Standardized processing method and device for article attribute information

Abstract

The invention discloses a standardized processing method and device for article attribute information, and relates to the technical field of data processing. A specific embodiment of the method comprises the following steps: trying to search corresponding standard attribute information based on acquired to-be-standardized article attribute information, acquiring description information of the to-be-standardized article under the condition that the corresponding standard attribute information is not searched, generating preliminary standard attribute information, obtaining corresponding evaluation parameters through calculation, and determining the to-be-standardized article according to the preliminary standard attribute information. And when the preliminary standard attribute information and the evaluation parameter meet a preset auditing condition, determining that the preliminary standard attribute information is standard attribute information corresponding to the to-be-standardized article attribute information for subsequent standardization processing. According to the embodiment, accurate standardization processing can be carried out on the attribute information of the to-be-processed article, and the error rate is reduced.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and device for standardizing object attribute information. Background Art

[0002] In recent years, with the increasing frequency of international trade, customs authorities around the world have tightened their requirements for the accuracy of declared import and export commodity data and the consistency of documents and goods. For example, some customs have explicitly prohibited the declaration of vague and general product names and imposed heavy fines for violations. Existing methods for cleaning and correcting commodity names declared by customs primarily rely on manual review and revisions based on regulations, which are significantly inefficient and inaccurate. Furthermore, manually reviewing regulatory documents is time-consuming and laborious, and the requirements for commodity name declaration vary from country to country. Finding and determining the correct declared commodity name requires considerable time and effort, especially for complex or uncommon commodities. Furthermore, the manual query and correction process carries a high risk of error, as the sheer volume of information, unclear descriptions, or misunderstandings can lead to errors in the final query. Existing methods for cleaning and correcting commodity names declared by customs have significant deficiencies in efficiency, accuracy, and international adaptability. Summary of the Invention

[0003] In view of this, embodiments of the present invention provide a method and apparatus for standardizing item attribute information, which can improve the accuracy of the standardized processing of item attribute information.

[0004] To achieve the above-mentioned object, according to one aspect of an embodiment of the present invention, a method for standardizing item attribute information is provided, comprising:

[0005] Obtaining attribute information of items to be standardized, and attempting to find standard attribute information corresponding to the attribute information of the items to be standardized;

[0006] In response to failure to find the corresponding standard attribute information, obtaining description information of the item to be standardized, and generating preliminary standard attribute information based on the description information of the item to be standardized;

[0007] Determine the evaluation parameters of preliminary standard attribute information;

[0008] In response to the evaluation parameters of the preliminary standard attribute information satisfying the preset review conditions, the preliminary standard attribute information is determined as the standard attribute information corresponding to the item attribute information to be standardized, so as to perform standardization processing on the item attribute information.

[0009] Optionally, the method further comprises:

[0010] In response to being able to find the corresponding standard attribute information, the corresponding standard attribute information is determined as the standard attribute information corresponding to the item attribute information to be standardized, so as to perform standardized processing on the item attribute information.

[0011] Optionally, the method further comprises:

[0012] In response to the evaluation parameters of the preliminary standard attribute information not satisfying the preset review conditions, the preliminary standard attribute information is regenerated, and the evaluation parameters of the regenerated preliminary standard attribute information are determined until the evaluation parameters of the regenerated preliminary standard attribute information satisfy the preset review conditions.

[0013] Optionally, the method further comprises:

[0014] A mapping relationship between the attribute information of the item to be standardized and the standard attribute information corresponding to the attribute information of the item to be standardized is established, and the mapping relationship is stored.

[0015] Optionally, generating preliminary standard attribute information based on the description information of the item to be standardized includes:

[0016] Recognition prompt words are generated based on the description information of the object to be standardized, and preliminary standard attribute information is generated based on the recognition prompt words.

[0017] Optionally, before attempting to find standard attribute information corresponding to the item attribute information to be standardized, the method further comprises: acquiring standard attribute information, standard attribute information mapping rules, and historical standard attribute information mapping data;

[0018] Building a training data set based on standard attribute information, standard attribute information mapping rules, and historical standard attribute information mapping data, and training a recognition model based on the training data set;

[0019] Generate preliminary standard attribute information based on the recognition prompt words, including:

[0020] The trained recognition model is called to generate preliminary standard attribute information based on the recognition prompt words.

[0021] Optionally, the method further comprises:

[0022] Calculate the score of standard attribute information based on preset review conditions and evaluation parameters;

[0023] The item attribute information to be standardized, the standard attribute information, and the score of the standard attribute information are added to the training data set, so that the recognition model can be incrementally trained based on the added training data set.

[0024] According to another aspect of an embodiment of the present invention, a device for standardizing item attribute information is provided, comprising:

[0025] An acquisition module, configured to acquire attribute information of items to be standardized and attempt to find standard attribute information corresponding to the attribute information of the items to be standardized;

[0026] a generating module for, in response to failure to find the corresponding standard attribute information, obtaining description information of the item to be standardized, and generating preliminary standard attribute information based on the description information of the item to be standardized;

[0027] a determination module for determining evaluation parameters of preliminary standard attribute information;

[0028] The determination module is further configured to determine the preliminary standard attribute information as the standard attribute information corresponding to the item attribute information to be standardized in response to the evaluation parameters of the preliminary standard attribute information satisfying the preset review conditions, so as to perform standardized processing on the item attribute information.

[0029] According to another aspect of an embodiment of the present invention, an electronic device is provided, comprising: one or more processors; a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the standardized processing method for item attribute information provided by an embodiment of the present invention.

[0030] According to another aspect of an embodiment of the present invention, a computer-readable medium is provided, on which a computer program is stored. When the program is executed by a processor, the method for standardizing the item attribute information provided by the embodiment of the present invention is implemented.

[0031] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program, which, when executed by a processor, implements the method for standardizing item attribute information provided by an embodiment of the present invention.

[0032] One embodiment of the above invention has the following advantages or beneficial effects: when the attribute information of the item to be standardized is obtained, the corresponding standard attribute information can be first attempted to be queried. If the information cannot be found, preliminary standard attribute information is generated using the recognition model. The calculated evaluation parameters and preset review conditions are used to determine whether the preliminary standard attribute information is the standard attribute information corresponding to the attribute information of the item to be standardized. If so, the preliminary standard attribute information is determined to be the standard attribute information corresponding to the attribute information of the item to be standardized. This can accurately determine the standard attribute information of the item, improve the efficiency of determining the standard attribute information, ensure the consistency of the item and the standard attribute information, and facilitate subsequent further verification of the item. In addition, before using the recognition model to generate preliminary standard attribute information, first attempting to query the standard attribute information can increase the processing speed of the process. That is, if the corresponding standard attribute information can be directly queried, there is no need to use the recognition model to generate it, which improves processing efficiency.

[0033] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The accompanying drawings are provided for a better understanding of the present invention and are not intended to limit the present invention.

[0035] Figure 1 1 is a schematic diagram of the main steps of a method for standardizing item attribute information according to an embodiment of the present invention;

[0036] Figure 2 Shown is a flow chart of a method for generating a standard commodity name according to an embodiment of the present invention;

[0037] Figure 3 2 is a schematic diagram of main modules of a device for standardizing item attribute information according to an embodiment of the present invention;

[0038] Figure 4 is an exemplary system architecture diagram in which embodiments of the present invention may be applied;

[0039] Figure 5 It is a schematic diagram of the structure of a computer system of a terminal device or a server suitable for implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0040] The following description of exemplary embodiments of the present invention is made in conjunction with the accompanying drawings, in which various details of the embodiments of the present invention are included to facilitate understanding. These details should be considered as merely exemplary. Therefore, it should be appreciated by those skilled in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0041] It should be noted that the collection, collection, updating, analysis, processing, use, transmission, and storage of user personal information involved in the technical solutions disclosed in this invention all comply with relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. Necessary measures are taken with respect to user personal information to prevent unauthorized access to user personal information data and to safeguard the security of user personal information, network security, and national security.

[0042] During customs inspections, the accuracy of declared commodity data and the consistency of the order and goods are strictly enforced. These inspections allow for accurate determination of whether the goods contain prohibited items. Currently, the primary method employed is manual review and correction, where the declared commodity names are corrected to obtain accurate product names. However, manual review places high demands on personnel and can lead to subjective errors, which can affect the accuracy of the product name corrections and hinder subsequent processes.

[0043] Figure 1 FIG. 1 is a schematic diagram of the main steps of the method for standardizing item attribute information according to an embodiment of the present invention. Figure 1 As shown, it includes steps S101 to S104.

[0044] Step S101: Acquire attribute information of an item to be standardized, and try to find standard attribute information corresponding to the attribute information of the item to be standardized.

[0045] Item attribute information to be standardized is obtained through the query system. The item attribute information may be information from data uploaded by the user through the query system's visual interface. Specifically, it may be the name of the product in the product data uploaded by the user, or information such as the product's size unit, weight unit, etc. This information may be set based on the actual application scenario and is not limited thereto. The obtained item attribute information to be standardized is subject to subsequent standardization operations to generate corresponding attribute information in a unified format that complies with relevant specifications, such as a standardized product name.

[0046] After obtaining the item attribute information to be standardized, the first step is to query whether there is corresponding standard attribute information. Specifically, the query scope can be a pre-stored standard attribute information mapping library. The standard attribute information mapping library stores mappings between item attribute information and standard attribute information, for example: silicone watch straps suitable for various watches (original product name / item attribute information) versus silicone watch straps (standard product name / standard attribute information). The mappings can be sourced from officially released standard name specification documents, historically processed standard data, and other sources. The mappings in the corresponding standard attribute information mapping library can be dynamically adjusted as specification documents are updated and as more standard data is processed to provide more accurate mapping data. During the query, it is necessary to determine whether any of the mappings contains information that is identical to the item attribute information to be standardized. Specifically, taking the product name as an example, the query involves checking whether the original product name in the mapping contains the same product name as the item attribute information to be standardized, and also checking whether the standard product name in the mapping contains the same product name as the item attribute information to be standardized. If any of the standard product names in the mapping contains the same product name as the item attribute information to be standardized, the corresponding standard attribute information is considered to be found.

[0047] Step S102 : in response to failure to find the corresponding standard attribute information, obtaining description information of the object to be standardized, and generating preliminary standard attribute information based on the description information of the object to be standardized.

[0048] If the standard attribute information corresponding to the item attribute information to be standardized cannot be found in the standard attribute information mapping library, the final standard attribute information needs to be generated based on the item attribute information to be standardized. In the subsequent embodiments of the present invention, the item attribute information is taken as an example of a product name.

[0049] It is necessary to obtain descriptive information for the items to be standardized. This descriptive information is the information in the data uploaded by the user through the query system's visual interface. Specifically, it can include the item's attributes and / or reference information. For example, in addition to the product name, the descriptive information may also include reference information such as the material and purpose of the product. By identifying the obtained descriptive information, preliminary standard attribute information is generated.

[0050] Step S103: determining evaluation parameters of preliminary standard attribute information.

[0051] An evaluation parameter is calculated based on the generated preliminary standard attribute information to determine the accuracy of the preliminary standard attribute information. In an embodiment of the present invention, the evaluation parameter is, for example, a confidence parameter. When determining the evaluation parameter for the preliminary standard attribute information, a pre-trained evaluation model can be used to perform calculations to determine the confidence level of the generated preliminary standard attribute information, such as high confidence, medium confidence, or low confidence. This allows for the subsequent determination of a confidence score for the preliminary standard attribute information based on the confidence parameter.

[0052] The prompt words for the evaluation model can be designed as follows:

[0053] Task Description: Based on the provided product name, reference information, and preliminary judgment results, determine whether {{#173.text#}} is correct. If correct, output "high confidence"; if incorrect, output "low confidence". Do not modify the original result; only provide the reason for the judgment and the judgment result.

[0054] Output format: Output results in XML format, where the correct field can only be "high confidence", "medium confidence" or "low confidence";

[0055] Output example: xxx;

[0056] Requirements: 1. Determine whether the preliminary results are correct; 2. Provide reasons for the judgment; 3. The output results must conform to the specified XML format;

[0057] Reference information (can be empty): {{#173.content#}}.

[0058] Step S104 , in response to the evaluation parameters of the preliminary standard attribute information satisfying the preset review conditions, the preliminary standard attribute information is determined as the standard attribute information corresponding to the item attribute information to be standardized, so as to perform standardization of the item attribute information.

[0059] The evaluation parameters of the generated preliminary standard attribute information are used to determine whether it meets the preset review conditions. If so, the generated preliminary standard attribute information is accurate and is determined as the final generated standard attribute information. If not, the generated preliminary standard attribute information is inaccurate. If the preset review conditions are met, the preliminary standard attribute information is determined as the standard attribute information of the item, and subsequent standardization processing is performed based on this preliminary standard attribute information.

[0060] Optionally, the preset review condition may be whether the evaluation parameter satisfies a value not less than a preset threshold value. For example, if the preset threshold value is medium confidence, then only high confidence and medium confidence values satisfy the preset review condition.

[0061] According to the method for standardizing the attribute information of items provided by an embodiment of the present invention, when the attribute information of the item to be standardized is obtained, the method can first attempt to query the corresponding standard attribute information. If the information cannot be found, the recognition model is used to generate preliminary standard attribute information. The calculated evaluation parameters and preset review conditions are used to determine whether the preliminary standard attribute information is the standard attribute information corresponding to the attribute information of the item to be standardized. If so, the preliminary standard attribute information is determined to be the standard attribute information corresponding to the attribute information of the item to be standardized. This method can accurately determine the standard attribute information of the item, improve the efficiency of determining the standard attribute information, ensure the consistency of the item with the standard attribute information, and facilitate subsequent further verification of the item. In addition, before using the recognition model to generate preliminary standard attribute information, first attempting to query the standard attribute information can increase the processing speed of the process. That is, if the corresponding standard attribute information can be directly queried, there is no need to use the recognition model to generate it, thereby improving processing efficiency.

[0062] Optionally, the method for standardizing item attribute information may further include: in response to being able to find corresponding standard attribute information, determining the corresponding standard attribute information as the standard attribute information corresponding to the item attribute information to be standardized, so as to perform standardized processing of the item attribute information.

[0063] If the corresponding standard attribute information can be found in the standard attribute information mapping library, the standard attribute information is directly obtained and used as the standard attribute information corresponding to the item attribute information to be standardized, so that subsequent standardization processing based on the standard attribute information can be carried out. Specifically, if the original product name is found to be the same as the item attribute information to be standardized (the product name to be standardized) during the query, then the corresponding standard product name is determined to be the standard attribute information corresponding to the item attribute information to be standardized based on the mapping relationship of the queried original product name; if the standard product name is found to be the same as the item attribute information to be standardized during the query, then it is determined that the item attribute information to be standardized is the standard name of the product, and subsequent standardization processing can be directly carried out based on this item attribute information.

[0064] According to the standardized processing method of item attribute information provided by an embodiment of the present invention, the standard attribute information mapping library can be queried based on the acquired item attribute information to be standardized. If the corresponding standard attribute information can be queried, the standard attribute information of the item attribute information can be determined, thereby improving the efficiency of determining the standard attribute information and also ensuring the accuracy of determining the standard attribute information.

[0065] Optionally, the method for standardizing item attribute information may also include: in response to the evaluation parameters of the preliminary standard attribute information not meeting the preset review conditions, regenerating the preliminary standard attribute information, and determining the evaluation parameters of the regenerated preliminary standard attribute information until the evaluation parameters of the regenerated preliminary standard attribute information meet the preset review conditions.

[0066] If the evaluation parameters of the generated preliminary standard attribute information do not meet the preset audit conditions, it means that the preliminary standard attribute information is inaccurate. At this time, it is necessary to use the recognition model to regenerate preliminary standard attribute information, and use the evaluation model to calculate the evaluation parameters of the regenerated preliminary standard attribute information to determine again whether the preset audit conditions are met. If it is still not met, continue to regenerate the preliminary standard attribute information until the evaluation parameters corresponding to the generated preliminary standard attribute information meet the preset audit conditions. Among them, when regenerating the preliminary standard attribute information, the preliminary standard attribute information that did not meet the preset audit conditions last time can be added to the prompt words of the recognition model as reference data to improve the accuracy of the regenerated preliminary standard attribute information.

[0067] According to the standardized processing method for item attribute information provided by an embodiment of the present invention, the accuracy of the preliminary standard attribute information generated by the recognition model can be judged based on evaluation parameters and preset audits. When it is judged that the preset audit conditions are not met, that is, when the generated information is inaccurate, it can be regenerated until the generated preliminary standard attribute information is accurate, thereby improving the accuracy of the final standard attribute information.

[0068] Optionally, the method for standardizing item attribute information may further include: establishing a mapping relationship between the item attribute information to be standardized and standard attribute information corresponding to the item attribute information to be standardized, and storing the mapping relationship.

[0069] After determining the standard attribute information corresponding to the item attribute information to be standardized, a mapping relationship is established between the two and this mapping relationship is added to the standard attribute information mapping library. This allows subsequent queries to directly retrieve the corresponding standard attribute information based on the mapping relationship. Adding this established mapping relationship to the standard attribute information mapping library enriches the data in the mapping library, increasing the probability of directly retrieval of standard attribute information in subsequent queries, reducing the process of generating standard attribute information and improving the efficiency of standardization.

[0070] Optionally, generating preliminary standard attribute information based on the description information of the object to be standardized includes: generating recognition prompt words based on the description information of the object to be standardized, and generating preliminary standard attribute information based on the recognition prompt words.

[0071] Specifically, recognition prompts are generated based on the acquired descriptive information, allowing a pre-trained recognition model to generate preliminary standard attribute information based on these prompts. For the generated prompts, the backend LLM (Large Language Model) to be connected is selected (Claude 3.5 (an artificial intelligence language model), gpt-4o (an artificial intelligence language model)), and an interface is exposed for connection to the customs system. Prompts are written according to three dimensions: roles, instructions, constraints, and example output. The roles and capabilities dimensions are generally universal. Roles: The LLM is required to act as an experienced customs expert, providing professional recommendations on product names from the perspective of customs commodity declarations. Instructions: Specific action requirements for the LLM, such as requiring the output of the standard declaration name corresponding to the product name. Constraints: Limitations on the LLM's output, such as requiring the output of several standard product names. Constraints can also be adjusted based on actual circumstances.

[0072] Take an example to illustrate the specific content of the recognition prompt words: Task: 1 Refer to the reference information in {{#173.content#}} to determine whether {{#173.query#}} is a fuzzy product name. A fuzzy product name generally refers to a product name that is very broad and difficult to determine the specific product. 2. If it is not a fuzzy product name, determine its standard product name based on the product name and reference information in {{#173.content#}}. The standard product name generally consists of the product material and product category. If there is a standard product name in the name and reference information, use it directly. If there is no standard product name, summarize the product material and type yourself to get the standard product name. 3. The above results are output in the specified XML (Extensible Markup Language) format;

[0073] Background information: Material list: Materials in the reference information and various common standard materials; Type list: Specific product types such as chairs, mobile phone holders, tops, underwear, mobile phones, books, etc.; Standard product names are composed of material and type;

[0074] Output format: Please output materials and types in XML format;

[0075] Output result restrictions: Do not output any additional information except the result information in XML format;

[0076] Reference information (can be empty){{#173.content#}};

[0077] Input: The current product name is {{{#173.query#}}}.

[0078] Optionally, before attempting to find standard attribute information corresponding to the item attribute information to be standardized, the method for standardizing the item attribute information may also include: obtaining standard attribute information, standard attribute information mapping rules and historical standard attribute information mapping data; constructing a training data set based on the standard attribute information, standard attribute information mapping rules and historical standard attribute information mapping data, and training a recognition model based on the training data set; generating preliminary standard attribute information based on recognition prompt words, including: calling the trained recognition model to generate preliminary standard attribute information based on the recognition prompt words.

[0079] Before querying the standard attribute information mapping library, the recognition model must be trained. Prior to training, the standard attribute information and standard attribute information mapping rules specified in various regulatory documents are obtained. Combined with historical standard attribute information mapping data, a variety of data with mapping relationships is constructed. A training dataset for the recognition model is then built based on this data. The recognition model is trained using this training dataset for subsequent use. The trained recognition model generates preliminary standard information based on recognition prompt words. The standard attribute information mapping rules are regularly updated data. These rules include: 1. Special letters, symbols, adjectives, and modifiers are removed from the Chinese product name; 2. Material, purpose, and physical name are retained; 3. The highest-weighted result is selected based on semantic and part-of-speech analysis of the product name and output; 4. If the cleaned Chinese product name is meaningless, the English product name is determined to be meaningful; 5. If the cleaned Chinese or English product name is meaningful, the next step is performed; 6. The system identifies and labels the product (e.g., food, cosmetics, tape, powder, liquid, etc.); and 7. The cleaned standard Chinese and English product names are output.

[0080] Optionally, data from each obtained regulatory document (such as officially released standard product name documents) needs to be extracted and organized to construct a training dataset. Specifically, documents are classified by sub-item number and sub-item clause, and keywords are extracted using the TF-IDF (statistical method) algorithm to construct a structured training set. The classified documents are then uploaded to the knowledge base as training sets for Claude 3.5 and GPT-4o. These classified documents help the system identify the category of the currently declared product name and assist in generating standard product names.

[0081] According to the standardized processing method of item attribute information provided by the embodiment of the present invention, a training data set can be constructed based on various specification documents and historical data to train a recognition model, so that the recognition model can output more accurate recognition results and improve the reliability of determining standard attribute information.

[0082] Optionally, the method for standardizing item attribute information may also include: calculating the evaluation score of the standard attribute information based on preset review conditions and evaluation parameters; adding the item attribute information to be standardized, the standard attribute information and the evaluation score of the standard attribute information to the training data set, so that the recognition model can be incrementally trained based on the added training data set.

[0083] For the determined evaluation parameters, an evaluation score needs to be further calculated to determine the accuracy of the standard attribute information. The preset review conditions also include the evaluation score calculation method. Specifically, the evaluation score calculation also requires integration with the manual review parameters of the standard attribute information. The final evaluation score is determined based on the evaluation parameters output by the evaluation model and the manual review parameters. For example, the relationship between the manual review result and the GPT-4O evaluation result is set on a scale of 1 to 9: High evaluation (large model) - High evaluation (review): 9 points; Medium evaluation (large model) - Medium evaluation (review): 9 points; Low evaluation (large model) - Low evaluation (review): 9 points; High evaluation (large model) - Medium evaluation (review): 4 points; Medium evaluation (large model) - High evaluation (review): 4 points; Medium evaluation (large model) - Low evaluation (review): 4 points; Low evaluation (large model) - Medium evaluation (review): 4 points; High evaluation (large model) - Low evaluation (review): 1 point; Low evaluation (large model) - High evaluation (review): 1 point. The above-mentioned setting of the evaluation score is only used as an embodiment. In actual application, the specific score can be set to different values according to requirements, and there is no limitation on this.

[0084] Optionally, after obtaining the evaluation score, incremental data is constructed by combining the attribute information of the item to be standardized and the corresponding standard attribute information, and added to the training data set, so as to use the added training data set to train the recognition model so that it can output more accurate results and improve the accuracy of the standard attribute information.

[0085] According to the method for standardizing the attribute information of items provided by an embodiment of the present invention, when the attribute information of the item to be standardized is obtained, it is possible to first attempt to query the corresponding standard attribute information from the standard attribute mapping library. If the query fails, the model is used to generate the standard attribute information, thereby improving the reliability of determining the standard attribute information and improving the processing efficiency. Specifically, when the model is used to generate the standard attribute information, the calculated evaluation parameters and preset review conditions are used to determine whether the preliminary standard attribute information is the standard attribute information corresponding to the attribute information of the item to be standardized. This can further accurately determine the standard attribute information of the item to facilitate subsequent further verification of the item. In addition, during the processing process, incremental data can be constructed based on the generated standard attribute information to perform incremental training on the model, greatly improving the accuracy of the generated results.

[0086] like Figure 2 The flowchart of the method for generating standard product names according to an embodiment of the present invention is shown. A user uploads unprocessed product data through a query system. The system then queries the product name mapping library. If it determines that a corresponding standard product name can be found, the found standard product name is returned to the user as a result. If no corresponding standard product name can be found, the product name generation system invokes the LLM model to generate a recommendation result and evaluation parameters for the generated recommendation result. The generated recommendation result is the preliminary standard product name. Specifically, the LLM model includes an identification model and an evaluation model. After the identification model generates a recommendation result, the recommendation result is input into the evaluation model to obtain the evaluation parameters for the recommendation result. The recommendation result and evaluation parameters are passed for manual review to calculate the evaluation score of the recommendation result. If the evaluation score is not less than a preset threshold, the recommendation result is determined to be passed, and the passed data is returned to the user. The passed data is also added to the training set data as incremental data to further optimize and train the LLM model. The training set data includes various types of specification file data, mapping rules, standard product name data, and product name mapping library data.

[0087] Figure 3 FIG. 3 is a schematic diagram of main modules of a device 300 for standardizing item attribute information according to an embodiment of the present invention. The device 300 for standardizing item attribute information mainly includes: an acquisition module 301 , a generation module 302 and a determination module 303 .

[0088] An acquisition module 301 is configured to acquire attribute information of an item to be standardized and attempt to find standard attribute information corresponding to the attribute information of the item to be standardized;

[0089] A generating module 302 is configured to obtain description information of the item to be standardized in response to failure to find the corresponding standard attribute information, and generate preliminary standard attribute information based on the description information of the item to be standardized;

[0090] Determination module 303, for determining evaluation parameters of preliminary standard attribute information;

[0091] The determination module 303 is further configured to determine the preliminary standard attribute information as the standard attribute information corresponding to the item attribute information to be standardized in response to the evaluation parameters of the preliminary standard attribute information satisfying the preset review conditions, so as to perform standardization of the item attribute information.

[0092] The device for standardizing the attribute information of an item provided by an embodiment of the present invention can, upon obtaining the attribute information of the item to be standardized, first attempt to query the corresponding standard attribute information. If the information cannot be found, the recognition model is used to generate preliminary standard attribute information. The calculated evaluation parameters and preset review conditions are used to determine whether the preliminary standard attribute information is the standard attribute information corresponding to the attribute information of the item to be standardized. If so, the preliminary standard attribute information is determined to be the standard attribute information corresponding to the attribute information of the item to be standardized. This device can accurately determine the standard attribute information of the item, improve the efficiency of determining the standard attribute information, ensure the consistency of the item with the standard attribute information, and facilitate subsequent further verification of the item. In addition, before using the recognition model to generate preliminary standard attribute information, first attempting to query the standard attribute information can increase the processing speed of the process. That is, if the corresponding standard attribute information can be directly queried, there is no need to use the recognition model to generate it, thereby improving processing efficiency.

[0093] Optionally, the determination module 303 is further configured to, in response to being able to find the corresponding standard attribute information, determine the corresponding standard attribute information as the standard attribute information corresponding to the item attribute information to be standardized, so as to perform standardization of the item attribute information.

[0094] Optionally, the generation module 302 is also used to regenerate the preliminary standard attribute information in response to the evaluation parameters of the preliminary standard attribute information not meeting the preset review conditions, and determine the evaluation parameters of the regenerated preliminary standard attribute information until the evaluation parameters of the regenerated preliminary standard attribute information meet the preset review conditions.

[0095] Optionally, the device 300 for standardizing item attribute information may further include: an establishing module 304 (not shown in the figure), configured to establish a mapping relationship between the item attribute information to be standardized and the standard attribute information corresponding to the item attribute information to be standardized and store the mapping relationship.

[0096] Optionally, the generating module 302 is further configured to generate recognition prompt words based on the description information of the item to be standardized, and generate preliminary standard attribute information based on the recognition prompt words.

[0097] Optionally, the acquisition module 301 is also used to acquire standard attribute information, standard attribute information mapping rules and historical standard attribute information mapping data; the standardization processing device 300 for item attribute information may also include: a construction module 305 (not shown in the figure), which is used to construct a training data set based on the standard attribute information, standard attribute information mapping rules and historical standard attribute information mapping data, and train a recognition model based on the training data set; the generation module 302 is also used to call the trained recognition model to generate preliminary standard attribute information based on the recognition prompt word.

[0098] Optionally, the device 300 for standardizing the item attribute information may further include: a calculation module 306 (not shown in the figure), configured to calculate a score of the standard attribute information based on preset review conditions and evaluation parameters;

[0099] The device 300 for standardizing item attribute information may further include: an adding module 307, which is further used to add the item attribute information to be standardized, the standard attribute information and the score of the standard attribute information to the training data set, so that the recognition model can be incrementally trained based on the added training data set.

[0100] According to the device for standardizing the attribute information of an item provided by an embodiment of the present invention, when the attribute information of the item to be standardized is obtained, it can first attempt to query the corresponding standard attribute information from the standard attribute mapping library. If the query fails, the model is used to generate the standard attribute information, thereby improving the reliability of the determination of the standard attribute information and improving the processing efficiency. Specifically, when the model is used to generate the standard attribute information, the calculated evaluation parameters and the preset review conditions are used to determine whether the preliminary standard attribute information is the standard attribute information corresponding to the attribute information of the item to be standardized. This can further accurately determine the standard attribute information of the item and facilitate subsequent further verification of the item. In addition, during the processing process, incremental data can be constructed based on the generated standard attribute information to perform incremental training on the model, which greatly improves the accuracy of the generated results.

[0101] Figure 4 An exemplary system architecture 400 is shown to which the method for standardizing item attribute information or the apparatus for standardizing item attribute information according to an embodiment of the present invention may be applied.

[0102] like Figure 4 As shown, system architecture 400 may include terminal devices 401, 402, 403, a network 404, and a server 405. Network 404 is used to provide a medium for communication links between terminal devices 401, 402, 403 and server 405. Network 404 may include various connection types, such as wired or wireless communication links or fiber optic cables.

[0103] Users can use terminal devices 401, 402, and 403 to interact with server 405 via network 404 to receive or send messages, etc. Terminal devices 401, 402, and 403 can be installed with various communication client applications, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only as examples).

[0104] The terminal devices 401 , 402 , and 403 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, and desktop computers.

[0105] Server 405 may be a server that provides various services, such as a backend management server (for example only) that supports shopping websites browsed by users using terminal devices 401, 402, and 403. The backend management server may query, generate, and process received data such as generation requests, and provide feedback (for example, standard attribute information) to the terminal device.

[0106] It should be noted that the method for standardizing item attribute information provided in the embodiment of the present invention is generally executed by the server 405 . Accordingly, the device for standardizing item attribute information is generally disposed in the server 405 .

[0107] It should be understood that Figure 4 The number of terminal devices, networks and servers in the embodiment is merely illustrative. Any number of terminal devices, networks and servers may be provided as required.

[0108] Reference below Figure 5 , which shows a schematic structural diagram of a computer system 500 of a terminal device or server suitable for implementing an embodiment of the present invention. Figure 5 The terminal device or server shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0109] like Figure 5 As shown, computer system 500 includes a central processing unit (CPU) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage unit 508 into a random access memory (RAM) 503. Various programs and data required for the operation of system 500 are also stored in RAM 503. CPU 501, ROM 502, and RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to bus 504.

[0110] The following components are connected to the I / O interface 505: an input section 506 including a keyboard, mouse, and the like; an output section 507 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and speakers; a storage section 508 including devices such as a hard disk; and a communication section 509 including a network interface card such as a LAN card or a modem. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as needed. Removable media 511, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 510 as needed, so that computer programs read from the removable media can be installed in the storage section 508 as needed.

[0111] In particular, according to embodiments disclosed herein, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed herein include a computer program product comprising a computer program embodied on a computer-readable medium, the computer program containing program code for executing the methods illustrated in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 509 and / or installed from removable media 511. When executed by central processing unit (CPU) 501, the computer program performs the aforementioned functions defined in the system of the present invention.

[0112] It should be noted that the computer-readable medium described in the present invention may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. Computer-readable storage media may include, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more conductors, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device, or component. In the present invention, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such a propagated data signal may take a variety of forms, including, but not limited to, electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. Program code embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wireline, optical fiber cable, RF, or any suitable combination thereof.

[0113] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the above-mentioned module, program segment, or a part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0114] The units or modules involved in the embodiments of the present invention may be implemented in software or hardware. The units or modules described may also be provided in a processor. For example, they may be described as follows: a processor includes an acquisition module, a generation module, and a determination module. The names of these modules do not, in some cases, constitute limitations on the units or modules themselves. For example, the generation module may also be described as "a module for obtaining descriptive information of an item to be standardized in response to a failure to find corresponding standard attribute information from a standard attribute information mapping library, generating identification prompt words based on the descriptive information, and generating preliminary standard attribute information based on the identification prompt words."

[0115] As another aspect, the present invention further provides a computer-readable medium, which may be included in the device described in the above embodiments, or may exist independently without being incorporated into the device. The computer-readable medium carries one or more programs, and when the one or more programs are executed by the device, the device includes:

[0116] Obtaining attribute information of items to be standardized, and attempting to find standard attribute information corresponding to the attribute information of the items to be standardized;

[0117] In response to failure to find the corresponding standard attribute information, obtaining description information of the object to be standardized, and generating preliminary standard attribute information based on the description information of the object to be standardized;

[0118] Determine the evaluation parameters of preliminary standard attribute information;

[0119] In response to the evaluation parameters of the preliminary standard attribute information satisfying the preset review conditions, the preliminary standard attribute information is determined as the standard attribute information corresponding to the item attribute information to be standardized, so as to perform standardization processing on the item attribute information.

[0120] According to the technical solution of the embodiment of the present invention, when the attribute information of the item to be standardized is obtained, the corresponding standard attribute information can be first attempted to be queried. If the information cannot be found, the recognition model is used to generate preliminary standard attribute information. The calculated confidence parameter and the preset review conditions are used to determine whether the preliminary standard attribute information is the standard attribute information corresponding to the attribute information of the item to be standardized. If so, the preliminary standard attribute information is determined to be the standard attribute information corresponding to the attribute information of the item to be standardized. This can accurately determine the standard attribute information of the item, improve the efficiency of determining the standard attribute information, ensure the consistency of the item and the standard attribute information, and facilitate subsequent further verification of the item. In addition, before using the recognition model to generate preliminary standard attribute information, first attempting to query the standard attribute information can increase the processing speed of the process. That is, if the corresponding standard attribute information can be directly queried, there is no need to use the recognition model to generate it, which improves processing efficiency.

[0121] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for standardizing item attribute information, characterized in that: include: Acquire attribute information of items to be standardized, and attempt to find standard attribute information corresponding to the attribute information of the items to be standardized; In response to failure to find the corresponding standard attribute information, obtaining description information of the object to be standardized, and generating preliminary standard attribute information based on the description information of the object to be standardized; determining evaluation parameters of the preliminary standard attribute information; In response to the evaluation parameters of the preliminary standard attribute information satisfying the preset review conditions, the preliminary standard attribute information is determined as the standard attribute information corresponding to the item attribute information to be standardized, so as to perform standardization processing on the item attribute information.

2. The method according to claim 1, characterized in that The method further comprises: In response to being able to find the corresponding standard attribute information, the corresponding standard attribute information is determined as the standard attribute information corresponding to the item attribute information to be standardized, so as to perform standardization processing on the item attribute information.

3. The method according to claim 1, characterized in that The method further comprises: In response to the evaluation parameters of the preliminary standard attribute information not meeting the preset review conditions, the preliminary standard attribute information is regenerated, and the evaluation parameters of the regenerated preliminary standard attribute information are determined until the evaluation parameters of the regenerated preliminary standard attribute information meet the preset review conditions.

4. The method according to claim 1, wherein The method further comprises: A mapping relationship between the item attribute information to be standardized and the standard attribute information corresponding to the item attribute information to be standardized is established, and the mapping relationship is stored.

5. The method according to any one of claims 1 to 4, characterized in that The generating of preliminary standard attribute information based on the description information of the object to be standardized includes: Recognition prompt words are generated based on the description information of the object to be standardized, and preliminary standard attribute information is generated based on the recognition prompt words.

6. The method according to claim 5, characterized in that Before attempting to find standard attribute information corresponding to the attribute information of the item to be standardized, the method further includes: Obtaining standard attribute information, standard attribute information mapping rules, and historical standard attribute information mapping data; constructing a training data set based on the standard attribute information, the standard attribute information mapping rule, and the historical standard attribute information mapping data, and training a recognition model based on the training data set; The generating of preliminary standard attribute information based on the recognition prompt word includes: The trained recognition model is called to generate preliminary standard attribute information based on the recognition prompt word.

7. The method according to claim 6, characterized in that The method further comprises: Calculating a score for the standard attribute information based on the preset review conditions and the evaluation parameters; The item attribute information to be standardized, the standard attribute information, and the score of the standard attribute information are added to the training data set, so that the recognition model can be incrementally trained based on the added training data set.

8. A device for standardizing item attribute information, characterized in that: include: an acquisition module, configured to acquire attribute information of items to be standardized and attempt to find standard attribute information corresponding to the attribute information of the items to be standardized; a generating module configured to, in response to failure to find the corresponding standard attribute information, obtain description information of the item to be standardized, and generate preliminary standard attribute information based on the description information of the item to be standardized; a determination module, configured to determine evaluation parameters of the preliminary standard attribute information; The determination module is further configured to determine the preliminary standard attribute information as the standard attribute information corresponding to the item attribute information to be standardized in response to the evaluation parameters of the preliminary standard attribute information satisfying preset review conditions, so as to perform standardization of the item attribute information.

9. An electronic device, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7.

10. A computer-readable medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

11. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images