Object data determination method and device, storage medium and electronic device

Abstract

The invention discloses an object data determination method and device, a storage medium and an electronic device, and relates to the field of big data, and the method comprises the steps: obtaining marking labels from marking tasks for marking different objects, sending the marking labels and an index database to a first agent, the first intelligent agent performs semantic understanding on the marking label and generates a first object feature consistent with an index format in an index database; screening the first object features based on a preset feature database to obtain second object features; sending the marking label and the second object feature to a second intelligent agent, so that the second intelligent agent performs semantic understanding on the marking label and generates an execution instruction corresponding to the marking label according to the second object feature; and executing the marking task according to the execution instruction to obtain object data for classifying different objects. By adopting the technical scheme, the problem of how to improve the efficiency of marking people is solved.

Description

Technical Field

[0001] This application relates to the field of big data, and more specifically, to a method, apparatus, storage medium, and electronic device for determining object data. Background Technology

[0002] Currently, when enterprises target their operations or marketing departments, they need to obtain user groups with specific tags from audience data to send campaign information, aiming to improve user engagement or activity. The process of tagging and obtaining these user groups typically involves coordinating business personnel for business analysis, data engineers for data exploration and modeling, and algorithm engineers for algorithm design and computation. This process is lengthy and inefficient. Therefore, there is a challenge in related technologies to improve the efficiency of audience tagging.

[0003] Regarding the issue of how to improve the efficiency of tagging populations in related technologies, no effective solution has yet been proposed. Summary of the Invention

[0004] This application provides a method, apparatus, storage medium, and electronic device for determining object data, in order to at least solve the problem of how to improve the efficiency of tagging people in related technologies.

[0005] According to one embodiment of this application, a method for determining object data is provided, comprising: obtaining tagging labels from a tagging task that tags different objects; sending the tagging labels and an indicator database to a first intelligent agent, so that the first intelligent agent performs semantic understanding on the tagging labels and generates a first object feature consistent with the indicator format in the indicator database, wherein the first object feature is an object feature that has a semantic association with the tagging labels; filtering the first object feature based on a preset feature database to obtain a second object feature, wherein the second object feature matches the object features in the preset feature database; sending the tagging labels and the second object feature to a second intelligent agent, so that the second intelligent agent performs semantic understanding on the tagging labels and generates an execution instruction corresponding to the tagging labels based on the second object feature; and executing the tagging task according to the execution instruction to obtain object data that classifies the different objects.

[0006] In an exemplary embodiment, sending the tagging labels and indicator database to a first intelligent agent, so that the first intelligent agent performs semantic understanding on the tagging labels and generates a first object feature consistent with the indicator format in the indicator database, includes: constructing the first intelligent agent through a large language model and a first prompt word, wherein the first prompt word is used to instruct the large language model to generate an object feature that is semantically related to the tagging labels after performing semantic understanding on the tagging labels, and the format of the object feature is consistent with the indicator format in the indicator database; inputting the tagging labels and indicator information from the indicator database into the first intelligent agent to obtain the first object feature output by the first intelligent agent.

[0007] In an exemplary embodiment, filtering the first object features based on a preset feature database to obtain second object features includes: vectorizing each first object feature to obtain multiple first vectors; vectorizing each feature in the preset feature database to obtain multiple second vectors; calculating the vector similarity between each first vector and each second vector; and determining the second object feature from the first object features based on the vector similarity, wherein the vector similarity between the first vector corresponding to the second object feature and any second vector is greater than a preset vector similarity.

[0008] In one exemplary embodiment, sending the tag and the second object feature to a second agent, so that the second agent performs semantic understanding of the tag and generates an execution instruction corresponding to the tag based on the second object feature, includes: constructing the second agent through a large language model and a second prompt word, wherein the second prompt word is used to instruct the large language model to generate an execution instruction related to the second object feature after performing semantic understanding of the tag; inputting the tag and the second object feature into the first agent to obtain the execution instruction output by the first agent.

[0009] In an exemplary embodiment, after obtaining object data that classifies the different objects, the method further includes: updating the push group using the object data and saving the push group to a database, wherein the push group contains push objects for business activities; in response to a business activity push instruction, selecting a target push object indicated by the business activity push instruction from the push group and pushing the business activity to the target push object.

[0010] In an exemplary embodiment, after saving the groups to be pushed to the database, the method further includes: checking the groups to be pushed in the database according to a preset period, determining overdue push groups from the database, wherein the overdue push groups refer to the groups to be pushed whose last update time is greater than a preset time period; generating new push groups based on the tags corresponding to the overdue push groups and the latest original object data; and replacing the overdue push groups in the database with the new push groups.

[0011] In an exemplary embodiment, after obtaining object data obtained by classifying the different objects according to the tagging results, the method further includes: performing attribute distribution statistics on the object data according to preset data attributes to obtain distribution data of the object data; and displaying the distribution data of the object data on a visualization interface in the form of a preset chart.

[0012] According to another aspect of the embodiments of this application, an object data determination apparatus is also provided, comprising: a first generation module, configured to obtain tagging labels from a tagging task that tags different objects, and send the tagging labels and an indicator database to a first intelligent agent, so that the first intelligent agent performs semantic understanding on the tagging labels and generates a first object feature consistent with the indicator format in the indicator database, wherein the first object feature is an object feature that has a semantic association with the tagging labels; a filtering module, configured to filter the first object feature based on a preset feature database to obtain a second object feature, wherein the second object feature is an object feature existing in the preset feature database; a second generation module, configured to send the tagging labels and the second object feature to a second intelligent agent, so that the second intelligent agent performs semantic understanding on the tagging labels and generates an execution instruction corresponding to the tagging labels according to the second object feature; and an execution module, configured to execute the tagging task according to the execution instruction to obtain object data that classifies the different objects.

[0013] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, wherein a computer program is stored in the computer program, and the computer program is configured to execute the above-described method for determining object data at runtime.

[0014] According to another aspect of the embodiments of this application, an electronic device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the above-described method for determining object data through the computer program.

[0015] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps in any of the above method embodiments.

[0016] In this embodiment, the tagging labels and indicator database obtained from the tagging task can be sent to a first intelligent agent. The first intelligent agent performs semantic understanding on the tagging labels and generates a first object feature with the same indicator format as the indicator in the indicator database. Then, a second object feature matching the object features in a preset feature database is selected from the first object feature. After the second intelligent agent performs semantic understanding on the tagging labels, it generates corresponding execution instructions based on the second object features. The execution instructions are then used to execute the target-achieving task to obtain object data categorized for different objects. This technical solution solves the problem of how to improve the efficiency of tagging populations, thereby achieving the effect of improving the efficiency of population tagging. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the hardware environment for a method of determining object data according to an embodiment of this application.

[0020] Figure 2 This is a flowchart of a method for determining object data according to an embodiment of this application;

[0021] Figure 3 This is a schematic diagram of a method for determining object data according to an embodiment of this application;

[0022] Figure 4 This is a structural block diagram of an object data determination device according to an embodiment of this application. Detailed Implementation

[0023] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0024] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0025] According to one aspect of the embodiments of this application, a method for determining object data is provided. This method for determining object data is widely used in application scenarios such as software development. Optionally, in this embodiment, the above-described method for determining object data can be applied to, for example... Figure 1 The hardware environment shown consists of terminal device 102 and server 104. For example... Figure 1 As shown, server 104 is connected to terminal device 102 via a network and can be used to provide services (such as application services) to the terminal or clients installed on the terminal. A database can be set up on the server or independently of the server to provide data storage services for server 104. Cloud computing and / or edge computing services can be configured on the server or independently of the server to provide data processing services for server 104.

[0026] The aforementioned network may include, but is not limited to, at least one of the following: wired network, wireless network. The aforementioned wired network may include, but is not limited to, at least one of the following: wide area network, metropolitan area network, local area network. The aforementioned wireless network may include, but is not limited to, at least one of the following: Wi-Fi (Wireless Fidelity), Bluetooth. The terminal device 102 may not be limited to PC, mobile phone, tablet computer, smart air conditioner, smart range hood, smart refrigerator, smart oven, smart stove, smart washing machine, smart water heater, smart washing equipment, smart dishwasher, smart projector, smart TV, smart clothes rack, smart curtains, smart audio-visual equipment, smart socket, smart speaker, smart speaker box, smart fresh air equipment, smart kitchen and bathroom equipment, smart bathroom equipment, smart robot vacuum cleaner, smart window cleaning robot, smart mopping robot, smart air purifier, smart steam oven, smart microwave oven, smart water heater, smart air purifier, smart water dispenser, smart door lock, etc.

[0027] This embodiment provides a method for determining object data, applied to the aforementioned terminal device. Figure 2 This is a flowchart of a method for determining object data according to an embodiment of this application, the process including the following steps:

[0028] Step S202: Obtain tagging labels from tagging tasks that tag different objects, and send the tagging labels and indicator database to the first intelligent agent so that the first intelligent agent can perform semantic understanding of the tagging labels and generate a first object feature that is consistent with the indicator format in the indicator database. The first object feature is an object feature that has a semantic association with the tagging labels.

[0029] Optionally, in step S202 above, the tagging task is based on the selection input by the user. For example, if the operator inputs "I need to find female users over 30 years old who frequently buy maternity and baby products" on the interactive page, then the received information is semantically parsed through a large language model, and the extracted tags include "frequently buy maternity and baby products", "over 30 years old", "female", etc.

[0030] Optionally, in step S202 above, the indicators in the indicator database include indicator names, indicator meanings, indicator calculation rules, etc. For example, the indicator name is the number of purchases, the indicator meaning is the total number of orders of the user, and the calculation rule is to sum the number of orders of the user.

[0031] Step S204: Filter the first object features based on a preset feature database to obtain second object features, wherein the second object features match the object features in the preset feature database;

[0032] Step S206: Send the tag and the second object feature to the second intelligent agent so that the second intelligent agent can perform semantic understanding of the tag and generate an execution instruction corresponding to the tag based on the second object feature;

[0033] Step S208: Execute the marking task according to the execution instruction to obtain object data for classifying the different objects.

[0034] Through the above steps, the tagging labels and indicator database obtained from the tagging task can be sent to the first intelligent agent. The first intelligent agent performs semantic understanding on the tagging labels and generates first object features that are consistent with the indicator format in the indicator database. Then, second object features that match the object features in the preset feature database are selected from the first object features. After the second intelligent agent performs semantic understanding on the tagging labels, it generates corresponding execution instructions based on the second object features. The execution instructions are then used to execute the target-achieving task to obtain object data categorized for different objects. This technical solution solves the problem of how to improve the efficiency of tagging populations, thereby achieving the effect of improving the efficiency of population tagging.

[0035] In an exemplary embodiment, the specific implementation process of sending the tag and indicator database to the first agent in step S202 above, so that the first agent performs semantic understanding on the tag and generates a first object feature consistent with the indicator format in the indicator database, includes: constructing the first agent through a large language model and a first prompt word, wherein the first prompt word is used to instruct the large language model to generate an object feature that is semantically related to the tag after performing semantic understanding on the tag, and the format of the object feature is consistent with the indicator format in the indicator database; inputting the tag and indicator information from the indicator database into the first agent to obtain the first object feature output by the first agent.

[0036] Optionally, in the above embodiments, a pre-trained large language model is first used as the foundation, combined with a first prompt word (e.g., "Please convert the following tags into feature descriptions that match the indicator database.") to construct a first intelligent agent. The task of this agent is to understand the meaning of the tags and convert them into a format consistent with the indicator database. For example, the input tags might be "frequently purchases maternity and baby products," "over 30 years old," and "female." Meanwhile, the indicator database may contain indicators such as "age," "number of purchases," and "product categories." These tags and indicator information are input into the first intelligent agent. After semantic understanding by the first intelligent agent, first object features consistent with the format of the indicator database are generated. For example, "frequently purchases maternity and baby products" will generate features related to this tag, such as "number of purchases" and "product categories."

[0037] In an exemplary embodiment, the process of filtering the first object features based on a preset feature database to obtain a second object feature table in step S204 can be implemented as follows: each first object feature is vectorized to obtain multiple first vectors; each feature in the preset feature database is vectorized to obtain multiple second vectors; the vector similarity between each first vector and each second vector is calculated, and the second object feature is determined from the first object features based on the vector similarity, wherein the vector similarity between the first vector corresponding to the second object feature and any second vector is greater than a preset vector similarity.

[0038] Optionally, in the above embodiments, since the first object features generated by the first intelligent agent may have model illusions, the generated features may not necessarily be determined by the collected population data. That is, the population data does not cover all aspects of data, and the features that can be determined based on the population data are limited. The features in the preset feature database are features that can be determined by the existing population data. This screening step reduces noise and avoids object features that are unrelated or weakly related to the preset features from entering the next step, thereby improving the accuracy and effectiveness of subsequent instruction generation.

[0039] Optionally, in the above embodiments, vectorizing both the first object features and the features in the preset feature database can be performed using models such as Word2Vec or BERT to convert text into numerical vectors. The similarity between the generated first vector set and the second vector set in the preset database is calculated; this step can use cosine similarity to measure whether the directions of the two vectors are similar. A threshold is set; when the similarity between two vectors is higher than this threshold, the two vectors are considered semantically sufficiently close. Then, features with high similarity to features in the preset feature database are selected from the first object features; these are the second object features.

[0040] In an exemplary embodiment, the process of sending the tag and the second object feature to the second agent in step S206, so that the second agent can perform semantic understanding of the tag and generate an execution instruction corresponding to the tag based on the second object feature, specifically includes: constructing the second agent through a large language model and a second prompt word, wherein the second prompt word is used to instruct the large language model to generate an execution instruction related to the second object feature after performing semantic understanding of the tag; inputting the tag and the second object feature into the first agent to obtain the execution instruction output by the first agent.

[0041] Optionally, in the above embodiments, the creation of the second intelligent agent first uses a pre-trained large language model as a foundation, combined with a first prompt word, such as: "For the following object characteristics, please provide an execution instruction to determine the object data related to the tag." For example, the input tag may be "frequently purchases maternal and infant products," "over 30 years old," or "female." The second object characteristics may be "number of purchases," "category of purchases," "gender," or "age." Then, the second intelligent agent, through semantic understanding, defines "frequently purchases maternal and infant products" as purchasing maternal and infant products more than 3 times a month, and then generates an execution instruction. For example, this instruction may search for people in the database through code. For example, the code logic may be to search for users whose gender field is female, whose age field is greater than 30, and whose number of purchases of maternal and infant products in their order data is more than 3 times.

[0042] In an exemplary embodiment, after obtaining object data for classifying the different objects by performing the above step S208, the method further includes: updating the push group using the object data and saving the push group to a database, wherein the push group contains push objects for business activities; in response to a business activity push instruction, selecting the target push object indicated by the business activity push instruction from the push group and pushing the business activity to the target push object.

[0043] Optionally, in the above embodiments, the filtered data of different objects can be saved as push groups. For example, push groups may include "female users over 30 years old who frequently buy maternal and infant products" or "male users who frequently exercise". Operators can push business activities to users corresponding to the selected push groups, thereby achieving precise push and improving the operational efficiency of business activities.

[0044] In an exemplary embodiment, after saving the groups to be pushed to the database, the method further includes: checking the groups to be pushed in the database according to a preset period, determining overdue push groups from the database, wherein the overdue push groups refer to the groups to be pushed whose last update time is greater than a preset time period; generating new push groups based on the tags corresponding to the overdue push groups and the latest original object data; and replacing the overdue push groups in the database with the new push groups.

[0045] Optionally, in the above embodiments, the groups to be pushed are time-sensitive, and user data will change over time. Therefore, the groups to be pushed need to be updated regularly to ensure that the groups to be pushed can match the tags corresponding to the tagging task, so as to achieve the goal of improving operational effectiveness.

[0046] In an exemplary embodiment, after performing the above step S206 to obtain object data obtained by classifying the different objects according to the tagging results, the method further includes: performing attribute distribution statistics on the object data according to preset data attributes to obtain distribution data of the object data; and displaying the distribution data of the object data on a visualization interface in the form of a preset chart.

[0047] Optionally, in the above embodiments, the tagged users can also be visualized, for example, displayed according to user attributes such as geographical distribution, gender, and age group. The display method can be a bar chart, pie chart, etc., to facilitate the operation personnel to understand the attribute characteristics of the tagged users.

[0048] To better understand the process of determining the object data described above, the implementation flow of the method for determining the object data will be described below in conjunction with optional embodiments, but this is not intended to limit the technical solution of the embodiments of this application.

[0049] In an optional embodiment, Figure 3 This is a schematic diagram of a method for determining object data according to an embodiment of this application, such as... Figure 3 As shown, the specific process includes the following steps:

[0050] Step 1: The user creates a group selection task (equivalent to a tagging task) and enters the topic to be selected.

[0051] For example, an operations staff member can create a new user segmentation task on the interactive interface and enter the segmentation topic as "users with high spending power who are interested in environmentally friendly products".

[0052] Step 2.1: Pass the metadata of the theme "users with high spending power and interest in environmentally friendly products" and the indicator database to the feature selection agent (a software module that performs a specific task, equivalent to the first intelligent agent mentioned above).

[0053] Optionally, when transmitting the selected topic in the above steps, you can either transmit the original topic text entered by the user, or you can decompose the original topic text into multiple tag texts before transmitting it.

[0054] Step 2.2: Pass the indicator name (such as "average consumption amount" or "number of times environmental protection products are purchased"), indicator meaning, and indicator calculation rules to the RAG (Retrieval-Augmented Generation) module.

[0055] Step 2.3: Pass the audience selection topic to the audience selection agent (equivalent to the second intelligent agent mentioned above).

[0056] Step 3.1: Feature Selection Agent receives the selected topic and the metadata of the indicator library. Using semantic understanding capabilities, it generates a preliminary feature list (equivalent to the first object features). The features in this feature list are those that the first agent believes may be related to the selected topic, such as "spending amount", "shopping category", and "shopping frequency".

[0057] Step 3.2: Pass the preliminary feature list to the RAG module.

[0058] Step 4.1: The RAG module receives the initial feature list and uses vector matching technology to find more accurate indicators in the indicator library. After filtering, indicator features that do not exist in the indicator library are removed, generating a more accurate feature list (equivalent to the second object features).

[0059] Step 4.2: Pass this high-precision feature list to the crowd selection agent.

[0060] Step 5.1: Audience Selection Agent understands the selected topics and the second feature list.

[0061] Step 5.2: Construct the data mining steps and generate data mining code.

[0062] Step 5.3: Submit the crowd selection task to the task scheduling module.

[0063] Step 6.1: The task scheduling module monitors the progress of the crowd selection task and pushes the crowd list data.

[0064] Step 6.2: The task scheduling module monitors the execution progress of the crowd selection task to ensure that the task is completed on time.

[0065] Step 6.3: Once the task is completed, push the generated crowd list data to the BI (Business Intelligence) system.

[0066] Step 7: The BI system receives the audience list data, which includes detailed information on selected users with high spending power and an interest in environmentally friendly products. Based on this data, the BI system generates audience distribution charts, such as user geographic distribution maps and consumption habit charts. These charts are then presented to operations staff to help develop more effective marketing strategies.

[0067] Through the above embodiments, intelligent tagging of users and generation of audience packages are achieved, optimizing the traditional audience tagging and selection process, improving marketing and operational efficiency, reducing operating costs, improving the accuracy of audience tagging, and enhancing user activity and stickiness.

[0068] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of this application.

[0069] Figure 4 This is a structural block diagram of an object data determination device according to an embodiment of this application; as shown below. Figure 4 As shown, it includes:

[0070] The first generation module 42 is used to obtain tagging labels from tagging tasks that tag different objects, and send the tagging labels and indicator database to the first intelligent agent so that the first intelligent agent can perform semantic understanding of the tagging labels and generate a first object feature that is consistent with the indicator format in the indicator database. The first object feature is an object feature that has a semantic association with the tagging labels.

[0071] The filtering module 44 is used to filter the first object features based on a preset feature database to obtain the second object features, wherein the second object features are object features existing in the preset feature database;

[0072] The second generation module 46 is used to send the tag and the second object feature to the second intelligent agent, so that the second intelligent agent can perform semantic understanding of the tag and generate an execution instruction corresponding to the tag based on the second object feature.

[0073] The execution module 48 is used to execute the marking task according to the execution instructions to obtain object data for classifying the different objects.

[0074] The aforementioned device allows the sending of tagging labels and an indicator database obtained from a tagging task to a first intelligent agent. The first agent performs semantic understanding on the tagging labels and generates first object features with the same format as the indicators in the indicator database. Then, it filters out second object features from the first object features that match object features in a preset feature database. Finally, the second agent performs semantic understanding on the tagging labels and generates corresponding execution instructions based on the second object features. These instructions are then used to execute the target-achieving task, resulting in object data categorized into different objects. This technical solution addresses the problem of improving the efficiency of tagging populations, thereby enhancing the overall efficiency of population tagging.

[0075] In an exemplary embodiment, the first generation module 42 is further configured to construct the first intelligent agent using a large language model and a first prompt word, wherein the first prompt word is used to instruct the large language model to generate object features that are semantically related to the tag after performing semantic understanding on the tag, and the format of the object features is consistent with the index format in the index database; the tag and the index information in the index database are input into the first intelligent agent to obtain the first object features output by the first intelligent agent.

[0076] In an exemplary embodiment, the filtering module 44 is further configured to perform vectorization processing on each first object feature to obtain multiple first vectors; perform vectorization processing on each feature in the preset feature database to obtain multiple second vectors; calculate the vector similarity between each first vector and each second vector; and determine the second object feature from the first object features based on the vector similarity, wherein the vector similarity between the first vector corresponding to the second object feature and any second vector is greater than the preset vector similarity.

[0077] In an exemplary embodiment, the second generation module 46 is further configured to construct the second agent using a large language model and a second prompt word, wherein the second prompt word is used to instruct the large language model to generate an execution instruction related to the second object feature after semantic understanding of the tag; and input the tag and the second object feature into the first agent to obtain the execution instruction output by the first agent.

[0078] In an exemplary embodiment, the object data determining device is further configured to update the push group using the object data and save the push group to a database, wherein the push group includes push objects for business activities; in response to a business activity push instruction, select the target push object indicated by the business activity push instruction from the push group and push the business activity to the target push object.

[0079] In an exemplary embodiment, the object data determining device is further configured to check the groups to be pushed in the database according to a preset period, determine the overdue push groups from the database, wherein the overdue push groups represent the groups to be pushed whose last update time is greater than a preset time period; generate new push groups based on the tag corresponding to the overdue push groups and the latest original object data; and replace the overdue push groups in the database with the new push groups.

[0080] In an exemplary embodiment, the object data determining device is further configured to perform attribute distribution statistics on the object data according to preset data attributes to obtain the distribution data of the object data; and display the distribution data of the object data on a visualization interface in the form of a preset chart.

[0081] Embodiments of this application also provide a storage medium including a stored program, wherein the program executes any of the methods described above when it is run.

[0082] Optionally, in this embodiment, the storage medium may be configured to store program code for performing the following steps:

[0083] S1, obtain tagging labels from tagging tasks that tag different objects, and send the tagging labels and indicator database to a first intelligent agent so that the first intelligent agent can perform semantic understanding of the tagging labels and generate a first object feature that is consistent with the indicator format in the indicator database, wherein the first object feature is an object feature that has a semantic association with the tagging labels.

[0084] S2, the first object features are filtered based on a preset feature database to obtain second object features, wherein the second object features match the object features in the preset feature database;

[0085] S3, the tag and the second object feature are sent to the second intelligent agent so that the second intelligent agent can perform semantic understanding of the tag and generate an execution instruction corresponding to the tag based on the second object feature;

[0086] S4, execute the marking task according to the execution instruction to obtain object data for classifying the different objects.

[0087] Embodiments of this application also provide an electronic device including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the steps in any of the above method embodiments.

[0088] Optionally, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor and the input / output device is connected to the processor.

[0089] Optionally, in this embodiment, the processor can be configured to perform the following steps via a computer program:

[0090] S1, obtain tagging labels from tagging tasks that tag different objects, and send the tagging labels and indicator database to a first intelligent agent so that the first intelligent agent can perform semantic understanding of the tagging labels and generate a first object feature that is consistent with the indicator format in the indicator database, wherein the first object feature is an object feature that has a semantic association with the tagging labels.

[0091] S2, the first object features are filtered based on a preset feature database to obtain second object features, wherein the second object features match the object features in the preset feature database;

[0092] S3, the tag and the second object feature are sent to the second intelligent agent so that the second intelligent agent can perform semantic understanding of the tag and generate an execution instruction corresponding to the tag based on the second object feature;

[0093] S4, execute the marking task according to the execution instruction to obtain object data for classifying the different objects.

[0094] Optionally, in this embodiment, the storage medium may include, but is not limited to, various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0095] Embodiments of this application also provide a computer program product, which includes a computer program that, when executed by a processor, implements the steps in any of the above method embodiments.

[0096] Embodiments of this application also provide another computer program product, including a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in any of the above method embodiments.

[0097] The embodiments described herein also provide a computer program that includes computer instructions stored in a computer-readable storage medium; a processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps in any of the above method embodiments.

[0098] Optionally, specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated here.

[0099] Obviously, those skilled in the art should understand that the modules or steps of this application described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those presented here, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, this application is not limited to any particular combination of hardware and software.

[0100] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for determining object data, characterized in that, include: Obtain labeling tags from labeling tasks that label different objects, and send the labeling tags and indicator database to a first intelligent agent so that the first intelligent agent can perform semantic understanding of the labeling tags and generate a first object feature that is consistent with the indicator format in the indicator database. The first object feature is an object feature that has a semantic association with the labeling tags. The first object features are filtered based on a preset feature database to obtain the second object features, wherein the second object features match the object features in the preset feature database. The label and the second object feature are sent to the second intelligent agent so that the second intelligent agent can perform semantic understanding of the label and generate an execution instruction corresponding to the label based on the second object feature. The marking task is executed according to the execution instructions to obtain object data for classifying the different objects.

2. The method for determining object data according to claim 1, characterized in that, Sending the tagging labels and indicator database to a first intelligent agent, enabling the first intelligent agent to perform semantic understanding on the tagging labels and generate a first object feature consistent with the indicator format in the indicator database, including: The first intelligent agent is constructed using a large language model and a first prompt word, wherein the first prompt word is used to instruct the large language model to generate object features that are semantically related to the tag after performing semantic understanding on the tag, and the format of the object features is consistent with the index format in the index database. The first intelligent agent is input with the tag and the indicator information in the indicator database to obtain the first object feature output by the first intelligent agent.

3. The method for determining object data according to claim 1, characterized in that, Based on a preset feature database, the features of the first object are filtered to obtain the features of the second object, including: Each feature of the first object is vectorized to obtain multiple first vectors; Each feature in the preset feature database is vectorized to obtain multiple second vectors; the vector similarity between each first vector and each second vector is calculated, and the second object feature is determined from the first object feature based on the vector similarity, wherein the vector similarity between the first vector corresponding to the second object feature and any second vector is greater than the preset vector similarity.

4. The method for determining object data according to claim 1, characterized in that, Sending the tag and the second object feature to the second agent, so that the second agent can perform semantic understanding of the tag and generate an execution instruction corresponding to the tag based on the second object feature, including: The second intelligent agent is constructed using a large language model and a second prompt word, wherein the second prompt word is used to instruct the large language model to generate execution instructions related to the features of the second object after semantically understanding the tag. The marking label and the second object feature are input into the first intelligent agent to obtain the execution instruction output by the first intelligent agent.

5. The method for determining object data according to claim 1, characterized in that, After obtaining the object data for classifying the different objects, the method further includes: The object data is used to update the push group, and the push group is saved to the database, wherein the push group contains the push objects of the business activities; In response to a business activity push instruction, the target push object indicated by the business activity push instruction is selected from the group to be pushed, and the business activity is pushed to the target push object.

6. The method for determining object data according to claim 5, characterized in that, After saving the groups to be pushed to the database, the method further includes: The database of groups to be pushed is checked according to a preset period, and groups that have been overdue for push are identified from the database. The overdue groups are groups that were last updated more than a preset time period away from the current time. A new group to be pushed is generated based on the tagging labels corresponding to the previously expected push group and the latest original object data; Replace the overdue push groups in the database with the new push groups.

7. The method for determining object data according to claim 1, characterized in that, After obtaining the object data obtained by classifying the different objects based on the labeling results, the method further includes: Based on preset data attributes, perform attribute distribution statistics on the object data to obtain the distribution data of the object data; The distribution data of the object data is displayed on the visualization interface in the form of a preset chart.

8. A device for determining object data, characterized in that, include: The first generation module is used to obtain tagging labels from tagging tasks that tag different objects, and send the tagging labels and indicator database to the first intelligent agent so that the first intelligent agent can perform semantic understanding of the tagging labels and generate a first object feature that is consistent with the indicator format in the indicator database. The first object feature is an object feature that has a semantic association with the tagging labels. The filtering module is used to filter the first object features based on a preset feature database to obtain the second object features, wherein the second object features are object features existing in the preset feature database; The second generation module is used to send the tag and the second object feature to the second intelligent agent, so that the second intelligent agent can perform semantic understanding of the tag and generate an execution instruction corresponding to the tag based on the second object feature. An execution module is used to execute the labeling task according to the execution instructions to obtain object data for classifying the different objects.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein the program, when executed, performs the method described in any one of claims 1 to 7.

10. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to execute the method described in any one of claims 1 to 7 through the computer program.

Images