the unapplied sub-Bloom filter is called as the target Bloom filter. By setting multiple sub-Bloom filters, multiple hash functions and dynamic expansion mechanisms are implemented, and the fault tolerance and false alarm rate of the Bloom filter can be improved.
[0071] Step 103, in response to receiving the product query instruction sent by the system, determining whether the product corresponding to the product query instruction matches any target product in the target product set.
[0072] Step 104 , if the product corresponding to the product query instruction matches any target product in the target product set, the target product tag and the binary object relationship table corresponding to the target product matching the product corresponding to the product query instruction are reported to the system.
[0073] Specifically, in response to receiving a product query instruction sent by the system, the product query instruction is parsed to obtain the product that needs to be queried corresponding to the product query instruction, and then the target product set is traversed to determine whether the product corresponding to the product query instruction matches any target product in the target product set, that is, to determine whether the target product set includes the product that needs to be queried corresponding to the product query instruction. When the product corresponding to the product query instruction matches the target product in the target product set, the target product label and binary object corresponding to the target product that matches the product corresponding to the product query instruction are reported to the system.
[0074] In one embodiment, if the product corresponding to the product query instruction does not match any target product in the target product set; obtain the product information of the product corresponding to the product query instruction from the supplementary product set, convert the product information of the product corresponding to the product query instruction obtained from the supplementary product into a binary object to obtain a supplementary object; import the supplementary object into the target product label and binary object relationship table; report the target product label and binary object corresponding to the target product matching the product corresponding to the product query instruction to the system.
[0075] See also Figure 2 In a specific implementation, the label management process is as follows: The label management method of the present application can be applied to a label management system. The label management process includes two links. First, in the first link, a sub-plan threshold selection is created to perform a custom label, that is, to obtain a label frame from the label pool, and then obtain the label frame attribute corresponding to the label frame. In response to the label management system obtaining the externally transmitted product, a product label is generated according to the product feature information corresponding to the product (product), which refers to the generation of a supplementary product label. Specifically, the product feature information corresponding to the supplementary product is extracted, and a label frame whose label frame attribute and the product attribute of the supplementary product are selected from the label pool as a supplementary label frame whose similarity is greater than a threshold value, and the product feature information of the supplementary product is embedded in the supplementary label frame to form a supplementary product label, and the supplementary product label is stored in the database (mysql). Among them, the supplementary product label is stored in the database in the form of (a report label information table and a report label product table). The report label information table includes the label ID of the product, and the report label product table includes product details. The supplementary product label is converted into a binary object to obtain a report label product BF table, and the table is stored in the database.
[0076] In the second link, the target product set is obtained, and the product information of multiple target products included in the target product set is input into the BI business system to perform data processing tasks, a BI product label table is generated, and it is determined whether the generated BI product label table contains invalid labels. If it contains invalid labels, the invalid labels are removed from the BI product label table, and the BI product label table after the removal operation is input into the Bloom filter, the product information originally in the form of strings is converted into binary objects, and the binary objects converted from the product information corresponding to the multiple target products are associated with the product labels corresponding to the target products, a target product label and binary object relationship table (label-BloomFilter table) is generated and stored in the database. When the label management system needs to provide a product label query service in response to receiving a product query instruction, it is selected to combine the target product label and binary object relationship table, or the target product label and binary object relationship table with the reported label product BF table (binary relationship table of supplementary product labels and supplementary products) to report to the system according to actual conditions.
[0077] It should be understood that although Figure 1-Figure 2 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1-Figure 2 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.
[0078] In one embodiment, Figure 3 As shown, a label management device is provided, including: an acquisition module 30, a business module 31, a conversion module 32, a judgment module 33 and a reporting module 34, wherein:
[0079] An acquisition module 30 is used to acquire a target commodity set, where the target commodity set includes a plurality of target commodities;
[0080] Business module 31, used to input multiple target commodity information into the BI business system to perform a data processing task and generate a BI commodity label table; the BI commodity label table includes multiple target commodity information and multiple target commodity labels corresponding to the multiple target commodity information;
[0081] The conversion module 32 is used to process the BI product label table through a Bloom filter, the Bloom filter converts multiple target product information into binary objects, and generates a relationship table between target product labels and binary objects;
[0082] A judgment module 33 is used to judge whether the product corresponding to the product query instruction matches any target product in the target product set in response to receiving the product query instruction sent by the system;
[0083] The reporting module 34 is used to match the product corresponding to the product query instruction with the target product in the target product set, and report the target product label and the binary object relationship table corresponding to the target product matching the product corresponding to the product query instruction to the system.
[0084] In one embodiment, another implementation of the tag management method that can be implemented by the above device includes the following specific steps:
[0085] Before obtaining the target product set, the following steps are also included:
[0086] Get the label frame and label frame properties;
[0087] In response to receiving a supplementary product set, the supplementary product set includes a plurality of supplementary products, obtaining product attributes corresponding to the plurality of supplementary products;
[0088] Calculate the similarity between the product attributes of multiple supplementary products and the label frame attributes, and use the label frame whose similarity between the label frame attribute and the supplementary product attribute is greater than a threshold as the supplementary label frame;
[0089] Extracting supplementary product feature information, generating supplementary product tags based on the supplementary product feature information and a supplementary tag framework, associating the supplementary product tags with the supplementary products, and generating a supplementary product set.
[0090] In one embodiment, another implementation of the tag management method that can be implemented by the above device includes the following specific steps:
[0091] Extracting the supplementary product feature information and generating a supplementary product label according to the supplementary product feature information of the supplementary product includes:
[0092] Acquire image information, text information, and location information of the supplementary product, and perform multimodal fusion on the image information and text information of the supplementary product through modal joint learning to obtain feature information of the supplementary product, wherein the feature information of the supplementary product includes features of the supplementary product in multiple modalities;
[0093] Inputting the supplementary label frame into the deep learning model, and outputting feature information of the supplementary label frame;
[0094] The feature information of the supplementary product is matched with the feature information of the supplementary label frame. When the feature information of the supplementary product matches the feature information of the supplementary label frame, the feature information of the supplementary product is embedded in the supplementary label frame to form a supplementary product label.
[0095] In one embodiment, another implementation of the tag management method that can be implemented by the above device includes the following specific steps:
[0096] The method also includes:
[0097] In response to the BI business system receiving the target product set, the BI business system obtains target product label attribute information corresponding to the target product set;
[0098] Determine whether the attribute information of the target product label information is consistent with the preset attribute information;
[0099] If the attribute information of the target product label information is consistent with the preset attribute information, the target product label attribute information is displayed in the BI product label table in the first field corresponding to the preset attribute information;
[0100] If the attribute information of the target product label information is inconsistent with the preset attribute information, a second field matching the target product label attribute information is queried from the preset field table, and the second field is displayed in the BI product label table.
[0101] In one embodiment, another implementation of the tag management method that can be implemented by the above device includes the following specific steps:
[0102] The method also includes:
[0103] In response to receiving a product query instruction sent by the system, determining whether the product corresponding to the product query instruction matches any target product in the target product set;
[0104] If the product corresponding to the product query instruction does not match any target product in the target product set;
[0105] Acquire product information of a product corresponding to the product query instruction from the supplementary product set, convert the product information of the product corresponding to the product query instruction acquired from the supplementary product set into a binary object, and obtain a supplementary object;
[0106] Importing the supplementary object into the target product tag and binary object relationship table;
[0107] The target product label and the binary object corresponding to the target product matching the product corresponding to the product query instruction are reported to the system.
[0108] In one embodiment, another implementation of the tag management method that can be implemented by the above device includes the following specific steps:
[0109] The BI system also includes a counting Bloom filter, and the method includes:
[0110] In response to reporting the target product label and the binary object relationship table to the system, calculating the number of products corresponding to each target product label reported to the system;
[0111] The counting Bloom filter obtains the number of commodities corresponding to each target commodity tag of the reporting system, and adds the value of the number of commodities corresponding to each target commodity tag of the reporting system to the initial count value of the counting Bloom filter to update the count value of the counting Bloom filter.
[0112] In one embodiment, another implementation of the tag management method that can be implemented by the above device includes the following specific steps:
[0113] The method also includes:
[0114] Splitting the counting Bloom filter into a preset number of sub-Bloom filters;
[0115] Select any sub-Bloom filter from the multiple sub-Bloom filters as the target Bloom filter for the current application;
[0116] When the target Bloom filter satisfies the condition that causes the target Bloom filter to have a false alarm, the unapplied sub-Bloom filter is called as the target Bloom filter.
[0117] For the specific definition of the label management device, please refer to the definition of the label management method above, which will not be repeated here. Each module in the above-mentioned label management device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.
[0118] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 4As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a label management method is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.
[0119] Those skilled in the art will understand that Figure 4 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
[0120] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:
[0121] Step 101, obtain a target product set, which includes multiple target products, input the multiple target product information into the BI business system to perform a data task, and generate a BI product label table; the BI product label table includes multiple target product information and multiple target product labels corresponding to the multiple target product information.
[0122] Step 102, the BI product label table is processed by a Bloom filter, the Bloom filter converts multiple target product information into binary objects, and generates a relationship table between target product labels and binary objects.
[0123] Step 103, in response to receiving the product query instruction sent by the system, determining whether the product corresponding to the product query instruction matches any target product in the target product set.
[0124] Step 104 , if the product corresponding to the product query instruction matches any target product in the target product set, the target product tag and the binary object relationship table corresponding to the target product matching the product corresponding to the product query instruction are reported to the system.
[0125] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:
[0126] Before obtaining the target product set, the following steps are also included:
[0127] Get the label frame and label frame properties;
[0128] In response to receiving a supplementary product set, the supplementary product set includes a plurality of supplementary products, obtaining product attributes corresponding to the plurality of supplementary products;
[0129] Calculate the similarity between the product attributes of multiple supplementary products and the label frame attributes, and use the label frame whose similarity between the label frame attribute and the supplementary product attribute is greater than a threshold as the supplementary label frame;
[0130] Extracting supplementary product feature information, generating supplementary product tags based on the supplementary product feature information and a supplementary tag framework, associating the supplementary product tags with the supplementary products, and generating a supplementary product set.
[0131] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:
[0132] Extracting the supplementary product feature information and generating a supplementary product label according to the supplementary product feature information of the supplementary product includes:
[0133] Acquire image information, text information, and location information of the supplementary product, and perform multimodal fusion on the image information and text information of the supplementary product through modal joint learning to obtain feature information of the supplementary product, wherein the feature information of the supplementary product includes features of the supplementary product in multiple modalities;
[0134] Inputting the supplementary label frame into the deep learning model, and outputting feature information of the supplementary label frame;
[0135] The feature information of the supplementary product is matched with the feature information of the supplementary label frame. When the feature information of the supplementary product matches the feature information of the supplementary label frame, the feature information of the supplementary product is embedded in the supplementary label frame to form a supplementary product label.
[0136] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:
[0137] The method also includes:
[0138] In response to the BI business system receiving the target product set, the BI business system obtains target product label attribute information corresponding to the target product set;
[0139] Determine whether the attribute information of the target product label information is consistent with the preset attribute information;
[0140] If the attribute information of the target product label information is consistent with the preset attribute information, the target product label attribute information is displayed in the BI product label table in the first field corresponding to the preset attribute information;
[0141] If the attribute information of the target product label information is inconsistent with the preset attribute information, a second field matching the target product label attribute information is queried from the preset field table, and the second field is displayed in the BI product label table.
[0142] In one embodiment, when the processor executes the computer program, the following steps are also implemented:
[0143] The method also includes:
[0144] In response to receiving a product query instruction sent by the system, determining whether the product corresponding to the product query instruction matches any target product in the target product set;
[0145] If the product corresponding to the product query instruction does not match any target product in the target product set;
[0146] Acquire product information of a product corresponding to the product query instruction from the supplementary product set, convert the product information of the product corresponding to the product query instruction acquired from the supplementary product set into a binary object, and obtain a supplementary object;
[0147] Importing the supplementary object into the target product tag and binary object relationship table;
[0148] The target product label and the binary object corresponding to the target product matching the product corresponding to the product query instruction are reported to the system.
[0149] In one embodiment, when the processor executes the computer program, the following steps are also implemented:
[0150] The BI system also includes a counting Bloom filter, and the method includes:
[0151] In response to reporting the target product label and the binary object relationship table to the system, calculating the number of products corresponding to each target product label reported to the system;
[0152] The counting Bloom filter obtains the number of commodities corresponding to each target commodity tag of the reporting system, and adds the value of the number of commodities corresponding to each target commodity tag of the reporting system to the initial count value of the counting Bloom filter to update the count value of the counting Bloom filter.
[0153] In one embodiment, when the processor executes the computer program, the following steps are also implemented:
[0154] The method also includes:
[0155] Splitting the counting Bloom filter into a preset number of sub-Bloom filters;
[0156] Select any sub-Bloom filter from the multiple sub-Bloom filters as the target Bloom filter for the current application;
[0157] When the target Bloom filter satisfies the condition that causes the target Bloom filter to have a false alarm, the unapplied sub-Bloom filter is called as the target Bloom filter.
[0158] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:
[0159] Step 101, obtain a target product set, which includes multiple target products, input the multiple target product information into the BI business system to perform a data task, and generate a BI product label table; the BI product label table includes multiple target product information and multiple target product labels corresponding to the multiple target product information.
[0160] Step 102, the BI product label table is processed by a Bloom filter, the Bloom filter converts multiple target product information into binary objects, and generates a relationship table between target product labels and binary objects.
[0161] Step 103, in response to receiving the product query instruction sent by the system, determining whether the product corresponding to the product query instruction matches any target product in the target product set.
[0162] Step 104 , if the product corresponding to the product query instruction matches any target product in the target product set, the target product tag and the binary object relationship table corresponding to the target product matching the product corresponding to the product query instruction are reported to the system.
[0163] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:
[0164] Before obtaining the target product set, the following steps are also included:
[0165] Get the label frame and label frame properties;
[0166] In response to receiving a supplementary product set, the supplementary product set includes a plurality of supplementary products, obtaining product attributes corresponding to the plurality of supplementary products;
[0167] Calculate the similarity between the product attributes of multiple supplementary products and the label frame attributes, and use the label frame whose similarity between the label frame attribute and the supplementary product attribute is greater than a threshold as the supplementary label frame;
[0168] Extracting supplementary product feature information, generating supplementary product tags based on the supplementary product feature information and a supplementary tag framework, associating the supplementary product tags with the supplementary products, and generating a supplementary product set.
[0169] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:
[0170] Extracting the supplementary product feature information and generating a supplementary product label according to the supplementary product feature information of the supplementary product includes:
[0171] Acquire image information, text information, and location information of the supplementary product, and perform multimodal fusion on the image information and text information of the supplementary product through modal joint learning to obtain feature information of the supplementary product, wherein the feature information of the supplementary product includes features of the supplementary product in multiple modalities;
[0172] Inputting the supplementary label frame into the deep learning model, and outputting feature information of the supplementary label frame;
[0173] The feature information of the supplementary product is matched with the feature information of the supplementary label frame. When the feature information of the supplementary product matches the feature information of the supplementary label frame, the feature information of the supplementary product is embedded in the supplementary label frame to form a supplementary product label.
[0174] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:
[0175] The method also includes:
[0176] In response to the BI business system receiving the target product set, the BI business system obtains target product label attribute information corresponding to the target product set;
[0177] Determine whether the attribute information of the target product label information is consistent with the preset attribute information;
[0178] If the attribute information of the target product label information is consistent with the preset attribute information, the target product label attribute information is displayed in the BI product label table in the first field corresponding to the preset attribute information;
[0179] If the attribute information of the target product label information is inconsistent with the preset attribute information, a second field matching the target product label attribute information is queried from the preset field table, and the second field is displayed in the BI product label table.
[0180] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:
[0181] The method also includes:
[0182] In response to receiving a product query instruction sent by the system, determining whether the product corresponding to the product query instruction matches any target product in the target product set;
[0183] If the product corresponding to the product query instruction does not match any target product in the target product set;
[0184] Acquire product information of a product corresponding to the product query instruction from the supplementary product set, convert the product information of the product corresponding to the product query instruction acquired from the supplementary product set into a binary object, and obtain a supplementary object;
[0185] Importing the supplementary object into the target product tag and binary object relationship table;
[0186] The target product label and the binary object corresponding to the target product matching the product corresponding to the product query instruction are reported to the system.
[0187] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:
[0188] The BI system also includes a counting Bloom filter, and the method includes:
[0189] In response to reporting the target product label and the binary object relationship table to the system, calculating the number of products corresponding to each target product label reported to the system;
[0190] The counting Bloom filter obtains the number of commodities corresponding to each target commodity tag of the reporting system, and adds the value of the number of commodities corresponding to each target commodity tag of the reporting system to the initial count value of the counting Bloom filter to update the count value of the counting Bloom filter.
[0191] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:
[0192] The method also includes:
[0193] Splitting the counting Bloom filter into a preset number of sub-Bloom filters;
[0194] Select any sub-Bloom filter from the multiple sub-Bloom filters as the target Bloom filter for the current application;
[0195] When the target Bloom filter satisfies the condition that causes the target Bloom filter to be falsely reported, the unused sub-Bloom filter is called as the target Bloom filter. A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it may include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in the present application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0196] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0197] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.
Claims
1. A tag management method, characterized in that: include: Obtain a target commodity set, wherein the target commodity set includes multiple target commodities, input the multiple target commodity information into the BI business system to perform a data processing task, and generate a BI commodity label table; The BI product label table includes a plurality of target product information and a plurality of target product labels corresponding to the plurality of target product information; The BI product label table is processed by a Bloom filter, the Bloom filter converts the target product information into binary objects, and generates a relationship table between target product labels and binary objects; In response to receiving a product query instruction sent by the system, determining whether the product corresponding to the product query instruction matches any target product in the target product set; If the product corresponding to the product query instruction matches the target product in the target product set, the target product label and the binary object relationship table corresponding to the target product matching the product corresponding to the product query instruction are reported to the system.
2. The method according to claim 1, characterized in that: Before obtaining the target product set, the following steps are further included: Get the label frame and label frame properties; In response to receiving a supplementary commodity set, the supplementary commodity set includes a plurality of supplementary commodities, obtaining commodity attributes corresponding to the plurality of supplementary commodities; Calculate the similarity between the product attributes of multiple supplementary products and the label frame attributes, and use the label frame whose similarity between the label frame attribute and the supplementary product attribute is greater than a threshold as the supplementary label frame; Extracting supplementary product feature information, generating a supplementary product tag based on the supplementary product feature information and a supplementary tag framework, associating the supplementary product tag with the supplementary product, and generating a supplementary product set.
3. The method according to claim 2, characterized in that The extracting of the supplementary commodity feature information and generating the supplementary commodity label according to the supplementary commodity feature information of the supplementary commodity comprises: Acquire image information, text information, and location information of the supplementary product, and perform multimodal fusion on the image information and text information of the supplementary product through modal joint learning to obtain feature information of the supplementary product, wherein the feature information of the supplementary product includes features of the supplementary product in multiple modalities; Inputting the supplementary label frame into the deep learning model, and outputting feature information of the supplementary label frame; The feature information of the supplementary product is matched with the feature information of the supplementary label frame. When the feature information of the supplementary product matches the feature information of the supplementary label frame, the feature information of the supplementary product is embedded in the supplementary label frame to form a supplementary product label.
4. The method according to claim 1, characterized in that: The method further comprises: In response to the BI business system receiving the target product set, the BI business system obtains target product label attribute information corresponding to the target product set; Determine whether the attribute information of the target product label information is consistent with the preset attribute information; If the attribute information of the target product label information is consistent with the preset attribute information, the target product label attribute information is displayed in the BI product label table in the first field corresponding to the preset attribute information; If the attribute information of the target product label information is inconsistent with the preset attribute information, a second field matching the target product label attribute information is queried from the preset field table, and the second field is displayed in the BI product label table.
5. The method according to claim 1, characterized in that The method further comprises In response to receiving a product query instruction sent by the system, determining whether the product corresponding to the product query instruction matches any target product in the target product set; If the product corresponding to the product query instruction does not match any target product in the target product set; Acquire product information of the product corresponding to the product query instruction from the supplementary product set, convert the product information of the product corresponding to the product query instruction acquired from the supplementary product set into a binary object to obtain a supplementary object; Importing the supplementary object into the target product tag and binary object relationship table; The target product label and the binary object corresponding to the target product matching the product corresponding to the product query instruction are reported to the system.
6. The method according to claim 1, characterized in that The BI system further includes a counting Bloom filter, and the method includes: In response to reporting the target product label and the binary object relationship table to the system, calculating the number of products corresponding to each target product label reported to the system; The counting Bloom filter obtains the number of commodities corresponding to each target commodity tag of the reporting system, and adds the numerical value of the number of commodities corresponding to each target commodity tag of the reporting system to the initial count value of the counting Bloom filter to update the count value of the counting Bloom filter.
7. The method according to claim 6, characterized in that The method further comprises: Splitting the counting Bloom filter into a preset number of sub-Bloom filters; Selecting any one of the plurality of sub-Bloom filters as a target Bloom filter for current application; When the target Bloom filter satisfies the condition that causes the target Bloom filter to have a false alarm, the unapplied sub-Bloom filter is called as the target Bloom filter.
8. A label management device, characterized in that: The device comprises: An acquisition module, used to acquire a target commodity set, wherein the target commodity set includes a plurality of target commodities; A business module, used for inputting the plurality of target commodity information into the BI business system to perform a data processing task and generate a BI commodity label table; the BI commodity label table includes a plurality of target commodity information and a plurality of target commodity labels corresponding to the plurality of target commodity information; A conversion module, used to process the BI product label table through a Bloom filter, wherein the Bloom filter converts the target product information into binary objects and generates a relationship table between target product labels and binary objects; A judgment module, configured to, in response to receiving a product query instruction sent by the system, judge whether the product corresponding to the product query instruction matches any target product in the target product set; The reporting module is used to match the product corresponding to the product query instruction with the target product in the target product set, and report the target product label and binary object relationship table corresponding to the target product matching the product corresponding to the product query instruction to the system.
9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.