Label generation method and device based on large model and configuration, equipment and product

Abstract

A label generation method, device and equipment based on a large model and configuration, and a product, relate to the technical field of large models and computers, and the label generation method comprises: acquiring label demand information of a label required by a business scenario; generating a label based on the label demand information through at least one of the following ways: in the case that the label demand information includes a preset keyword, performing semantic understanding on the preset keyword through a large model to generate a similar keyword similar in semantics to the preset keyword; obtaining a first label according to the similar keyword; in the case that the label demand information includes label caliber information, generating a second label according to a label generation rule corresponding to the label caliber information, wherein the label generation rule is determined in response to a first configuration operation based on the label caliber information in a first configuration page. Not only does it improve the development efficiency and iteration efficiency of the label, but also improves the satisfaction rate of the label to the business, effectively solving the problem of increasing label demand of the business.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of large models and computer technology, in particular, to a label generation method and device based on large models and configuration, equipment and products. BACKGROUND

[0002] A label is a keyword or identifier used to classify, mark and retrieve data, which can help users quickly locate, manage and understand data. Labels can include rule-based labels and indicator-based labels. Rule-based labels are generated based on predefined rules or conditions, such as the classification label of product A being "electronic products". This label can help merchants quickly classify products or make it easier for users to search and browse products on the website by category. Indicator-based labels are generated based on statistical or calculated results of data, such as "product A sales in the past month". This label can help merchants understand product sales, and in turn help merchants adjust inventory and promotion strategies.

[0003] With the widespread popularity of labels in practical applications, the demand for labels continues to rise. However, the manual development of labels by developers in related technologies is inefficient and outdated, and cannot meet the growing demand for labels. SUMMARY

[0004] This summary is provided to introduce a selection of concepts that are further described below in the detailed description. This summary is not intended to identify key or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.

[0005] In a first aspect, the present disclosure provides a label generation method based on large models and configuration, comprising:

[0006] Obtaining label requirement information of labels required by a business scenario;

[0007] Generating labels based on the label requirement information by at least one of the following ways:

[0008] In the case where the label requirement information includes a preset keyword, performing semantic understanding on the preset keyword by a large model to generate a similar keyword similar in semantics to the preset keyword; obtaining a first label according to the similar keyword;

[0009] In a case where the label demand information comprises label caliber information, a second label is generated according to a label generation rule corresponding to the label caliber information, wherein the label generation rule is determined in response to a first configuration operation based on the label caliber information in a first configuration page, and the first configuration page is used for configuring a generation rule of a label.

[0010] In a second aspect, the present disclosure provides a label generation device based on a large model and configuration, comprising:

[0011] A obtaining module is configured to obtain label demand information of a label required by a business scenario;

[0012] A generating module is configured to generate a label based on the label demand information by at least one of the following ways:

[0013] In a case where the label demand information comprises a preset keyword, a similar keyword similar in semantics to the preset keyword is generated by performing semantic understanding on the preset keyword by a large model; and a first label is obtained according to the similar keyword;

[0014] In a case where the label demand information comprises label caliber information, a second label is generated according to a label generation rule corresponding to the label caliber information, wherein the label generation rule is determined in response to a first configuration operation based on the label caliber information in a first configuration page, and the first configuration page is used for configuring a generation rule of a label.

[0015] In a third aspect, the present disclosure provides a computer readable medium having a computer program stored thereon, wherein the program is executed by a processing device to implement the steps of the method in the first aspect.

[0016] In a fourth aspect, the present disclosure provides an electronic device, comprising:

[0017] A storage device having a computer program stored thereon;

[0018] A processing device configured to execute the computer program in the storage device to implement the steps of the method in the first aspect.

[0019] In a fifth aspect, the present disclosure provides a computer program product comprising a computer program, wherein the computer program is executed by a processor to implement the steps of the method in the first aspect.

[0020] By the technical solution, in a case where the label demand information includes preset keywords, the similar keywords similar in semantics to the preset keywords can be generated by the large model based on semantic understanding of the preset keywords, and then the first label can be obtained according to the similar keywords. In a case where the label demand information includes label caliber information, the second label can be generated according to the label generation rule configured based on the label caliber information. By using the method, the label similar in semantics to the preset keywords can be automatically expanded by the text processing capability of the large model, and the label generation rule can be configured according to the label caliber information, and then the label meeting the label generation rule can be automatically generated. The method not only improves the development efficiency and iteration efficiency of the label, but also improves the satisfaction rate of the label to the business, and effectively solves the problem of increasing label demand of the business.

[0021] Other features and advantages of the present disclosure will be described in detail in the following detailed description section. BRIEF DESCRIPTION OF DRAWINGS

[0022] The above and other features, advantages, and aspects of embodiments of the present disclosure will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings in which:

[0023] Figure 1 is a schematic flow chart of a label generation method based on a large model and configuration according to an exemplary embodiment of the present disclosure;

[0024] Figure 2 is a schematic diagram of a label generation process based on a large model according to an exemplary embodiment of the present disclosure;

[0025] Figure 3 is a schematic diagram of a first configuration page according to an exemplary embodiment of the present disclosure;

[0026] Figure 4 is a comparative schematic diagram of a related technology and a label generation process of the present embodiment according to an exemplary embodiment of the present disclosure;

[0027] Figure 5 is a process schematic diagram of label quality detection according to an exemplary embodiment of the present disclosure;

[0028] Figure 6 is a schematic diagram of a second configuration page according to an exemplary embodiment of the present disclosure;

[0029] Figure 7 is a process schematic diagram of a division development label according to an exemplary embodiment of the present disclosure;

[0030] Figure 8 is a structural block diagram of a large model and configuration-based label generation apparatus according to an exemplary embodiment of the disclosure;

[0031] Figure 9 is a structural schematic diagram of an electronic device according to an exemplary embodiment of the disclosure. DETAILED DESCRIPTION

[0032] Embodiments of the disclosure will be described in more detail with reference to the drawings. While certain embodiments of the disclosure will be illustrated in the drawings, it is understood that the disclosure can be embodied in various forms and should not be interpreted in a limited sense as set forth in the embodiments set forth herein, but rather, the embodiments are provided to more thoroughly and completely understand the disclosure. It should be understood that the drawings and embodiments of the disclosure are for exemplary purposes only and are not intended to limit the scope of protection of the disclosure.

[0033] It should be understood that each of the steps described in the method embodiments of the disclosure can be performed in different orders and / or in parallel. In addition, the method embodiments can include additional steps and / or omit the steps shown. The scope of the disclosure is not limited in this respect.

[0034] The term "comprising" and variations thereof as used herein are open-ended, that is, "including, but not limited to". The term "based on" is "based, at least in part, on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Related definitions will be given in the description below.

[0035] It should be noted that the concepts of "first", "second", etc. mentioned in the disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0036] It should be noted that the modification of "one" or "multiple" mentioned in the disclosure is illustrative and not limiting, and those skilled in the art should understand that unless otherwise explicitly indicated in the context, it should be understood as "one or more".

[0037] The names of the messages or information exchanged between the devices in the embodiments of the disclosure are only for illustrative purposes, and are not used to limit the scope of the messages or information.

[0038] It can be understood that before using the technical solutions disclosed in the embodiments of the disclosure, the type, scope of use, use scenario, etc. of the personal information involved in the disclosure should be informed to the user and the authorization of the user should be obtained in a proper manner according to relevant laws and regulations.

[0039] For example, in response to receiving an active request of a user, prompt information is sent to the user to explicitly prompt the user that the operation requested to be performed will require obtaining and using personal information of the user. Thus, the user can autonomously select whether to provide personal information to the software or hardware such as an electronic device, an application program, a server or a storage medium performing the operation of the technical solution of the present disclosure according to the prompt information.

[0040] As an optional but non-limiting implementation manner, in response to receiving an active request of a user, the manner of sending prompt information to the user may, for example, be a pop-up window manner, in which the prompt information may be presented in a text manner. In addition, the pop-up window may also carry selection controls for the user to select “agree” or “disagree” to provide personal information to the electronic device.

[0041] It can be understood that the above notification and obtaining user authorization process is only illustrative and does not limit the implementation manner of the present disclosure, and other manners meeting relevant laws and regulations can also be applied to the implementation manner of the present disclosure.

[0042] Meanwhile, it can be understood that the data (including but not limited to the data itself, the obtaining or use of the data) involved in the technical solution should comply with the requirements of relevant laws and regulations and relevant provisions.

[0043] Taking an e-commerce business scenario as an example, tags can be used in core scenarios such as message pushing, finding potential users, activity circle selection, search and recommendation, and advertisement placement. With the increasing application of tags in e-commerce businesses, the business has more and more demands for tags, and along with this, there is a contradiction between the increasing demand for tags and the low efficiency of tag development.

[0044] Taking iterative updating of rule-based tags as an example, some rule-based tags need to list similar keywords to realize tag updating. Related technologies mainly realize tag iteration by manually sorting keywords and adding them to storage, which is time-consuming and inefficient, and tag updating is lagging. Alternatively, based on a developed and deployed data processing link, raw data collected from various data sources is processed to obtain tags and register them to a tag management platform. Although the data processing link is visible, it takes time to develop the data processing link, which is inefficient.

[0045] Therefore, the present disclosure provides a tag generation method and device based on a large model and configuration, and an equipment and product, to solve the above technical problems.

[0046] The embodiments of the present disclosure are further explained and described below with reference to the accompanying drawings.

[0047] Figure 1is a flowchart of a large model and configuration-based label generation method according to an example embodiment of the present disclosure, with reference to Figure 1 The label generation method can include the following steps:

[0048] S101: Obtain label requirement information of a label required by a business scenario.

[0049] For example, the label requirement information can be determined according to the label requirements of the business scenario, and specifically can be configured according to the requirements. The present disclosure does not limit this.

[0050] S102: Generate a label based on the label requirement information by at least one of the following ways: in the case that the label requirement information includes a preset keyword, perform semantic understanding on the preset keyword by a large model to generate a similar keyword similar in semantics to the preset keyword; obtain a first label according to the similar keyword; in the case that the label requirement information includes label caliber information, generate a second label according to a label generation rule corresponding to the label caliber information, wherein the label generation rule is determined in response to a first configuration operation based on the label caliber information in a first configuration page, and the first configuration page is used to configure the generation rule of the label.

[0051] For example, in an actual business scenario, rule-based labels are usually displayed in the form of words, so more semantically similar labels can be expanded after the large model performs semantic understanding on the keywords of the labels fed back by the business. Assuming that it is required to circle the users who purchase "infant milk powder" in the business scenario, "infant milk powder" can be used as a preset keyword to expand more similar keywords by a large model, and then more labels are obtained, such as "infant formula milk powder" and "infant milk powder", etc. The present disclosure does not limit this.

[0052] For example, in an actual business scenario, for rule-based labels and index-based labels, the specific standards and ranges of the required labels can be accurately defined and described by setting label caliber information, so as to configure the corresponding label generation rule based on the label caliber information, and then automatically generate the required labels based on the label generation rule. Assuming that it is required to adjust the inventory according to the historical sales of the goods in the business scenario, the label generation rule can include label source, algorithm, time range, filtering condition, etc., and then the label generated according to the label generation rule, such as "the past one month sales of commodity A", etc. The present disclosure does not limit this.

[0053] It is worth noting that the large model method or the configuration method can be used to generate labels, or the large model method and the configuration method can be used to generate labels, and specifically the selection can be made according to the requirements. The present disclosure does not limit this.

[0054] By using the above method, the text processing capability of the large model can be used to automatically expand labels similar in semantics to the preset keyword. In addition, the label generation rule can be configured according to the label caliber information, and then the label meeting the label generation rule can be automatically generated. This not only improves the development efficiency and iteration efficiency of the label, but also improves the satisfaction rate of the label to the business, effectively solving the problem of increasing label demand of the business.

[0055] In a possible manner, the large model is used to understand the semantics of the preset keyword, and a similar keyword similar in semantics to the preset keyword is generated. The method comprises: using the large model to understand the semantics of the preset keyword, obtaining a target long tail keyword similar in semantics to the preset keyword from a target data source, and determining the target long tail keyword as a similar keyword.

[0056] It should be understood that, since the same label may represent different meanings in different business scenarios, in order to generate a label that meets the label demand, the business scenario information can be limited in the label demand information, for example, an e-commerce business scenario, and then the large model is used to mine labels from the corresponding data source, thereby improving the accuracy of label generation. In addition, the large model can be input with a corpus corresponding to the business scenario to help the large model more accurately understand the semantics and further improve the accuracy of label generation.

[0057] As Figure 2 indicated, the preset keyword can be determined according to the rule-based label fed back by the business. Taking an e-commerce business scenario as an example, the text processing capability of the large model is used to understand the semantics of the preset keyword, and similar keywords similar in semantics to the preset keyword are perceived from the business data generated by the e-commerce business scenario. The large model can periodically perceive similar keywords from the business data generated by the business scenario to iteratively update the label and avoid the problem of update lag.

[0058] It should be understood that the keywords similar in semantics to the preset keyword perceived by the large model can be long tail keywords. Long tail keywords refer to keywords that are relatively specific, detailed, and have relatively small search volumes. The label determined based on the long tail keyword can improve the accuracy of the label and better meet the business demand of the label.

[0059] For example, continuing to take "infant milk powder" as the preset keyword as an example, if the keyword is directly perceived, the "milk powder" label can be expanded, and then the users who purchase "adult milk powder" and "milk powder for the middle-aged and the elderly" can be selected, which does not meet the actual business demand. By perceiving long tail keywords similar in semantics to the preset keyword, labels that better meet the business demand can be mined, such as "infant 1-stage milk powder" and "infant 2-stage milk powder", and then users meeting the business demand can be accurately selected.

[0060] To further ensure the accuracy of the label, a similarity algorithm can be built into the large model to output similar keywords after screening the long-tail words. For example, the semantic similarity between the preset keyword and each long-tail word is calculated based on the cosine similarity, and then the target long-tail word with a semantic similarity greater than a preset similarity is selected as a similar keyword, thereby further improving the accuracy of the label and improving the label development efficiency and the satisfaction rate of the label to the business. The preset similarity can be set according to the requirements, and the present disclosure does not limit this.

[0061] It should be noted that the above large model can be fine-tuned based on a general large model, combined with a corpus of a business scenario and a similarity algorithm, to meet the business needs of automatic expansion of labels.

[0062] In a possible manner, the label generation method further includes: in the case where the preset keyword does not include the target long-tail word, storing the preset keyword in a candidate label pool, wherein the target data source includes the candidate label pool.

[0063] For example, continuing to refer to Figure 2 In the case where the large model cannot perceive the long-tail words similar in semantics based on the preset keyword, it indicates that more labels cannot be expanded based on the preset keyword. At this time, the preset keyword can be stored in the candidate label pool as one of the data sources for subsequent expansion of labels, thereby improving the richness of the label sources. At the same time, the expansion process of the label is stopped.

[0064] In a possible manner, the first label is obtained according to the similar keyword, including: obtaining a verification result of the similar keyword, the verification result being used to represent whether the similar keyword matches the business scenario; and in the case where the verification result represents that the similar keyword matches the business scenario, taking the similar keyword as the first label.

[0065] For example, continuing to refer to Figure 2 The similar keyword can be verified, for example, the similar keyword and the business scenario information are input into a verification model to determine whether the similar keyword matches the business scenario, or manual verification can be performed, and the present disclosure does not limit this. In the case where the verification result represents that the similar keyword matches the business scenario, the similar keyword is determined as a label and stored in a label expansion queue or registered to a label management platform, so that the label is consumed by a downstream label consumption platform. The specific requirements can be set, and the present disclosure does not limit this.

[0066] Alternatively, in the case where the verification result represents that the similar keyword does not match the business scenario, the similar keyword is discarded.

[0067] Through the screening of the similar keyword, the accuracy of the expanded label is further ensured, and the expanded label is prevented from not matching the business scenario, thereby meeting the needs of the business scenario for the label.

[0068] In a possible implementation, the label generation rule is configured in the following manner: a rule creation page is displayed, the rule creation page displays options corresponding to preset rule templates, the preset rule templates are preset for different label generation scenarios, and the preset rule templates are associated with display of description information of corresponding label generation scenarios; in response to a trigger operation on an option corresponding to a target rule template in the preset rule templates based on the label caliber information, a first configuration page corresponding to the target rule template is displayed; and in response to a first configuration operation in the first configuration page based on the label caliber information, the label generation rule is determined according to configuration information corresponding to the first configuration operation.

[0069] For example, as shown in Figure 3 The rule creation page displays options corresponding to preset rule templates, each preset rule template is associated with display of description information of a corresponding label generation scenario, for example, template 1 is used for performing an aggregation operation on offline detailed data as a label value, and the like, which can be set according to requirements, and the disclosure does not limit this. In order to select a required template based on the description information of the template, improve the understandability of the configuration operation, and improve the configuration efficiency of the label generation rule.

[0070] For example, a template meeting a requirement can be selected from the preset rule templates according to the label caliber information, and referring back to Figure 3 Template 1 is selected to display a corresponding rule configuration page, and configuration items such as a label type and a label generation algorithm can be configured, which can be configured according to the label caliber information, and the disclosure does not limit this. Thus, the configuration production of the label can be performed according to the label requirement of a business scenario, and the development efficiency of the label is improved.

[0071] For example, a business requirement needs to adjust an advertising placement strategy according to the sales of product A in the past month, and the set label generation algorithm can perform an aggregation calculation on the sales field according to product A, and the time range is set to one month, to obtain an index label “sales of product A in the past month”. In actual application, the sales of product A is counted through the label, for example, if the sales is greater than a preset sales, the current advertising placement strategy can be maintained, and if the sales is less than the preset sales, the advertising placement strategy needs to be adjusted, and the like, which can be set according to requirements, and the disclosure does not limit this.

[0072] It is worth noting that, as shown in Figure 4As shown, in the related art, a developer develops the computing logic of a label by understanding the label caliber information, and then develops the corresponding data processing link, deploys the corresponding label generation task, and the development cycle is long and the label development efficiency is low. In the embodiment, the configured label generation method can configure the generation rule of the label on the visual configuration page by understanding the label caliber information, and then automatically convert the generation rule of the label into the computing logic of the label, automatically generate the corresponding data processing link and deploy the corresponding label generation task. The complex computing caliber label production can be realized through the configured rule, which is simple to operate, and the real-time and offline label production can be realized based on the accessed data source, thereby effectively improving the development efficiency of the label.

[0073] wherein, Figure 4 The label test can refer to the label verification process, and the specific implementation can be determined according to requirements, which is not limited in the present disclosure.

[0074] In a possible manner, the label generation method further includes: transmitting the quality detection rule corresponding to the second label and the label generation rule to the label management platform, the label management platform being configured to perform quality detection on the business data obtained based on the second label according to the quality detection rule, and performing label management on the second label according to the result of the quality detection; the quality detection rule corresponding to the second label is obtained by configuring as follows: displaying a second configuration page corresponding to the label generation rule, wherein the second configuration page is used to configure the quality detection rule corresponding to the label generation rule, and the second configuration page displays the detection parameter and the alarm threshold corresponding to the detection parameter; in response to a second configuration operation on the detection parameter and / or the alarm threshold in the second configuration page, determining the quality detection rule according to the configuration information corresponding to the second configuration operation.

[0075] It should be noted that in the related art, the quality detection rule of the label is manually configured after the label is registered to the label management platform, which is not efficient.

[0076] In the embodiment, the configuration process of the quality detection rule is integrated into the generation process of the label, as shown in Figure 5 As shown, after the label caliber information is sorted out according to the label requirements, the label generation specification is configured according to the label caliber information, as shown in Figure 3 The "next step" control is triggered, and the configuration page of the quality detection rule is displayed, as shown in Figure 6 After the label generation rule is configured, the corresponding quality detection rule can be further configured, so that the corresponding quality detection rule can be uniformly configured for the label generated based on the label generation rule, thereby greatly improving the configuration efficiency of the quality detection rule.

[0077] For example, as shown in Figure 6As shown, the quality detection rule can be filled with a unified quality detection rule by default, improving the configuration efficiency of the quality detection rule, and can also support modifying the default quality detection rule on demand, such as modifying the detection parameters, the alarm threshold corresponding to the detection parameters, adding a quality detection rule, and deleting a quality detection rule. The specific determination can be made according to the requirements, and the present disclosure does not limit this, thereby meeting different quality detection requirements.

[0078] It is worth noting that different detection intensities of quality monitoring rules can be used for risk labels and non-risk labels. For non-risk labels, because their impact on online business is small, conventional quality detection rules can be used for detection, for example, fine-grained detection can be carried out around the number of data queried through the label, the data fluctuation rate, and the enumeration value coverage rate. For risk labels, because their impact on online business is large, additional heavy protection detection rules can be added on the basis of conventional quality detection rules to strengthen the detection intensity of risk labels. The heavy protection detection rules can include bypass cross detection (i.e., using different data sources / fields to produce the same label through different production links and cross-verify each other) and rollback detection, thereby meeting different intensity quality detection requirements.

[0079] The above-mentioned conventional quality detection rules and heavy protection detection rules can be configured through the quality detection rule configuration page as shown in Figure 6 The conventional quality detection rules do not need to be configured, and the heavy protection detection rules can be configured for labels that need to be detected, and the specific determination can be made according to the requirements, and the present disclosure does not limit this,

[0080] In the embodiments of the present disclosure, the quality detection rule can be used to continuously detect the quality of the label after the label is produced and put online, so that the quality problems existing in the label production link can be found according to the quality detection, and the quality problems can be improved according to the quality problems, thereby further improving the control effect of the label quality.

[0081] In a possible manner, the label requirement information further includes a label type, and the label generation method further includes: in a case where the label type is a first label type, generating first prompt information based on the label requirement information, the first prompt information being used to prompt a first developer to develop a label based on the label requirement information, the first label type representing a label type corresponding to a label with a usage frequency lower than or equal to a preset threshold in a business scenario; and in a case where the label type is a second label type, generating second prompt information based on the label requirement information, the second prompt information being used to prompt a second developer to develop a label based on the label requirement information, the second label type representing a label type corresponding to a label with a usage frequency higher than the preset threshold in the business scenario.

[0082] For example, as shown in Figure 7As shown, for the label with low frequency of use in the business scenario, the first developer can be used for label development, and for the label with high frequency of use in the business scenario, the second developer can be used for label development. The preset threshold can be set according to the needs, and the prompt information can be the corresponding development task, and the present disclosure does not limit this.

[0083] It should be noted that the first developer can be a non-business scenario developer or a developer with low understanding of the business scenario. This part of the developer has a large number of developers and high development efficiency, but the label quality is low. The second developer can be a business scenario developer or a developer who understands the business scenario. This part of the developer is familiar with the business scenario, but the number is small, the development efficiency is low, but the matching degree of the label developed by the second developer with the business scenario is higher, and the label quality is higher.

[0084] In the related art, pure first developers or pure second developers are usually used for label development. Although the division of labor is clear, the labels developed by the pure first developers have obvious quality problems, and the pure second developers are limited by the development efficiency and cannot meet the label requirements of the business.

[0085] Since the labels with high frequency have a wider impact on the business scope when the quality problem occurs, the labels with low frequency can be developed by the first developer, and the labels with high frequency can be developed by the second developer, thereby reducing the risk of business abnormalities caused by label quality problems. Further, the first developer and the second developer jointly complete the label development, which effectively improves the label development efficiency while ensuring the label quality and meets the label requirements of the business.

[0086] In a possible manner, the label generation method further includes: obtaining a target label developed by the first developer and / or the second developer; determining the business object information corresponding to the target label according to the label requirement information, and the business object information is used for quality management of the target label.

[0087] For example, continuing to refer to Figure 7 The corresponding quality management personnel can be specified for each label entity, for example, the corresponding quality management personnel is specified for the commodity label entity. In this way, when the business object information corresponding to the label developed by the first developer and the second developer determines that the label belongs to the commodity label entity, the corresponding quality management personnel is used for unified quality management of the label. In this way, whether the label is developed by the first developer or the second developer, unified quality management can be realized, and the uniformity of the label quality is ensured.

[0088] Through unified quality management of the development of labels according to use frequency, the label quality can be ensured, and the label can be quickly put online to meet the label demand of the business. The subsequent label consumption platform can produce data based on the label, which is not described herein again.

[0089] By using the above method, the development efficiency of the label is significantly improved through automatic expansion of the rule label, production of the label in a configuration mode, and development and unified quality management according to use frequency. The label satisfaction rate of the business is improved, the difficulty and cycle of label development and online are reduced, the contradiction between the increasing label demand of the business and the low label development efficiency is solved through overall efficiency improvement, and the business growth is better assisted.

[0090] Based on the same concept, the embodiment of the disclosure also provides a label generation device based on a large model and a configuration mode, as shown in Figure 8 The label generation device 800 can include:

[0091] The acquisition module 801 is configured to acquire label demand information of a label required by a business scenario.

[0092] The generation module 802 is configured to generate a label based on the label demand information through at least one of the following ways:

[0093] In a case where the label demand information includes a preset keyword, a large model is used to perform semantic understanding on the preset keyword to generate a similar keyword similar in semantics to the preset keyword; and a first label is obtained according to the similar keyword.

[0094] In a case where the label demand information includes label caliber information, a second label is generated according to a label generation rule corresponding to the label caliber information, wherein the label generation rule is determined in response to a first configuration operation based on the label caliber information in a first configuration page, and the first configuration page is used to configure the generation rule of the label.

[0095] Optionally, the generation module 802 is configured to:

[0096] The large model is used to perform semantic understanding on the preset keyword, to obtain a target long tail keyword with a semantic similarity greater than a preset similarity from a target data source, and to determine the target long tail keyword as the similar keyword, wherein the target data source is determined based on scene information representing the business scenario in the label demand information.

[0097] Optionally, the label generation device 800 further includes a storage module, and the storage module is configured to:

[0098] In a case where the preset keyword does not exist in the target long tail keyword, the preset keyword is stored in a candidate label pool, and the target data source comprises the candidate label pool.

[0099] Optionally, the generation module 802 is configured to:

[0100] obtain a verification result of the similar keyword, the verification result being used to represent whether the similar keyword matches the business scenario;

[0101] In a case where the verification result represents that the similar keyword matches the business scenario, the similar keyword is taken as the first label.

[0102] Optionally, the label generation rule is configured in the following manner:

[0103] display a rule creation page, wherein the rule creation page displays options corresponding to preset rule templates, the preset rule templates being preset for different label generation scenarios, and the preset rule templates are associated with display of description information of corresponding label generation scenarios;

[0104] in response to a trigger operation of a target rule template corresponding option in the preset rule templates based on the label caliber information, display a first configuration page corresponding to the target rule template;

[0105] in response to a first configuration operation in the first configuration page based on the label caliber information, determine the label generation rule according to configuration information corresponding to the first configuration operation.

[0106] Optionally, the label generation apparatus 800 further comprises a transmission module, configured to:

[0107] transmit the second label and a quality detection rule corresponding to the label generation rule to a label management platform, the label management platform being configured to perform quality detection on business data acquired based on the second label according to the quality detection rule, and perform label management on the second label according to a result of the quality detection;

[0108] the quality detection rule corresponding to the second label is configured in the following manner:

[0109] display a second configuration page corresponding to the label generation rule, wherein the second configuration page is used to configure a quality detection rule corresponding to the label generation rule, and the second configuration page displays a detection parameter and an alarm threshold corresponding to the detection parameter;

[0110] In response to a second configuration operation on the detection parameter and / or the alarm threshold in the second configuration page, the quality detection rule is determined according to configuration information corresponding to the second configuration operation.

[0111] Optionally, the label requirement information further comprises a label type, and the label generation apparatus 800 further comprises a development module, configured to:

[0112] In a case where the label type is a first label type, generate first prompt information based on the label requirement information, the first prompt information being used to prompt a first developer to develop a label based on the label requirement information, the first label type representing a label type corresponding to a label with a usage frequency lower than or equal to a preset threshold in the business scenario;

[0113] In a case where the label type is a second label type, generate second prompt information based on the label requirement information, the second prompt information being used to prompt a second developer to develop a label based on the label requirement information, the second label type representing a label type corresponding to a label with a usage frequency higher than a preset threshold in the business scenario.

[0114] Optionally, the label generation apparatus 800 further comprises a determination module, configured to:

[0115] obtain a target label developed by the first developer and / or the second developer;

[0116] determine, according to the label requirement information, business object information corresponding to the target label, the business object information being used for quality management of the target label.

[0117] Based on the same idea, the embodiments of the present disclosure further provide a computer readable medium having a computer program stored thereon, the program being executed by a processing device to implement the steps of any of the above label generation methods based on a large model and configuration.

[0118] Based on the same idea, the embodiments of the present disclosure further provide an electronic device, which can include:

[0119] a storage device having a computer program stored thereon;

[0120] a processing device configured to execute the computer program in the storage device to implement the steps of any of the above label generation methods based on a large model and configuration.

[0121] Based on the same idea, the embodiments of the present disclosure further provide a computer program product, comprising a computer program, the computer program being executed by a processor to implement the steps of any of the above label generation methods based on a large model and configuration.

[0122] Reference will now be made to Figure 9 which shows a structural diagram of an electronic device 900 suitable for use in implementing embodiments of the present disclosure. The terminal device in embodiments of the present disclosure can include, but is not limited to, a mobile terminal such as a mobile phone, a notebook computer, a digital broadcast receiver, a PDA (Personal Digital Assistant), a PAD (Tablet Personal Computer), a PMP (Portable Multimedia Player), a car terminal (e.g., a car navigation terminal), and the like, as well as a stationary terminal such as a digital TV, a desktop computer, and the like. Figure 9 The illustrated electronic device is merely an example and should not impose any limitation on the functions and the range of use of embodiments of the present disclosure.

[0123] As shown in Figure 9 , the electronic device 900 can include a processing device (e.g., a central processor, a graphic processor, etc.) 901 that can perform various appropriate actions and processes according to programs stored in a read-only memory (ROM) 902 or loaded into a random access memory (RAM) 903 from a storage device 908. In the RAM 903, various programs and data required for the operation of the electronic device 900 are also stored. The processing device 901, the ROM 902, and the RAM 903 are connected to each other through a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0124] In general, the following devices can be connected to the I / O interface 905: input devices 906 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, and the like; output devices 907 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, and the like; storage devices 908 including, for example, a magnetic tape, a hard disk, and the like; and communication devices 909. The communication devices 909 can allow the electronic device 900 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 9 The electronic device 900 is shown with various devices, but it is understood that not all of the shown devices are required to be implemented or present. More or fewer devices can alternatively be implemented or present.

[0125] In particular, the processes described above with reference to the flowcharts can be implemented as a computer software program according to embodiments of the present disclosure. For example, embodiments of the present disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods illustrated by the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network through the communication devices 909, or installed from the storage devices 908, or installed from the ROM 902. When the computer program is executed by the processing device 901, the above-described functions defined in the methods of the present disclosure are performed.

[0126] Note that the computer readable medium described above in the present disclosure can be a computer readable signal medium or a computer readable storage medium or any combination thereof. The computer readable storage medium can be, for example and without limitation, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or any suitable combination thereof. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present disclosure, the computer readable storage medium can be any tangible medium that contains or stores a program used by or in connection with an instruction execution system, apparatus or device. In the present disclosure, the computer readable signal medium can include a data signal propagated in baseband or propagated as a carrier wave in a propagated data signal, in which the computer readable program code is contained. Such a propagated data signal can take any of a variety of forms, including but not limited to electro-magnetic, optical, or any suitable combination thereof. The computer readable signal medium can also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate or transport a program for use by or in connection with an instruction execution system, apparatus or device. The program code contained on the computer readable medium can be transmitted by any suitable medium, including but not limited to wire, cable, RF, etc., or any suitable combination thereof.

[0127] In some embodiments, communication can be conducted using any currently known or future developed network protocol, such as HTTP (HyperText Transfer Protocol), and can be interconnected with digital data communication (e.g., a communication network) of any form or medium, such as the Internet. Examples of communication networks include local area networks ("LAN"), wide area networks ("WAN"), internetworks (e.g., the Internet), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future developed networks.

[0128] The computer readable medium described above can be included in the electronic device described above; or can exist separately from the electronic device without being incorporated into the electronic device.

[0129] The computer readable medium described above carries one or more programs, when the one or more programs are executed by the electronic device, cause the electronic device to: obtain label demand information of a label required by a business scenario; and generate a label based on the label demand information by at least one of the following: in a case where the label demand information includes a preset keyword, performing semantic understanding on the preset keyword by a large model to generate a similar keyword similar in semantics to the preset keyword; obtaining a first label according to the similar keyword; and in a case where the label demand information includes label caliber information, generating a second label according to a label generation rule corresponding to the label caliber information, wherein the label generation rule is determined in response to a first configuration operation based on the label caliber information in a first configuration page used for configuring a generation rule of a label.

[0130] Computer program code for carrying out operations of the present disclosure can be written in any one or more of a variety of programming languages or combinations of languages, including an object-oriented programming language such as Java, Smalltalk, C++, or the like, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0131] The flow diagrams and the block diagrams in the drawings are illustrations of architectures, functionalities, and operations of possible implementations of systems, methods, and computer program products according to various embodiments of present disclosure. In this regard, each block in the flow diagrams or block diagrams can represent a module, a procedure, or a portion of code, which comprises one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may

[0132] The modules described in the embodiments of this disclosure can be implemented in software or hardware. The names of the modules are not, in some cases, intended to limit the functionality of the module itself.

[0133] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0134] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0135] The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features disclosed in this disclosure that have similar functions.

[0136] Furthermore, while the operations are described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or in a sequential order. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.

[0137] Although the subject matter has been described in language specific to structural features and / or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims. With respect to the devices in the above-described embodiments, in which various modules perform operations, the specific manner in which the various modules perform the operations has been described in detail in the embodiments relating to the method. Here, no detailed explanation will be given.

Claims

1. A tag generation method based on large models and configurability, characterized in that, The tag generation method includes: Obtain the tag requirement information for the business scenario; Based on the tag requirement information, tags are generated using at least one of the following methods: When the tag requirement information includes preset keywords, the preset keywords are semantically understood using the large model to obtain initial long-tail keywords from the target data source; target long-tail keywords with a semantic similarity greater than the preset similarity to the preset keywords are selected from the initial long-tail keywords based on a preset similarity algorithm; a verification result for the target long-tail keywords is obtained, which is used to characterize whether the target long-tail keywords match the business scenario; if the verification result indicates that the target long-tail keywords match the business scenario, the target long-tail keywords are used as the first tag; wherein, the target data source is determined based on the scenario information characterizing the business scenario in the tag requirement information; When the label requirement information includes label caliber information, a second label is generated according to the label generation rule corresponding to the label caliber information. The label generation rule is determined in response to a first configuration operation based on the label caliber information in a first configuration page. The first configuration page is used to configure the label generation rule, and the configuration items of the first configuration page include the label generation algorithm. The tag generation rules are configured as follows: The rule creation page displays options corresponding to preset rule templates. These preset rule templates are pre-defined for different tag generation scenarios, and they are associated with descriptions of the corresponding tag generation scenarios. In response to a trigger operation on the option corresponding to the target rule template in the preset rule template based on the label caliber information, the first configuration page corresponding to the target rule template is displayed; In response to a first configuration operation on the first configuration page based on the label caliber information, the label generation rule is determined according to the configuration information corresponding to the first configuration operation; The label generation method further includes: The second tag and the quality inspection rule corresponding to the tag generation rule are transmitted to the tag management platform. The tag management platform is used to perform quality inspection on the business data obtained based on the second tag according to the quality inspection rule, and to manage the second tag according to the result of the quality inspection. The quality inspection rules corresponding to the second label are configured as follows: The second configuration page corresponding to the tag generation rule is displayed. The second configuration page is used to configure the quality detection rule corresponding to the tag generation rule, and the second configuration page displays the detection parameters and the alarm thresholds corresponding to the detection parameters. In response to a second configuration operation on the second configuration page for the detection parameters and / or the alarm threshold, the quality detection rule is determined based on the configuration information corresponding to the second configuration operation.

2. The tag generation method based on a large model and configurability according to claim 1, characterized in that, The label generation method further includes: If the target long-tail keyword does not exist in the preset keyword, the preset keyword is stored in the candidate tag pool, wherein the target data source includes the candidate tag pool.

3. The tag generation method based on a large model and configurability according to claim 1 or 2, characterized in that, The label requirement information also includes label type, and the label generation method further includes: When the tag type is a first tag type, a first prompt message is generated based on the tag requirement information. The first prompt message is used to prompt the first developer to develop tags based on the tag requirement information. The first tag type represents the tag type corresponding to a tag whose usage frequency in the business scenario is lower than or equal to a preset threshold. When the tag type is the second tag type, a second prompt message is generated based on the tag requirement information. The second prompt message is used to prompt the second developer to develop tags based on the tag requirement information. The second tag type represents the tag type corresponding to the tag whose usage frequency in the business scenario is higher than a preset threshold.

4. The tag generation method based on a large model and configurability according to claim 3, characterized in that, The label generation method further includes: Obtain the target tags developed by the first developer and / or the second developer; Based on the tag requirement information, the business object information corresponding to the target tag is determined, and the business object information is used for quality management of the target tag.

5. A tag generation device based on a large model and configurability, characterized in that, The label generation device includes: The acquisition module is used to acquire tag requirement information for the tags needed in the business scenario; The generation module is used to generate tags based on the tag requirement information using at least one of the following methods: When the tag requirement information includes preset keywords, the preset keywords are semantically understood using the large model to obtain initial long-tail keywords from the target data source; target long-tail keywords with a semantic similarity greater than the preset similarity to the preset keywords are selected from the initial long-tail keywords based on a preset similarity algorithm; a verification result for the target long-tail keywords is obtained, which is used to characterize whether the target long-tail keywords match the business scenario; if the verification result indicates that the target long-tail keywords match the business scenario, the target long-tail keywords are used as the first tag; wherein, the target data source is determined based on the scenario information characterizing the business scenario in the tag requirement information; When the label requirement information includes label caliber information, a second label is generated according to the label generation rule corresponding to the label caliber information. The label generation rule is determined in response to a first configuration operation based on the label caliber information in a first configuration page. The first configuration page is used to configure the label generation rule, and the configuration items of the first configuration page include the label generation algorithm. The tag generation rules are configured as follows: a rule creation page is displayed, wherein the rule creation page displays options corresponding to a preset rule template, the preset rule template is pre-set for different tag generation scenarios, and the preset rule template is associated with description information of the corresponding tag generation scenario; in response to a trigger operation on the option corresponding to the target rule template in the preset rule template based on the tag caliber information, a first configuration page corresponding to the target rule template is displayed; in response to a first configuration operation on the first configuration page based on the tag caliber information, the tag generation rules are determined according to the configuration information corresponding to the first configuration operation. The tag generation device further includes a transmission module, which is used to transmit the second tag and the quality detection rule corresponding to the tag generation rule to the tag management platform. The tag management platform is used to perform quality detection on the business data obtained based on the second tag according to the quality detection rule, and to perform tag management on the second tag according to the result of the quality detection. The quality detection rule corresponding to the second label is configured as follows: a second configuration page corresponding to the label generation rule is displayed, wherein the second configuration page is used to configure the quality detection rule corresponding to the label generation rule, and the second configuration page displays detection parameters and alarm thresholds corresponding to the detection parameters; in response to a second configuration operation on the detection parameters and / or the alarm thresholds in the second configuration page, the quality detection rule is determined according to the configuration information corresponding to the second configuration operation.

6. A computer-readable medium having a computer program stored thereon, characterized in that, When executed by a processing device, the computer program performs the steps of the method described in any one of claims 1-4.

7. An electronic device, characterized in that, include: A storage device on which computer programs are stored; A processing device for executing the computer program in the storage device to implement the steps of the method according to any one of claims 1-4.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-4.

Images