Object recognition method, apparatus, and storage medium

Abstract

The application provides an object identification method and device and a storage medium, relates to the technical field of communication, and is used for improving the technical problems of low efficiency and low accuracy in general technology. The method comprises the following steps: acquiring service data of a first object; the service data comprises service usage data and service signaling data; inputting the service data into a target classification model for classification processing to obtain a classification result of the first object; the classification result comprises a service type or a common type; in the case that the first object is of the service type, determining a potential object of a communication service corresponding to the service type from a plurality of second objects having call records with the first object.

Description

Technical Field

[0001] This application belongs to the field of communication technology, and in particular relates to an object identification method, apparatus and storage medium. Background Technology

[0002] With the continuous development of big data technology, telecommunications operators have applied technologies such as data mining to conduct data analysis in their telecommunications services in order to identify potential customers.

[0003] Currently, rule-based methods are commonly used for data analysis to identify potential customers. This approach typically involves determining whether a customer is a potential customer by judging whether their data conforms to pre-defined rules based on dimensions such as age group, income level, and residential location. However, because these rules rely on extensive manual surveys and the dimensions referenced are often limited, this method is frequently inefficient and struggles to accurately identify potential customers. Summary of the Invention

[0004] This application provides an object recognition method, apparatus, and storage medium to improve the technical problems of low efficiency and low accuracy in general technologies.

[0005] To achieve the above objectives, this application adopts the following technical solution:

[0006] In a first aspect, an object identification method is provided, comprising: acquiring business data of a first object; the business data including business usage data and business signaling data; inputting the business data into a target classification model for classification processing to obtain a classification result of the first object; the classification result including service type or ordinary type; and, if the first object is of service type, determining potential objects of communication services corresponding to the service type from multiple second objects that have call records with the first object.

[0007] Optionally, the method for determining potential objects of communication services corresponding to the service type from multiple second objects that have call records with the first object specifically includes:

[0008] The second object that meets the first preset condition among multiple second objects is identified as a potential object; the first preset condition includes that the number of calls with the first object within a first preset time period is greater than a preset number threshold, the average call duration with the first object within the first preset time period is less than a first time period threshold, and the second object belongs to a different communication network than the first object.

[0009] Optionally, the object recognition method further includes: acquiring multiple first sample data and multiple second sample data; the first sample data is used to represent business data of service type objects; the second sample data is used to represent business data of ordinary type objects; and training an initial classification model based on the multiple first sample data and multiple second sample data to obtain a target classification model.

[0010] Optionally, the first sample data meets the second preset conditions; the second preset conditions include: the number of called parties within the second preset time period is greater than a preset threshold for the number of called parties; the average outbound call duration within the second preset time period is less than or equal to a second time period threshold; the number of access network devices on which outbound calls are based within the second preset time period is less than or equal to a preset threshold for the number of access network devices; and / or the weighted score of the number of called parties, the number of called parties, the average outbound call duration, and the number of access network devices is greater than a preset score threshold; the method for obtaining multiple first sample data specifically includes:

[0011] Retrieve multiple first-candidate data; the first-candidate data includes the type label of the service type;

[0012] The first candidate data that meets the second preset condition is determined as the first sample data, and multiple first sample data are obtained.

[0013] Optionally, the service type may include logistics service type, customer service type, or training service type.

[0014] In a second aspect, an object recognition device is provided, comprising: an acquisition unit, a processing unit, and a determination unit;

[0015] The acquisition unit is used to acquire the business data of the first object; the business data includes business usage data and business signaling data.

[0016] The processing unit is used to input the business data obtained by the acquisition unit into the target classification model for classification processing, and obtain the classification result of the first object; the classification result includes service type or ordinary type.

[0017] The determining unit is used to determine, when the first object is a service type, a potential object of a communication service corresponding to the service type from a plurality of second objects that have call records with the first object.

[0018] Optionally, the unit is defined, specifically for:

[0019] The second object that meets the first preset condition among multiple second objects is identified as a potential object; the first preset condition includes that the number of calls with the first object within a first preset time period is greater than a preset number threshold, the average call duration with the first object within the first preset time period is less than a first time period threshold, and the second object belongs to a different communication network than the first object.

[0020] Optionally, the acquisition unit is further configured to acquire multiple first sample data and multiple second sample data; the first sample data is used to represent business data of service type objects; the second sample data is used to represent business data of ordinary type objects;

[0021] The processing unit is also used to train the initial classification model based on multiple first sample data and multiple second sample data to obtain the target classification model.

[0022] Optionally, the first sample data meets the second preset conditions; the second preset conditions include: the number of called parties within the second preset time period is greater than a preset threshold for the number of called parties; the average outbound call duration within the second preset time period is less than or equal to a second time period threshold; the number of access network devices on which outbound calls are based within the second preset time period is less than or equal to a preset threshold for the number of access network devices; and / or the weighted score of the number of called parties, the number of called parties, the average outbound call duration, and the number of access network devices is greater than a preset score threshold; the acquisition unit is specifically used for:

[0023] Retrieve multiple first-candidate data; the first-candidate data includes the type label of the service type;

[0024] The first candidate data that meets the second preset condition is determined as the first sample data, and multiple first sample data are obtained.

[0025] Optionally, the service type may include logistics service type, customer service type, or training service type.

[0026] Thirdly, an object recognition device is provided, including a memory and a processor; the memory is used to store computer execution instructions, and the processor is connected to the memory via a bus; when the object recognition device is running, the processor executes the computer execution instructions stored in the memory to cause the object recognition device to perform the object recognition method as described in the first aspect.

[0027] The object identification device can be a network device or a component of a network device, such as a chip system within the network device. The chip system supports the network device in implementing the functions involved in the first aspect and any possible implementation thereof, such as receiving, identifying, and routing the data and / or information involved in the aforementioned object identification method. The chip system includes a chip, but may also include other discrete devices or circuit structures.

[0028] Fourthly, a computer-readable storage medium is provided, including computer-executable instructions that, when executed on a computer, cause the computer to perform the object recognition method as described in the first aspect.

[0029] It should be noted that the aforementioned computer instructions may be stored, in whole or in part, on the first computer-readable storage medium. The first computer-readable storage medium may be packaged together with the processor of the object identification device, or it may be packaged separately from the processor of the object identification device; this application does not impose any limitations on this.

[0030] In this application, the name of the aforementioned object recognition device does not limit the device or functional module itself. In actual implementation, these devices or functional modules may appear under other names. As long as the function of each device or functional module is similar to that of this application, it falls within the scope of the claims of this application and its equivalents.

[0031] These or other aspects of this application will become more readily apparent in the following description.

[0032] The technical solution provided in this application brings at least the following beneficial effects:

[0033] Based on any of the above, in this application, business data of the first object can be obtained and input into the target classification model for classification processing to obtain the classification result of the first object, so as to further determine the potential object of the communication service corresponding to the service type from multiple second objects that have call records with the first object when the first object is a service type.

[0034] Since business data includes business usage data and business signaling data, the target classification model can accurately determine the classification result of the first object based on information from multiple dimensions. Furthermore, type determination based on the target classification model offers high processing efficiency.

[0035] Furthermore, if the first object is a service type, it indicates that the first object is more likely to communicate with users who have business needs, and potential objects are more likely to be identified from multiple second objects that have call records with the first object. Based on this, this application can support efficient and accurate identification of potential objects. Therefore, this application can be used to improve the technical problems of low efficiency and low accuracy existing in general technologies. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the structure of an object recognition system provided in an embodiment of this application;

[0037] Figure 2 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application;

[0038] Figure 3 A flowchart illustrating an object recognition method provided in an embodiment of this application;

[0039] Figure 4 A flowchart illustrating another object recognition method provided in an embodiment of this application;

[0040] Figure 5 This is a schematic diagram of the structure of an object recognition device provided in an embodiment of this application. Detailed Implementation

[0041] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0042] It should be noted that in the embodiments of this application, the words "exemplary" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0043] To facilitate a clear description of the technical solutions of the embodiments of this application, the terms "first" and "second" are used in the embodiments of this application to distinguish the same or similar items with essentially the same function and effect. Those skilled in the art can understand that the terms "first" and "second" are not intended to limit the quantity or execution order.

[0044] Furthermore, the terms "comprising" and "having" in the embodiments, claims, and drawings of this application are not exclusive. For example, a process, method, system, product, or device that includes a series of steps or modules is not limited to the listed steps or modules, but may also include steps or modules not listed.

[0045] To facilitate understanding of this application, the relevant elements involved in this application are described below.

[0046] Lead identification is a crucial business development technique that helps companies identify potential customers and predict their needs and purchasing behavior, thereby enabling them to better formulate business development strategies.

[0047] With the continuous development of big data technology, more and more enterprises are beginning to use machine learning and data mining techniques such as clustering, classification, and association rule mining to analyze and model customer data in order to identify potential customers. For example, telecommunications operators have applied machine learning and data mining techniques to their communication services for data analysis to identify potential customers.

[0048] Currently, rule-based and statistical methods are commonly used for data analysis to identify potential customers. This approach typically involves determining whether a customer is a potential customer by judging whether their data conforms to pre-defined rules based on dimensions such as age group, income level, and residential location. For example, a customer aged 25 to 40, with an income exceeding 100,000 yuan, and residing in a specific area is considered a potential customer. However, because these rules rely on extensive manual surveys and the dimensions considered are relatively limited, this method often struggles to accurately identify potential customers.

[0049] Statistical methods refer to identifying potential customers based on statistical patterns in customer data. For example, by analyzing historical transaction information in customer data, customers with transaction records corresponding to specific products can be identified as potential customers.

[0050] However, rule-based methods rely on extensive manual surveys to develop rules, which need to be constantly updated as time and customer spending levels change. Statistical methods, on the other hand, often refer to basic statistical data such as the mean and mode, easily overlooking other factors such as business content and the location of access base stations, resulting in lower identification accuracy.

[0051] To address the aforementioned issues, this application provides an object identification method that can acquire business data of a first object and input the business data of the first object into a target classification model for classification processing to obtain the classification result of the first object. In order to further identify potential objects of communication services corresponding to the service type from multiple second objects that have call records with the first object when the first object is a service type.

[0052] Since business data includes business usage data and business signaling data, the target classification model can accurately determine the classification result of the first object based on information from multiple dimensions. Furthermore, type determination based on the target classification model offers high processing efficiency.

[0053] Furthermore, if the first object is a service type, it indicates that the first object is more likely to communicate with users who have business needs, and potential objects are more likely to be identified from multiple second objects that have call records with the first object. Based on this, this application can support efficient and accurate identification of potential objects. Therefore, this application can be used to improve the technical problems of low efficiency and low accuracy existing in general technologies.

[0054] This object recognition method is applicable to object recognition systems. Figure 1 One structure of the object recognition system 100 is shown. For example... Figure 1As shown, the object recognition system 100 may include a data configuration device 101 and an object recognition device 102. The data configuration device 101 and the object recognition device 102 can be communicatively connected.

[0055] In practical applications, the object identification device 102 can also communicate with multiple data configuration devices 101 simultaneously. For ease of understanding, this application will use the communication connection between the object identification device 102 and one data configuration device 101 as an example for illustration.

[0056] Optional, Figure 1 The object recognition device 102 can be a functional module integrated into the data configuration device 101, or it can be a device that is independently set up from the data configuration device 101. This disclosure does not limit this.

[0057] It is easy to understand that when the object identification device 102 is a functional module integrated within the data configuration device 101, the communication method between the data configuration device 101 and the object identification device 102 is the same as the communication method between the two when the data configuration device 101 and the object identification device 102 are set up independently.

[0058] For ease of understanding, this application will be described primarily using the example of the data configuration device 101 and the object recognition device 102 being configured independently of each other.

[0059] In one possible approach, the data configuration device 101 can be used to provide a data configuration service. This service can support business data for the configuration object (e.g., the first object), as well as data such as parameters and algorithms for the initial classification model.

[0060] In one possible way, Figure 1 The data configuration device 101 in the middle can be a terminal. Figure 1 The diagram shows an example of one device configuration of the data configuration device 101, and is not intended to limit the specific configuration of the data configuration device 101.

[0061] In one possible approach, the object recognition device 102 may be configured with a target classification model, a first preset condition, and a second preset condition, etc., in order to provide object recognition services.

[0062] Optionally, Figure 1 The object recognition device 102 can be a terminal, a server, or other electronic devices with object recognition capabilities. Figure 1 The diagram shows an example of one device configuration of the object recognition device 102, and is not intended to limit the specific configuration of the object recognition device 102.

[0063] Optionally, when the data configuration device 101 and the object identification device 102 are terminals, the terminal may be a device providing voice and / or data connectivity to a user, a handheld device with wireless connectivity, or other processing devices connected to a wireless modem. The wireless terminal can communicate with one or more core networks via a radio access network (RAN). The wireless terminal may be a mobile terminal, such as a computer with a mobile terminal, or a portable, pocket-sized, handheld, or computer-embedded mobile device that exchanges voice and / or data with the radio access network, such as a mobile phone, tablet computer, laptop computer, netbook, or personal digital assistant (PDA). This application embodiment does not impose any limitations on this.

[0064] When the object identification device 102 is a server, the server can be a single server or a server cluster consisting of multiple servers. In some embodiments, the server cluster can also be a distributed cluster. This application does not impose any limitations on this.

[0065] like Figure 2 The diagram shown is a hardware structure schematic of an electronic device provided in an embodiment of this application. This electronic device can be a data configuration device 101 or an object recognition device 102. The electronic device includes a processor 21, a memory 22, a communication interface 23, and a bus 24. The processor 21, memory 22, and communication interface 23 are connected via the bus 24.

[0066] Processor 21 is the control center of the electronic device. It can be a single processor or a collective term for multiple processing elements. For example, processor 21 can be a CPU or other general-purpose processors. Among them, general-purpose processors can be microprocessors or any conventional processors.

[0067] As one embodiment, processor 21 may include one or more CPUs, for example Figure 2 CPU0 and CPU1 are shown in the diagram.

[0068] The memory 22 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto.

[0069] In one possible implementation, the memory 22 can exist independently of the processor 21. The memory 22 can be connected to the processor 21 via a bus 24 and is used to store instructions or program code. When the processor 21 calls and executes the instructions or program code stored in the memory 22, it can implement the object recognition method provided in the following embodiments of this application.

[0070] In another possible implementation, the memory 22 can also be integrated with the processor 21.

[0071] Communication interface 23 is used for connecting electronic devices to other devices via a communication network, which can be Ethernet, wireless access network, wireless local area network (WLAN), etc. Communication interface 23 may include a receiving unit for receiving data and a transmitting unit for sending data.

[0072] Bus 24 can be an industry standard architecture (ISA) bus, a peripheral component interconnect (PCI) bus, or an extended industry standard architecture (EISA) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 2 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0073] It should be pointed out that, Figure 2 The structures shown do not constitute a limitation on electronic devices, except... Figure 2 In addition to the components shown, electronic devices may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0074] like Figure 3 The diagram shown is a flowchart illustrating an object recognition method provided in an embodiment of this application. This object recognition method can be applied to... Figure 1 The object recognition device 102 in the object recognition system 100 shown. The object recognition method includes: S301-S303.

[0075] S301, The object recognition device acquires the business data of the first object.

[0076] Business data can include business usage data and business signaling data.

[0077] In one possible approach, the first object could be the user's identity identifier within the communication network. For example, the first object could be a mobile phone number, an International Mobile Subscriber Identity (IMSI), etc.

[0078] Optionally, service usage data may include service package information, service usage duration (also known as network access duration), and service contract type. Service contract type can be individual or enterprise. Enterprise type can be further divided into logistics, food delivery, driver, and customer service types, etc. Service package information may include available call minutes and available network traffic subscribed by the user within the operator's network.

[0079] Optionally, the service signaling data may include the establishment time of the call by the first object, the call type, the local access network equipment, the peer number, the peer access network equipment, the call duration, and the communication network type.

[0080] In one possible way, combining Figure 1 Staff can pre-configure the business data of the first object into the object recognition device using a data configuration device, and configure trigger commands corresponding to the business data of the first object. Trigger commands can be conditional triggers or timed triggers, etc. Based on this, the object recognition device can read the business data of the first object when the trigger command is triggered, in order to perform object recognition.

[0081] Alternatively, staff can send an object identification request carrying the business data of the first object to the object identification device in real time via the data configuration device. Correspondingly, the object identification device can receive the object identification request, parse it to obtain the business data of the first object, and perform object identification in response to the request.

[0082] Specifically, the data configuration device may be equipped with an input module. Staff can perform data editing operations, instruction editing operations, and request sending operations through the input module. In response to a data editing operation performed by a staff member, the data configuration device can acquire the business data of the first object corresponding to the data editing operation and send the business data of the first object to the object recognition device. In response to an instruction editing operation performed by a staff member, the data configuration device can also generate a trigger instruction corresponding to the instruction editing operation and send the trigger instruction to the object recognition device. In response to a request sending operation performed by a staff member, the data configuration device can also generate an object recognition request corresponding to the request sending operation and send the object recognition request of the first object to the object recognition device. Correspondingly, the object recognition device can receive the business data of the first object, the trigger instruction, and the object recognition request from the data configuration device.

[0083] S302. The object recognition device inputs business data into the target classification model for classification processing to obtain the classification result of the first object.

[0084] The classification results can include service type or general type. Service type indicates that the first object belongs to a user within a specific service scenario. Specific service scenarios could be logistics services, training services, game customer service, video customer service, etc. General type indicates that the first object belongs to a general user. A general user is an individual user in a communication network who meets their personal calling and internet access needs.

[0085] It's important to note that, compared to ordinary types, when the first object is a service type, it indicates that the first object is more likely to communicate with users of a specific type, and potential objects are more likely to be identified from those with call records with the first object. For example, when the service type indicates that the first object belongs to a user in a specific service scenario, the first object will typically communicate with users who have online shopping needs. Users with online shopping needs often have a high demand for data traffic, so potential objects with data traffic needs are more likely to be identified from those with call records with the first object. Similarly, when the service type indicates that the first object belongs to a user in a game customer service scenario, the first object will typically communicate with game players. Game players often have a high demand for data traffic, so potential objects with data traffic needs are more likely to be identified from those with call records with the first object.

[0086] Based on this, in order to efficiently and accurately identify potential objects, the object recognition device can input the business data of the first object into the target classification model for classification processing to obtain the classification result of the first object. If the classification result of the first object is ordinary type, it indicates that it is difficult to identify potential objects from objects with call records with the first object, and the object recognition device can end the current recognition process. If the classification result of the first object is service type, it indicates that the possibility of identifying potential objects from objects with call records with the first object is relatively high, and the object recognition device can continue to execute S303 below.

[0087] In one possible approach, the target classification model can determine the confidence level (also known as the predicted probability) for the service type and the confidence level for the common type, and use the higher confidence level value as the classification result for the first object.

[0088] In one possible example, if the target classification model obtains a confidence score of 0.4 for the ordinary type and a confidence score of 0.6 for the service type after the first object is input into it, then the target classification model determines the classification result of the first object as the service type.

[0089] In one possible approach, the target classification model can be obtained by pre-training an initial classification model based on multiple sample data. For the specific process of training the target classification model, please refer to S501-S502 below for understanding; it will not be repeated here.

[0090] In one possible approach, service types could include logistics services, customer service services, or training services. Considering the varying needs of different types of users for the services, service types can be further subdivided into logistics services, training services, and customer service services (e.g., game customer service and video customer service). In this case, the target classification model can be trained based on different sample data from different scenarios; it is a multi-classification model, meaning the output classification result is either logistics service, customer service, training service, or a general type. Based on this, the type of the first object can be accurately identified, facilitating the accurate determination of the degree of need for the services among potential users identified based on the first object, thus enabling accurate service package recommendations.

[0091] In one possible approach, to improve the classification accuracy of the target classification model, the model can also be a binary classification model, meaning the output classification result is either service type or ordinary type. Furthermore, target classification models corresponding to different scenarios can be trained based on sample data from various scenarios such as logistics services, training services, game customer service, and video customer service. For example, a target classification model corresponding to the logistics service scenario, and a target classification model corresponding to the game customer service scenario, etc.

[0092] In this scenario, the first object can be an object belonging to a different enterprise type corresponding to different scenarios. For example, the enterprise type corresponding to a logistics service scenario could be a logistics type, while the enterprise type corresponding to a game customer service scenario could be a game service type, and so on. Considering that objects belonging to the same enterprise type may have different functions within the enterprise, objects belonging to the same enterprise type may be service type or general type. Based on this, the object recognition device can determine the classification result of the first object through a target classification model corresponding to the enterprise type of the first object, thereby performing object recognition.

[0093] S303. When the first object is a service type, the object identification device determines the potential object of the communication service corresponding to the service type from a plurality of second objects that have call records with the first object.

[0094] In one possible approach, if the classification result of the first object is a service type, it indicates that there is a high probability of identifying a potential object from the objects that have call records with the first object. In this case, the object identification device can determine the potential object of the communication service corresponding to the service type from multiple second objects that have call records with the first object.

[0095] Optionally, the object identification device can obtain the call setup time and peer number (i.e., the second object) of the first object from the service signaling data of the first object, and based on the call setup time, count the number of times each peer number appears within a first preset duration (e.g., 1 hour or 24 hours), that is, the number of calls between each second object and the first object within a specific duration, so as to further identify the second object whose number of calls is greater than a preset threshold as a potential object.

[0096] Alternatively, the object identification device can obtain the peer number (i.e., the second object) and communication network type of the call made by the first object from the service signaling data of the first object, that is, determine the communication network type of each second object, so as to further identify the second objects whose communication network type is different from that of the first object as potential objects.

[0097] In one possible approach, a potential target can be used to represent a second group of users who are more likely to subscribe to the communication services corresponding to the service type. In other words, the users to whom the potential target belongs are more likely to have a demand for the communication services corresponding to the service type. Subsequently, the operator can accurately target the services to the users to whom the potential target belongs.

[0098] In one possible approach, when the service type indicates that the first object belongs to a user in a logistics service scenario, the second object may belong to a user who frequently uses online shopping application services, and this user's demand for network traffic is moderate. In this case, the communication service corresponding to the service type could be a service that provides the first amount of network traffic. When the service type indicates that the first object belongs to a user in a game customer service scenario, the second object may belong to a user who frequently uses game application services, and this user's demand for network traffic is high. In this case, the communication service corresponding to the service type could be a service that provides the second amount of network traffic. The second amount can be greater than the first amount.

[0099] In one embodiment, in S303 above, when the object identification device determines the potential object of the communication service corresponding to the service type from a plurality of second objects that have a call record with the first object, this application provides an optional implementation method, including: S401.

[0100] S401, The object recognition device identifies the second object that meets the first preset condition among a plurality of second objects as a potential object.

[0101] The first preset condition may include the number of calls with the first object within a first preset time period being greater than a preset number threshold, the average call duration with the first object within a first preset time period being less than a first time period threshold, and belonging to a different communication network from the first object.

[0102] In one possible approach, the object identification device can, based on the service signaling data of the first object, count the number of calls, average call duration, and communication network type between each second object (i.e., peer number) and the first object within a first preset time period, and further determine whether each second object meets the first preset conditions.

[0103] It should be noted that, for a network operator, entities belonging to other network operators and with high network demand are often considered potential targets. The first preset condition can help filter out entities belonging to other network operators, with frequent calls, and demonstrating network traffic usage needs. Therefore, this application can efficiently and accurately identify potential targets, thereby accelerating the network operator's business development.

[0104] In one possible approach, the first preset duration, the preset number of times threshold, and the first duration threshold can be reasonably set by staff according to different scenarios.

[0105] In one embodiment, in order to train a target classification model, such as Figure 4 As shown, the object recognition method provided in this application embodiment further includes: S501-S502.

[0106] S501, The object recognition device acquires multiple first sample data and multiple second sample data.

[0107] The first sample data can be used to represent business data for service-type objects. The second sample data can be used to represent business data for ordinary-type objects. The number of multiple first sample data and multiple second sample data can be the same or similar to avoid model overfitting and improve the model's classification accuracy.

[0108] In one possible way, combining Figure 1 Staff can send multiple first sample data points and multiple second sample data points to the object recognition device via the data configuration device. Correspondingly, the object recognition device can receive these multiple first sample data points and multiple second sample data points.

[0109] In one possible approach, when service types are further subdivided into logistics service types, customer service service types, and training service types, the sample data can also be divided into logistics service types, customer service service types, training service types, and general types, etc., to facilitate training a multi-classification model. In this case, the number of sample data for each type can be the same or similar to avoid model overfitting and improve the model's classification accuracy.

[0110] S502, The object recognition device trains the initial classification model based on multiple first sample data and multiple second sample data to obtain the target classification model.

[0111] In one possible approach, aside from differing type labels, the first and second sample data can include data of the same dimensions. This includes information on each item in the sample object's business usage information and business signaling data. Examples include mobile phone number, network access duration, service package type, monthly spending limit, total number of calls, number of outgoing calls, number of incoming calls, total call duration, average call duration, number of calls with a duration less than a third duration threshold, access network device, the number of calls corresponding to the access network device, the number of unique peer numbers, and the number of peer numbers with a cumulative outgoing call count exceeding the outgoing call count threshold. When the sample data has diverse dimensions, this can improve the training effect of the initial classification model and increase classification accuracy.

[0112] In one possible approach, the object recognition device can randomly divide multiple first sample data and multiple second sample data into multiple subsets, and sequentially input the first sample data and second sample data from each subset into an initial classification model for classification processing to obtain a processing result. Based on the processing result, the type labels of each first sample data and each second sample data, a loss value is determined. If the loss value is greater than a preset loss threshold, the parameters of the initial classification model can be adjusted until the loss value is less than or equal to the preset loss threshold, at which point the object recognition device can obtain a target classification model.

[0113] In one possible approach, the initial classification model can be generated based on a tree model such as extreme gradient boosting (XGBoost), light gradient boosting machine (LightGBM), or gradient boosting (CatBoost).

[0114] In one embodiment, when acquiring multiple first sample data, this application provides an optional implementation method, including: S601-S602.

[0115] S601, The object recognition device acquires multiple first candidate data.

[0116] The first candidate data may include the type label of the service type.

[0117] In one possible way, combining Figure 1 Staff can use the data configuration device to perform tag editing operations, creating multiple types of tags that correspond one-to-one with multiple sample objects. In response to the data configuration device, it can generate multiple types of tags that correspond one-to-one with the sample objects, and combine them with the business data of each sample object to obtain multiple first candidate data, which are then sent to the object recognition device. Correspondingly, the object recognition device can receive the multiple first candidate data.

[0118] Optionally, when the sample object is a mobile phone number, the data configuration device can be used to provide number type identification service and type information editing service. The number type identification service can be implemented based on relevant application services. These relevant application services can display type labels for mobile phone numbers based on big data. For example, type labels could be fraudulent calls, express delivery, and food delivery. The type information editing service can support users manually labeling mobile phone numbers with type labels. For example, for all mobile phone numbers under a company's number segment, staff can manually label them with type labels corresponding to the company's type. For example, for a logistics service company, all mobile phone numbers under the number segment applied for by the company can be manually labeled as logistics service type.

[0119] S602, The object recognition device determines the first candidate data that meets the second preset conditions as the first sample data, and obtains multiple first sample data.

[0120] The first sample data may meet the second preset conditions. The second preset conditions may include: the number of called parties within the second preset time period is greater than a preset number threshold; the average outbound call duration within the second preset time period is less than or equal to the second time period threshold; the number of access network devices on which outbound calls are based within the second preset time period is less than or equal to a preset number threshold; and / or the weighted score of the number of call parties, the number of called parties, the average outbound call duration, and the number of access network devices is greater than a preset score threshold.

[0121] In one possible approach, the second preset condition could also include the number of outbound calls exceeding the threshold for the first call within different time periods. That is, making multiple outbound calls at irregular intervals.

[0122] In one possible approach, the object identification device can count the number of called parties, the average outbound call duration, and the number of access network devices on which the outbound call is based within the service signaling data of each first candidate data within a second preset time period. It can further determine whether the number of called parties corresponding to each first candidate data within the second preset time period is greater than a preset object number threshold, whether the average outbound call duration within the second preset time period is less than or equal to a second time period threshold, and whether the number of access network devices on which the outbound call is based within the second preset time period is less than or equal to a preset device number threshold, so as to determine whether the first candidate data meets the second preset conditions.

[0123] In one possible approach, the object recognition device can determine the weighted score corresponding to each first candidate data based on a first formula, the number of call objects (also known as the number of peer numbers), the number of called objects, the average outbound call duration, and the number of access network devices. It then determines whether the weighted score corresponding to each first candidate data is greater than a preset score threshold, thereby determining whether the first candidate data meets a second preset condition. Finally, it selects the first candidate data with a score higher than the preset score threshold as the first sample data. The first formula is:

[0124]

[0125] Among them, S i Let be the weighted score of the i-th first candidate data, where i is a positive integer. C i Let C be the number of call participants in the i-th first candidate data. min This is a preset threshold corresponding to the number of call participants. The higher the score, the better, as the number of call participants exceeds this preset threshold. T represents the number of called parties in the i-th first candidate data, i.e., the number of outgoing call numbers. i Let T be the average outbound call duration in the i-th first candidate data. max This is a preset threshold for the duration of a single outbound call. i This refers to the number of access network devices. (CELL) max A preset threshold corresponding to the number of access network devices.

[0126] In one possible approach, the second preset duration, the preset target number threshold, the second duration threshold, and the preset device number threshold can be reasonably set by staff according to different scenarios. For example, in a training service scenario, the second preset duration could be 1 hour, the preset target number threshold could be 10, the second duration threshold could be 3 minutes, and the preset device number threshold could be 1.

[0127] In one possible approach, after obtaining multiple first sample data, the object recognition device can randomly sample the business data of other sample objects based on the number of the multiple first sample data, and assign a common type label to the business data of the sampled sample objects to obtain multiple second sample data, so that the number of multiple first sample data and the number of multiple second sample data are the same.

[0128] In this embodiment of the application, the object recognition device can acquire the business data of the first object and input the business data of the first object into the target classification model for classification processing to obtain the classification result of the first object. In order to further determine the potential object of the communication service corresponding to the service type from multiple second objects that have call records with the first object when the first object is a service type.

[0129] Since business data includes business usage data and business signaling data, the target classification model can accurately determine the classification result of the first object based on information from multiple dimensions. Furthermore, type determination based on the target classification model offers high processing efficiency.

[0130] Furthermore, if the first object is a service type, it indicates that the first object is more likely to communicate with users who have business needs, and potential objects are more likely to be identified from multiple second objects that have call records with the first object. Based on this, this application can support efficient and accurate identification of potential objects. Therefore, this application can be used to improve the technical problems of low efficiency and low accuracy existing in general technologies.

[0131] To address the numerous shortcomings of general-purpose techniques, this application proposes a more comprehensive, efficient, and practical method for identifying potential objects, including steps such as sample data screening, type label definition, business data classification processing, and classification model training. These novel steps enable this application to achieve higher accuracy and a wider range of applications.

[0132] The foregoing mainly describes the solutions provided by the embodiments of this application from a methodological perspective. To achieve the above functions, it includes corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0133] This application embodiment can divide the object recognition device into functional modules according to the above method example. For example, each function can be divided into a separate functional module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or as a software functional module. Optionally, the module division in this application embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0134] like Figure 5 The diagram shown is a structural schematic of an object recognition device provided in an embodiment of this application. This object recognition device can be used to perform tasks such as... Figure 3 and Figure 4The object recognition method shown, and related implementations not shown in the accompanying drawings, are described above. The object recognition device includes: an acquisition unit 701, a processing unit 702, and a determination unit 703;

[0135] Acquisition unit 701 is used to acquire the service data of the first object; the service data includes service usage data and service signaling data; for example, combined with Figure 3 The acquisition unit 701 can be used to execute S301.

[0136] Processing unit 702 is used to input the business data acquired by acquisition unit 701 into the target classification model for classification processing to obtain the classification result of the first object; the classification result includes service type or ordinary type; for example, combined with Figure 3 The processing unit 702 can be used to execute S302.

[0137] The determining unit 703 is used to, when the first object is a service type, determine potential objects for communication services corresponding to the service type from a plurality of second objects that have call records with the first object. For example, in combination with Figure 3 Unit 703 can be used to execute S303.

[0138] Optionally, unit 703 is specifically used for:

[0139] The second object among multiple second objects that meets the first preset condition is identified as a potential object; the first preset condition includes that the number of calls with the first object within a first preset time period is greater than a preset number threshold, the average call duration with the first object within the first preset time period is less than a first time period threshold, and the second object belongs to a different communication network than the first object. For example, the determining unit 703 can be used to execute S401.

[0140] Optionally, the acquisition unit 701 is further configured to acquire multiple first sample data and multiple second sample data; the first sample data is used to represent business data of service type objects; the second sample data is used to represent business data of ordinary type objects; for example, combined with Figure 4 The acquisition unit 701 can be used to execute S501.

[0141] The processing unit 702 is also used to train the initial classification model based on multiple first sample data and multiple second sample data to obtain the target classification model. For example, combining... Figure 4 The processing unit 702 can be used to execute S502.

[0142] Optionally, the first sample data meets the second preset conditions; the second preset conditions include: the number of called parties within the second preset duration is greater than a preset threshold for the number of called parties; the average outbound call duration within the second preset duration is less than or equal to a second duration threshold; the number of access network devices on which outbound calls are based within the second preset duration is less than or equal to a preset threshold for the number of access network devices; and / or the weighted score of the number of called parties, the number of called parties, the average outbound call duration, and the number of access network devices is greater than a preset score threshold; the acquisition unit 701 is specifically used for:

[0143] Multiple first candidate data are obtained; the first candidate data includes the type label of the service type; for example, the obtaining unit 701 can be used to execute S601.

[0144] The first candidate data that meets the second preset condition is determined as the first sample data, and multiple first sample data are obtained. For example, the acquisition unit 701 can be used to execute S602.

[0145] Optionally, the service type may include logistics service type, customer service type, or training service type.

[0146] Those skilled in the art will recognize that, in one or more of the examples above, the functions described in this application can be implemented using hardware, software, firmware, or any combination thereof. When implemented in software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media include computer-readable storage media and communication media, wherein communication media include any medium that facilitates the transmission of a computer program from one place to another. Storage media can be any available medium accessible to a general-purpose or special-purpose computer.

[0147] Through the above description of the embodiments, those skilled in the art can clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0148] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units described above is only a logical functional division, and other division methods may exist in actual implementation. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the shown or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms. Units described as separate components may or may not be physically separate; components shown as units may be one physical unit or multiple physical units, i.e., they may be located in one place or distributed in multiple different places. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0149] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An object recognition method, characterized in that, include: Obtain the business data of the first object; the business data includes business usage data and business signaling data; The business data is input into the target classification model for classification processing to obtain the classification result of the first object; the classification result includes service type or general type; the service type includes logistics service type, customer service service type, or training service type. If the first object is the service type, then from a plurality of second objects that have call records with the first object, potential objects for communication services corresponding to the service type are determined; The method further includes: Obtain multiple first candidate data; the first candidate data includes the type label of the service type; The first candidate data that meets the second preset condition is determined as the first sample data, and multiple first sample data are obtained; the first sample data is used to represent the business data of the object of the service type; the second preset condition includes: the number of called objects corresponding to the second preset time period is greater than the preset object number threshold, the average outbound call duration within the second preset time period is less than or equal to the second time period threshold, the number of access network devices on which the outbound call is based within the second preset time period is less than or equal to the preset device number threshold, and / or the weighted score of the number of call objects, the number of called objects, the average outbound call duration and the number of access network devices is greater than the preset score threshold; Acquire multiple second sample data; the second sample data is used to represent the business data of the common type of object; Based on the multiple first sample data and the multiple second sample data, the initial classification model is trained to obtain the target classification model.

2. The object recognition method according to claim 1, characterized in that, The step of determining potential objects for communication services corresponding to the service type from a plurality of second objects that have call records with the first object includes: The second object that meets the first preset condition among the plurality of second objects is determined as the potential object; the first preset condition includes that the number of calls with the first object within a first preset time period is greater than a preset number threshold, the average call duration with the first object within the first preset time period is less than a first duration threshold, and the object belongs to a different communication network than the first object.

3. An object recognition device, characterized in that, include: Acquisition unit, processing unit, and determination unit; The acquisition unit is used to acquire the service data of the first object; the service data includes service usage data and service signaling data; The processing unit is used to input the business data acquired by the acquisition unit into the target classification model for classification processing to obtain the classification result of the first object; the classification result includes service type or ordinary type; the service type includes logistics service type, customer service type or training service type; The determining unit is configured to, when the first object is the service type, determine, from a plurality of second objects that have call records with the first object, a potential object of a communication service corresponding to the service type; The acquisition unit is further configured to acquire multiple first candidate data; the first candidate data includes the type label of the service type; The determining unit is further configured to determine the first candidate data that meets the second preset condition as the first sample data, thereby obtaining multiple first sample data; The first sample data is used to represent the business data of the object of the service type; The second preset conditions include: the number of called parties within the second preset duration is greater than a preset number threshold; the average outbound call duration within the second preset duration is less than or equal to a second duration threshold; the number of access network devices on which outbound calls are based within the second preset duration is less than or equal to a preset number threshold; and / or the weighted score of the number of call parties, the number of called parties, the average outbound call duration, and the number of access network devices is greater than a preset score threshold. The acquisition unit is further configured to acquire multiple second sample data; the second sample data is used to represent the business data of the ordinary type object; The processing unit is further configured to train the initial classification model based on the plurality of first sample data and the plurality of second sample data to obtain the target classification model.

4. The object recognition device according to claim 3, characterized in that, The determining unit is specifically used for: The second object that meets the first preset condition among the plurality of second objects is determined as the potential object; the first preset condition includes that the number of calls with the first object within a first preset time period is greater than a preset number threshold, the average call duration with the first object within the first preset time period is less than a first duration threshold, and the object belongs to a different communication network than the first object.

5. An object recognition device, characterized in that, It includes a memory and a processor; the memory is used to store computer execution instructions, and the processor is connected to the memory via a bus; when the object recognition device is running, the processor executes the computer execution instructions stored in the memory, so that the object recognition device performs the object recognition method as described in any one of claims 1-2.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes computer-executable instructions that, when executed on a computer, cause the computer to perform the object recognition method as described in any one of claims 1-2.

Images