Classification method, device and equipment of internet of things card, storage medium and product

Abstract

The application provides a method, device, equipment, storage medium and product for classifying an Internet of Things card. The method comprises: obtaining first card data of the Internet of Things card and second card data of a communication card number in a call card number white list of the Internet of Things card; the second card data comprises calling communication data and user track location information corresponding to the communication card number; based on the calling communication data and the user track location information corresponding to the communication card number, the social circle information of the communication card number is classified to obtain a social circle category; according to a preset abnormal call transfer index, target data corresponding to the preset abnormal call transfer index is determined from the first card data and the second card data; the preset abnormal call transfer index at least comprises a social circle index; the target data at least comprises the social circle category; and according to the preset abnormal call transfer index, an index weight and the target data, a risk category of the Internet of Things card is determined.

Description

Technical Field

[0001] This application relates to the field of mobile communication technology, and in particular to a method, apparatus, device, storage medium and product for classifying Internet of Things (IoT) cards. Background Technology

[0002] With the development of the Internet of Things (IoT) industry, the use of IoT SIM cards is becoming increasingly widespread. However, the regulation of IoT SIM cards is currently more lenient compared to that of traditional SIM cards, leading many criminals to attempt to use them for illegal activities. Because IoT SIM cards have a whitelist restriction, a common practice is to add a number to the whitelist and then set up call forwarding on that number. By using this number to forward calls, the whitelist restriction is bypassed, allowing the IoT SIM card number to make unrestricted calls to other numbers.

[0003] Currently, to prevent criminals from using the aforementioned methods for illegal activities, the number of call forwarding calls on IoT cards within a certain time period is typically tracked. If the number of call forwarding calls exceeds a preset value, the IoT card is shut down or otherwise disabled. However, ordinary users also have a need to use the call forwarding function when using IoT cards normally. The current method is prone to misjudgment and has low accuracy in risk identification, resulting in a negative impact on the use of many legitimate customers. Summary of the Invention

[0004] This application provides a method, apparatus, device, storage medium, and product for classifying Internet of Things (IoT) cards, which can improve the accuracy of risk identification.

[0005] Firstly, this application provides a method for classifying Internet of Things (IoT) cards, the method comprising:

[0006] Obtain first card data of the IoT card and second card data of the communication card number in the call card number whitelist of the IoT card; the second card data includes the calling call data and the user trajectory location information corresponding to the communication card number;

[0007] Based on the caller's data and the user's location information corresponding to the communication card number, the social circle information of the communication card number is classified to obtain the social circle category;

[0008] Based on a preset abnormal call forwarding index, target data corresponding to the preset abnormal call forwarding index is determined from the first card data and the second card data; the preset abnormal call forwarding index includes at least a social circle index; the target data includes at least a social circle category.

[0009] The risk category of the IoT card is determined based on the preset abnormal call transfer indicators, indicator weights, and target data.

[0010] The method described in this application obtains first card data of an IoT card and second card data of communication card numbers from the call card number whitelist of the IoT card. Then, based on the calling data and the user trajectory location information corresponding to the communication card number, the social circle information of the communication card number is categorized to obtain a social circle category. Next, target data corresponding to the preset abnormal call forwarding index is determined from the first and second card data according to the social circle category. Finally, the risk category of the IoT card is determined based on the preset abnormal call forwarding index, the index weight, and the target data. By introducing social circle information, misjudgment is avoided when normal users use call forwarding for numbers within their social circle, resulting in high accuracy in risk identification.

[0011] In some possible implementations, the calling call data includes the calling call time, and the social circle information of the communication card number is categorized based on the calling call data and the user trajectory location information corresponding to the communication card number to obtain social circle categories, including:

[0012] Extract the caller's time from the data on the second SIM card;

[0013] Search the second card data for the target user's trajectory location information that matches the caller's call time;

[0014] Based on the caller's call time and the matching target user's location information, the called numbers of the communication card number are classified to obtain the social circle category.

[0015] The method described in this application extracts the caller's call time and the corresponding target user's location information from the second card data. Then, based on these two pieces of information, it classifies the called numbers of the communication card number to obtain different social circle categories where the communication card number frequently makes calls at specific times and locations. This allows for accurate classification of the different social circle categories of the numbers called by the communication card number based on its calling patterns.

[0016] In some possible implementations, the step of classifying the called numbers of the communication card number according to the calling time and the matching target user trajectory location information to obtain social circle categories includes:

[0017] From the target user's location trajectory, filter out the user's work location trajectory information corresponding to the caller's call time being on a weekday;

[0018] Based on the user's work trajectory location information, determine the work location of the user with the communication card number;

[0019] The time when the user of the aforementioned communication card number is at their workplace is defined as the working period.

[0020] For each called number, count the first number of proactive calls between the communication card number and the called card number from the calling call data, and the second number of proactive calls between the communication card number and the called number at the work location and during the work hours.

[0021] For each called number, calculate the ratio of the second number to the first number;

[0022] The social circle category corresponding to the called number whose ratio is greater than a preset threshold is determined as the working relationship category.

[0023] The method described in this application involves filtering user work location information from the target user's trajectory location data, identifying the user's work location corresponding to the caller's call time being a weekday, determining the user's work location based on this trajectory location information, and then defining the time the user is at their work location as the work period. Next, for each called number, the number of second calls made at the work location and during the work period is counted, along with the total number of first calls. The ratio of the second call to the first call is calculated. When the ratio exceeds a preset threshold, the social circle category corresponding to this number is determined as the work relationship category. The number classified as the work relationship category is determined by calculating the proportion of calls made by each number at the work location and during the work period.

[0024] In some possible implementations, the step of classifying the called numbers of the communication card number according to the calling time and the matching target user trajectory location information to obtain social circle categories includes:

[0025] From the target user's location trajectory, filter out the location trajectory information of users whose call time was on a rest day;

[0026] Based on the user's rest trajectory location information, determine the rest location of the user with the communication card number;

[0027] The time when the user with the aforementioned communication card number is at a rest location is defined as the rest period.

[0028] For each called number, the number of the first proactive calls between the communication card number and the called card number are counted from the calling call data, as well as the number of the third proactive calls between the communication card number and the called number at the rest location and rest time.

[0029] For each called number, calculate the ratio of the third digit to the first digit;

[0030] Called numbers with a ratio greater than a preset threshold are identified as belonging to the social circle category of family and friends.

[0031] The method described in this application involves filtering user rest location information from the target user's trajectory location that corresponds to the caller's call time being a rest day, thereby determining the user's rest location and defining the time the user is at the rest location as the rest period. Then, for each called number, the number of third calls made at the rest location and rest time is counted, along with the total number of first calls. The ratio of the third call to the first call is calculated. When the ratio exceeds a preset threshold, the social circle category corresponding to this number is determined as the "family and friends" category. The number classified as a family and friends category is determined by calculating the proportion of calls made by each number at rest locations and rest times.

[0032] In some possible implementations, after determining the risk category of the IoT card based on the preset abnormal call transfer index, index weight, and target data, the method further includes:

[0033] If the risk category is Level 1, service to the first number shall be suspended;

[0034] When the risk category is classified as the second risk level, a risk warning will be issued.

[0035] In the case of risk category three, an offline confirmation notification is issued, wherein the levels of the first risk, the second risk, and the third risk increase sequentially.

[0036] The methods described in this application involve different actions depending on the risk situation, such as suspending service, issuing risk warnings, or sending offline confirmation notices. This allows for different risk management approaches based on different circumstances.

[0037] In some possible implementations, after determining the risk category of the IoT card based on the preset abnormal call transfer index, index weight, and target data, the method further includes:

[0038] Obtain the actual risk category of the IoT card;

[0039] If the actual risk category is inconsistent with the determined risk category, the weight corresponding to the preset abnormal call transfer index shall be adjusted according to the actual risk category and the risk category.

[0040] The method described in this application, after determining the risk category, obtains the actual risk category of the IoT card. If the actual risk category differs from the determined risk category, the weight of the preset abnormal call forwarding indicator is adjusted accordingly. This allows for the adjustment of relevant parameters based on the actual risk level, making the results closer to the actual situation.

[0041] In some possible implementations, after obtaining the first card data of the IoT card and the second card data of the communication card numbers in the call card number whitelist of the IoT card, the method further includes:

[0042] Extract the first user identifier from the first card data;

[0043] From the Internet of Things (IoT) system, extract IoT data features corresponding to the first user identifier to obtain IoT information;

[0044] Extract the second user identifier from the second card data;

[0045] From the human network system, extract the human network data features corresponding to the second user identifier to obtain human network information;

[0046] The IoT information and the human network information are combined to obtain a wide data table.

[0047] The method described in this application extracts IoT data features corresponding to a first user identifier from the IoT system, and similarly extracts human network data features corresponding to a second user identifier from the human network system, thereby obtaining IoT information and human network information. Finally, the IoT information and the human network information are combined to obtain a wide data table. This method enables data fusion between two different systems, breaking down data silos while avoiding data security issues.

[0048] Secondly, this application provides a sorting device for Internet of Things (IoT) cards, characterized in that the device comprises:

[0049] The acquisition module is used to acquire first card data of the IoT card and second card data of the communication card number in the call card number whitelist of the IoT card; the second card data includes the calling call data and the user trajectory location information corresponding to the communication card number;

[0050] The classification module is used to classify the social circle information of the communication card number based on the calling call data and the user trajectory location information corresponding to the communication card number, so as to obtain the social circle category;

[0051] The determination module is used to determine target data corresponding to the preset abnormal call transfer index from the first card data and the second card data based on the preset abnormal call transfer index; the preset abnormal call transfer index includes at least a social circle index; the target data includes at least a social circle category;

[0052] The determination module is also used to determine the risk category of the IoT card based on the preset abnormal call transfer index, index weight and target data.

[0053] Thirdly, embodiments of this application provide an IoT card classification device, the device comprising: a processor, and a memory storing computer program instructions; the processor reads and executes the computer program instructions to implement the IoT card classification method described above.

[0054] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer program instructions, which, when executed by a processor, implement the IoT card classification method described above.

[0055] Fifthly, embodiments of this application provide a computer program product, wherein instructions in the computer program product, when executed by a processor of an electronic device, cause the electronic device to perform the IoT card classification method described above.

[0056] The method described in this application obtains first card data of an IoT card and second card data of communication card numbers from the call card number whitelist of the IoT card. Then, based on the calling data and the user trajectory location information corresponding to the communication card number, the social circle information of the communication card number is categorized to obtain a social circle category. Next, target data corresponding to the preset abnormal call forwarding index is determined from the first and second card data according to the social circle category. Finally, the risk category of the IoT card is determined based on the preset abnormal call forwarding index, the index weight, and the target data. By introducing social circle information, misjudgment is avoided when normal users use call forwarding for numbers within their social circle, resulting in high accuracy in risk identification. Attached Figure Description

[0057] This application can be better understood from the following description of specific embodiments in conjunction with the accompanying drawings, wherein:

[0058] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings, wherein the same or similar reference numerals denote the same or similar features.

[0059] Figure 1 This is a flowchart illustrating an embodiment of an IoT card classification method provided in this application;

[0060] Figure 2 This is a diagram illustrating call forwarding using an IoT SIM card;

[0061] Figure 3 This is a schematic diagram of the overall process of a classification method for IoT cards provided in another embodiment of this application;

[0062] Figure 4This is a schematic diagram of the structure of an IoT card classification device provided in one embodiment of this application;

[0063] Figure 5 This is a schematic diagram of the hardware structure of the IoT card classification device provided in the embodiments of this application. Detailed Implementation

[0064] The features and exemplary embodiments of various aspects of this application will now be described in detail. Numerous specific details are set forth in the following detailed description to provide a comprehensive understanding of this application. However, it will be apparent to those skilled in the art that this application can be implemented without requiring some of these specific details. The following description of embodiments is merely intended to provide a better understanding of this application by illustrating examples. This application is by no means limited to any specific configurations and algorithms presented below, but covers any modifications, substitutions, and improvements to elements, components, and algorithms without departing from the spirit of this application. Well-known structures and techniques are not shown in the accompanying drawings and the following description in order to avoid unnecessary obfuscation of this application.

[0065] With the development of the Internet of Things (IoT) industry, the use of IoT SIM cards is becoming increasingly widespread. However, the current regulation of IoT SIM cards is more lenient compared to that of human-operated SIM cards, leading to frequent cases of criminals attempting to exploit the targeted voice function of IoT SIM cards to commit telecommunications fraud, seriously affecting the healthy development of IoT businesses. Because IoT SIM cards have call whitelist restrictions, a common practice is to add a number to the whitelist and then set up call forwarding on that number. By using this number to forward calls, the whitelist restriction can be bypassed, allowing the IoT SIM card number to make unrestricted calls to other numbers.

[0066] Currently, to prevent criminals from using the aforementioned methods for illegal activities, the number of call forwarding calls on IoT cards within a certain time period is typically tracked. If the number of call forwarding calls exceeds a preset value, the IoT card is shut down or otherwise disabled. However, ordinary users also have a need to use the call forwarding function when using IoT cards normally. The current method is prone to misjudgment and has low accuracy in risk identification, negatively impacting a large number of legitimate customers. Moreover, the data source used is limited, relying solely on call behavior data from targeted voice IoT cards, resulting in poor real-time performance. Shutdowns and actions are only taken after fraudulent risks occur, lacking early warning and real-time risk awareness.

[0067] To address the aforementioned problems, the inventors discovered that by combining voice detail data, signaling location data, and card basic data from directional voice IoT cards with similar data from human network IoT cards, and employing federated modeling technology, the fusion and modeling analysis of IoT card data and human network data can be achieved. This allows for the establishment of a whitelist number communication circle model for directional voice IoT cards and a data fusion model for IoT cards and human networks. Furthermore, by using data from identified fraudulent directional voice IoT card communication behavior, whitelist number communication behavior, and their whitelist number communication circle model as a training set, supervised machine learning algorithms can be used to extract risk indicators, indicator values, and their influence weights. A multi-dimensional indicator scoring algorithm model can then be used to achieve real-time identification and classification of fraudulent risks associated with directional voice IoT cards. After confirmation and handling, the results can be fed back, forming a closed-loop process. This approach avoids simplistic, one-size-fits-all risk assessments and reduces misjudgments.

[0068] In one embodiment of the classification method for IoT cards in this application, reference can be made to... Figure 1 In this embodiment, the method can be executed by the above-mentioned IoT card classification device, and the method includes the following steps S101 to S104.

[0069] Step S101: Obtain the first card data of the IoT card and the second card data of the communication card number in the call card number whitelist of the IoT card; the second card data includes the calling call data and the user trajectory location information corresponding to the communication card number.

[0070] In practice, the system hosting the IoT SIM card can be connected to its number to query the corresponding whitelist database. When data from a second SIM card is needed, the system retrieves call data and user location information for the specified SIM card number—this is the second data—through an interface with the mobile operator, based on the number information in the whitelist database.

[0071] Step S102: Based on the calling data and the user trajectory location information corresponding to the communication card number, classify the social circle information of the communication card number to obtain the social circle category.

[0072] In practical implementation, caller ID data may include call time, call duration, and call recipient information, while user location information may include latitude and longitude, timestamps, etc. By matching the timestamps in the call log with the timestamps in the user's location information, the geographical location information at the time of the call can be found. Based on factors such as call frequency, call duration, and the geographical location of the call recipient, it is determined whether the call recipient belongs to the user's social circle, and the user's social circle information is categorized. Categorization may include different categories such as family, friends, colleagues, and business partners, and in actual use, it can be further subdivided or merged according to needs.

[0073] Step S103: Based on the preset abnormal call transfer index, determine the target data corresponding to the preset abnormal call transfer index from the first card data and the second card data; the preset abnormal call transfer index includes at least the social circle index; the target data includes at least the social circle category.

[0074] In practical implementation, the first step is to define a series of abnormal call forwarding indicators. These indicators may include social circle indicators and other metrics related to abnormal call forwarding. Specifically, social circle indicators may involve call frequency, call duration, and the geographical location of the caller. Data can be queried and filtered from the first and second SIM card data to select data corresponding to the preset abnormal call forwarding indicators. The selected first and second SIM card data need to be matched to determine the correlation between them, such as matching phone numbers, ID card numbers, call record timestamps, and callers. By analyzing the matched data, the social circle category of each user can be determined, thereby obtaining the target data.

[0075] Step S104: Determine the risk category of the IoT card based on the preset abnormal call transfer index, index weight, and target data.

[0076] In practical implementation, for each abnormal call forwarding indicator, the system needs to determine its relative importance, i.e., indicator weight. Based on the indicator weight and target data, the system can calculate the risk score for each IoT SIM card. Specifically, the risk score can be calculated using a weighted summation method, that is, multiplying the value of each indicator by its corresponding weight and then adding the results to obtain the risk score. Finally, based on the calculated risk scores, the IoT SIM cards can be classified into different risk categories, such as setting different thresholds or score ranges, and classifying the IoT SIM cards into different categories such as low-risk, medium-risk, and high-risk according to the risk score.

[0077] The method described in this application obtains first card data of an IoT card and second card data of communication card numbers from the call card number whitelist of the IoT card. Then, based on the calling data and the user trajectory location information corresponding to the communication card number, the social circle information of the communication card number is categorized to obtain a social circle category. Next, target data corresponding to the preset abnormal call forwarding index is determined from the first and second card data according to the social circle category. Finally, the risk category of the IoT card is determined based on the preset abnormal call forwarding index, the index weight, and the target data. By introducing social circle information, misjudgment is avoided when normal users use call forwarding for numbers within their social circle, resulting in high accuracy in risk identification.

[0078] In some implementations, the calling call data includes the calling call time, and the social circle information of the communication card number is categorized based on the calling call data and the user trajectory location information corresponding to the communication card number to obtain social circle categories, including:

[0079] Extract the caller's call time from the data on the second SIM card.

[0080] In the specific implementation, all outgoing call records are filtered from the call log data in the second card's data set. This can be achieved by querying the database or traversing the call log data, filtering out records where the call type is outgoing. For each outgoing call record, relevant call time information is extracted, specifically the call start time and call end time, which can be obtained directly from the call log.

[0081] The target user's trajectory location information that matches the caller's call time is retrieved from the data of the second card.

[0082] In practice, the target user's trajectory location information is obtained, and this information is typically stored using base station positioning. The trajectory location information may be recorded in time-series format, including timestamps, latitude and longitude coordinates, and other location information. For each call record, the time period of the call is determined. For each defined call time period, the trajectory location information is searched chronologically to find location information matching the call time period.

[0083] Based on the caller's call time and the matching target user's location information, the called numbers of the communication card number are classified to obtain the social circle category.

[0084] In practice, based on information such as call frequency, duration, and location, the system can categorize called numbers into different social circles. For example, called numbers with high call frequency, long call duration, and similar call locations can be grouped into the same category.

[0085] The method described in this application extracts the caller's call time and the corresponding target user's location information from the second card data. Then, based on these two pieces of information, it classifies the called numbers of the communication card number to obtain different social circle categories where the communication card number frequently makes calls at specific times and locations. This allows for accurate classification of the different social circle categories of the numbers called by the communication card number based on its calling patterns.

[0086] In some implementations, classifying the called numbers of the communication card number into social circle categories based on the calling time and the matching target user location information includes:

[0087] From the target user's location trajectory, filter out the user's work location trajectory information corresponding to the caller's call time being on a weekday.

[0088] In the specific implementation, the system obtains the time information of the caller's call record, including the date and the specific call time. Based on the date information of the caller's call record, the system can determine whether the call occurred on a weekday or a non-weekday. For caller's call records from weekdays, the system needs to filter out the corresponding work trajectory location information from the target user's trajectory location information.

[0089] Based on the user's work trajectory location information, the user's work location is determined using the communication card number.

[0090] In practice, for the target user's trajectory location information during the workday, clustering analysis algorithms (such as K-means clustering) can be used to group this location information, thereby finding densely packed location points around the user's possible work locations, which are very likely to be the user's work locations.

[0091] The time when the user with the communication card number is at their workplace is defined as the working period.

[0092] In practice, once a user's work location is determined, a surrounding area can be defined. The user's location trajectory during the workday is then matched against this defined area. When the user's location trajectory falls within this area, the system marks that time period as the user's work period.

[0093] For each called number, the number of the first proactive calls between the communication card number and the called card number are counted from the calling call data, as well as the number of the second proactive calls between the communication card number and the called number at the work location and during the work hours.

[0094] In the specific implementation, for each called number, the number of all active calls between the communication card number and the called card number is counted, that is, the first count. Then, the calls are filtered, and the second count of active calls between the communication card number and the called card number at the workplace and during the working hours is counted in the filtered call records.

[0095] For each called number, calculate the ratio of the second number to the first number.

[0096] The social circle category corresponding to the called number whose ratio is greater than a preset threshold is determined as the working relationship category.

[0097] In the specific implementation, the first and second counts of each called number are obtained, and their ratio is calculated. For each called number, the calculated ratio is compared with a preset threshold. If the ratio is greater than the preset threshold, the called number is marked as a work relationship category; otherwise, no processing is performed or it is marked as another category.

[0098] The method described in this application involves filtering user work location information from the target user's trajectory location data, identifying the user's work location corresponding to the caller's call time being a weekday, determining the user's work location based on this trajectory location information, and then defining the time the user is at their work location as the work period. Next, for each called number, the number of second calls made at the work location and during the work period is counted, along with the total number of first calls. The ratio of the second call to the first call is calculated. When the ratio exceeds a preset threshold, the social circle category corresponding to this number is determined as the work relationship category. The number classified as the work relationship category is determined by calculating the proportion of calls made by each number at the work location and during the work period.

[0099] In some implementations, classifying the called numbers of the communication card number into social circle categories based on the calling time and the matching target user location information includes:

[0100] From the target user's location trajectory, filter out the location trajectory information of users whose call time was on a rest day.

[0101] In the implementation, the time range for rest days is first defined. Typically, rest days can be weekends or specific public holidays. The call records of the target user are iterated through, filtering out call records where the outgoing call occurred on a rest day. For the filtered call records that meet the criteria, the location information corresponding to the call time can be matched using timestamps.

[0102] Based on the user's rest location information, the rest location of the user with the communication card number is determined.

[0103] In the implementation, the user's location trajectory information on rest days is clustered to identify the user's rest location. During the clustering process, the trajectory locations can be divided into different clusters based on the similarity of the location coordinates, with each cluster representing a possible rest location. The identified rest locations are then associated with the communication card number to determine which user the rest location corresponds to.

[0104] The time when the user with the aforementioned communication card number is at a rest location is defined as the rest period.

[0105] For each called number, the number of the first proactive calls between the communication card number and the called card number are counted from the calling call data, as well as the number of the third proactive calls between the communication card number and the called number at the rest location and rest time.

[0106] For each called number, calculate the ratio of the third digit to the first digit.

[0107] Called numbers with a ratio greater than a preset threshold are identified as belonging to the social circle category of family and friends.

[0108] The specific execution process of the above steps is similar to that described above, and the specific execution process will not be repeated here.

[0109] The method described in this application involves filtering user rest location information from the target user's trajectory location that corresponds to the caller's call time being a rest day, thereby determining the user's rest location and defining the time the user is at the rest location as the rest period. Then, for each called number, the number of third calls made at the rest location and rest time is counted, along with the total number of first calls. The ratio of the third call to the first call is calculated. When the ratio exceeds a preset threshold, the social circle category corresponding to this number is determined as the "family and friends" category. The number classified as a family and friends category is determined by calculating the proportion of calls made by each number at rest locations and rest times.

[0110] In some implementations, after determining the risk category of the IoT card based on the preset abnormal call transfer index, index weight, and target data, the method further includes:

[0111] If the risk category is Level 1, service to the first number will be suspended.

[0112] In practice, risk assessment information related to the number is obtained, and when the risk category is classified as "Level 1 Risk," service operations for that number are terminated. This may specifically include: interrupting telephone service (stopping telephone communication services to the number); blocking SMS service (stopping sending or receiving SMS messages to the number); and blocking data service (stopping data services to the number, such as internet access).

[0113] A risk warning will be issued if the risk category is classified as Level 2.

[0114] In practice, when the risk category is classified as Level 2 risk, risk warning notifications are sent to relevant personnel via SMS, email, or app push notifications. Warning information can also be recorded in system logs or a database, including details such as the warning time, reason, and contact number.

[0115] In the case of risk category three, an offline confirmation notification is issued, wherein the levels of the first risk, the second risk, and the third risk increase sequentially.

[0116] In practice, when the risk category is classified as third risk, an offline confirmation notification is sent to relevant personnel, prompting them to confirm or conduct further investigation.

[0117] The methods described in this application involve different actions depending on the risk situation, such as suspending service, issuing risk warnings, or sending offline confirmation notices. This allows for different risk management approaches based on different circumstances.

[0118] In some implementations, after determining the risk category of the IoT card based on the preset abnormal call transfer index, index weight, and target data, the method further includes:

[0119] Obtain the actual risk category of the IoT card.

[0120] In practice, the actual risk category of the IoT card can be obtained by having relevant personnel assess the risk of the associated network card.

[0121] If the actual risk category is inconsistent with the determined risk category, the weight corresponding to the preset abnormal call transfer index shall be adjusted according to the actual risk category and the risk category.

[0122] In practical implementation, if the actual risk category differs from the determined risk category, the specific type of inconsistency needs to be identified. Weights are adjusted based on the inconsistency, and the weight adjustment strategy can be determined according to the actual situation; it can be linear, exponential, or based on other rules. For example, a fixed weight adjustment ratio can be set, or the weights can be dynamically adjusted based on the gap between the actual risk and the determined risk.

[0123] The method described in this application, after determining the risk category, obtains the actual risk category of the IoT card. If the actual risk category differs from the determined risk category, the weight of the preset abnormal call forwarding indicator is adjusted accordingly. This allows for the adjustment of relevant parameters based on the actual risk level, making the results closer to the actual situation.

[0124] In some embodiments, after obtaining the first card data of the IoT card and the second card data of the communication card numbers in the call card number whitelist of the IoT card, the method further includes:

[0125] Extract the first user identifier from the first card data.

[0126] From the Internet of Things (IoT) system, IoT data features corresponding to the first user identifier are extracted to obtain IoT information.

[0127] In the implementation, the identification information of the first user is identified and extracted from the data of the first card through methods such as database queries, API calls, or file reading. This can be a unique identifier for the user, such as a user ID or mobile phone number. Then, based on the identification information of the first user, a request is sent to the IoT system to obtain IoT data related to that user. Specific operations such as filtering queries and data aggregation can be used to ensure that the correct data is obtained.

[0128] Extract the second user identifier from the second card data.

[0129] From the human network system, extract the human network data features corresponding to the second user identifier to obtain human network information.

[0130] In the specific implementation, a request is sent from the human network system to the human network system using the second user identifier to obtain human network data related to that user. This can be achieved through operations such as filtering queries and data aggregation to ensure that the correct data is obtained.

[0131] The IoT information and the human network information are combined to obtain a wide data table.

[0132] The obtained IoT information and human network information are combined to obtain a wide data table containing both types of data.

[0133] The method described in this application extracts IoT data features corresponding to a first user identifier from the IoT system, and similarly extracts human network data features corresponding to a second user identifier from the human network system, thereby obtaining IoT information and human network information. Finally, the IoT information and the human network information are combined to obtain a wide data table. This method enables data fusion between two different systems, breaking down data silos while avoiding data security issues.

[0134] As another implementation of this application, you can refer to Figure 2 , Figure 2 This is a schematic diagram of call forwarding for an IoT card. In this embodiment, IoT data, namely the calling number data, is acquired, as shown in Table 1.

[0135] Table 1

[0136]

[0137]

[0138] Then, obtain the data from the human network, which is the data of the intermediate number. The specific data is shown in Table 2.

[0139] Table 2

[0140]

[0141]

[0142] The acquisition of data from both the Internet of People (IoT) and the Internet of Things (IoT) is highly sensitive because IoT data and IoT data may be stored in different companies and systems, and involve a great deal of user privacy information. To break down data silos without causing data security issues, a multi-system federated modeling approach can be adopted to achieve the integration and exchange of IoT SIM card data and IoT data.

[0143] Specifically, by using the caller and called party user identifiers, human network data and IoT data are fused to form a wide fact table. Based on the caller's user identifier, features are extracted from the data corresponding to this user identifier in the IoT system to obtain abstract feature information. Similarly, features are extracted from the data corresponding to the called party's user identifier in the human network system. Then, the two sets of data are fused to obtain the wide fact table. In this way, the data transmitted between different systems is abstract information after feature extraction; the original information remains stored only in its respective system. Even if the abstract information is intercepted, no security issues will arise.

[0144] In addition, the social circles of whitelisted numbers will be divided based on basic information from the Internet of Things (IoT) data, communication behavior data, and location trajectory data. These social circles are categorized into groups such as friends and family, colleagues, relatives, weak ties, and others. For example, numbers that frequently communicate during working hours (Monday to Friday) and whose location trajectories overlap at work, but communicate less on holidays and after get off work, can be classified as colleagues. Similarly, numbers that rarely contact each other during working hours (Monday to Friday) and whose location trajectories are not together, but frequently contact each other on weekends, holidays, and after get off work, and whose location trajectories overlap, can be classified as relatives. Social circles will serve as an important indicator of abnormal call forwarding. If calls are forwarded to numbers within the relatives' circle, the IoT card has a lower risk of fraud, meaning the indicator has a lower risk score and a lower weight. Conversely, if calls are forwarded to strangers frequently, the risk of fraud is higher, meaning the indicator has a higher risk score and a higher weight. Specific data in the whitelisted number social circle model is shown in Table 3 below.

[0145] Table 3

[0146]

[0147] Using communication behavior data of IoT cards identified as being involved in fraud, along with communication behavior data of their whitelisted numbers and their social circles, and whitelisted number social circle model data, as a training set, abnormal call forwarding risk indicators, indicator values, and the influence weight coefficients of each indicator can be extracted. Then, based on the multi-dimensional indicator scores and indicator weights, the risk score for each indicator is calculated, and the risk scores of each indicator are summed to calculate the total risk score in real time. The specific indicators, indicator values, and indicator weights are extracted through the "Training and Feature Extraction Module" step described above. For example, for the indicator "Frequent updates to whitelist numbers", if the number of changes in the last day N1 <= 2 or the number of changes in the last hour N2 <= 1, the indicator value is 3 points; if the number of changes in the last day 3 <= N1 <= 6 or the number of changes in the last hour 2 <= N2 <= 5, the indicator value is 8 points; if the number of changes in the last day N1 >= 7 or the number of changes in the last hour N2 >= 6, the indicator value is 10 points. The indicator weight is 0.15. The risk score of the indicator is obtained by multiplying the indicator value score by the indicator weight.

[0148] Based on the final comprehensive risk score, the risk of fraud is categorized into four levels: extremely high, high, medium, and low. The process then proceeds to risk handling. Extremely high-risk cards are immediately automatically shut down; high-risk cards trigger a risk warning, and management personnel immediately confirm and verify the information; medium-risk cards are confirmed offline. After confirmation and handling, the results are fed back, and the model parameters are optimized based on the confirmed results, ultimately forming a closed-loop process. The overall process can be referenced below. Figure 3 .

[0149] Based on the IoT card classification method provided in the above embodiments, this application also provides specific implementations of an IoT card classification device. Please refer to the following embodiments.

[0150] First see Figure 4 The IoT card classification device 400 provided in this application embodiment includes the following modules:

[0151] The acquisition module 401 is used to acquire the first card data of the IoT card and the second card data of the communication card number in the call card number whitelist of the IoT card; the second card data includes the calling call data and the user trajectory location information corresponding to the communication card number.

[0152] The classification module 402 is used to classify the social circle information of the communication card number based on the calling call data and the user trajectory location information corresponding to the communication card number, so as to obtain the social circle category.

[0153] The determining module 403 is used to determine target data corresponding to the preset abnormal call transfer index from the first card data and the second card data according to the preset abnormal call transfer index; the preset abnormal call transfer index includes at least a social circle index; the target data includes at least a social circle category.

[0154] The determination module 403 is further configured to determine the risk category of the IoT card based on the preset abnormal call transfer index, index weight and target data.

[0155] The method described in this application obtains first card data of an IoT card and second card data of communication card numbers from the call card number whitelist of the IoT card. Then, based on the calling data and the user trajectory location information corresponding to the communication card number, the social circle information of the communication card number is categorized to obtain a social circle category. Next, target data corresponding to the preset abnormal call forwarding index is determined from the first and second card data according to the social circle category. Finally, the risk category of the IoT card is determined based on the preset abnormal call forwarding index, the index weight, and the target data. By introducing social circle information, misjudgment is avoided when normal users use call forwarding for numbers within their social circle, resulting in high accuracy in risk identification.

[0156] As one implementation of this application, the IoT card classification device 400 further includes:

[0157] The extraction module is used to extract the caller's call time from the data of the second card.

[0158] The search module is used to search for the target user's trajectory location information that matches the caller's call time from the data of the second card.

[0159] The classification module is used to classify the called numbers of the communication card number according to the calling time and the matching target user trajectory location information to obtain the social circle category.

[0160] The method described in this application extracts the caller's call time and the corresponding target user's location information from the second card data. Then, based on these two pieces of information, it classifies the called numbers of the communication card number to obtain different social circle categories where the communication card number frequently makes calls at specific times and locations. This allows for accurate classification of the different social circle categories of the numbers called by the communication card number based on its calling patterns.

[0161] As one implementation of this application, the IoT card classification device 400 further includes:

[0162] The filtering module is used to filter out the work trajectory location information of users whose call time is a weekday from the target user's trajectory location.

[0163] The determination module is used to determine the work location of the user with the communication card number based on the user's work trajectory location information.

[0164] The determination module is also used to determine the time when the user of the communication card number is at the workplace as the working period.

[0165] The statistics module is used to count, from the calling call data, the first number of proactive calls between the communication card number and the called card number, and the second number of proactive calls between the communication card number and the called number at the work location and during the work hours for each called number.

[0166] The calculation module is used to calculate the ratio of the second number to the first number for each called number.

[0167] The determination module is also used to determine the social circle category corresponding to called numbers whose ratio is greater than a preset threshold as the working relationship category.

[0168] The method described in this application involves filtering user work location information from the target user's trajectory location data, identifying the user's work location corresponding to the caller's call time being a weekday, determining the user's work location based on this trajectory location information, and then defining the time the user is at their work location as the work period. Next, for each called number, the number of second calls made at the work location and during the work period is counted, along with the total number of first calls. The ratio of the second call to the first call is calculated. When the ratio exceeds a preset threshold, the social circle category corresponding to this number is determined as the work relationship category. The number classified as the work relationship category is determined by calculating the proportion of calls made by each number at the work location and during the work period.

[0169] As one implementation of this application, the IoT card classification device 400 further includes:

[0170] The filtering module is used to filter out the rest location information of users whose call time is a rest day from the target user's trajectory location.

[0171] The determination module is used to determine the rest location of the user with the communication card number based on the user's rest trajectory location information.

[0172] The determining module is also used to determine the time when the user of the communication card number is at the rest location as the rest period.

[0173] The statistics module is used to count, from the calling call data, the first number of proactive calls between the communication card number and the called card number for each called number, as well as the third number of proactive calls between the called number and the called number at the rest location and rest time.

[0174] The calculation module is used to calculate the ratio of the third digit to the first digit for each called number.

[0175] The determination module is also used to determine the social circle category corresponding to called numbers whose ratio is greater than a preset threshold as the family and friends relationship category.

[0176] The method described in this application involves filtering user rest location information from the target user's trajectory location that corresponds to the caller's call time being a rest day, thereby determining the user's rest location and defining the time the user is at the rest location as the rest period. Then, for each called number, the number of third calls made at the rest location and rest time is counted, along with the total number of first calls. The ratio of the third call to the first call is calculated. When the ratio exceeds a preset threshold, the social circle category corresponding to this number is determined as the "family and friends" category. The number classified as a family and friends category is determined by calculating the proportion of calls made by each number at rest locations and rest times.

[0177] As one implementation of this application, the IoT card classification device 400 further includes:

[0178] The stop module is used to stop providing service to the first number when the risk category is the highest risk level.

[0179] The early warning module is used to issue risk warnings when the risk category is the second risk level.

[0180] The notification module is used to issue an offline confirmation notification when the risk category is the third risk, wherein the levels of the first risk, the second risk, and the third risk increase sequentially.

[0181] The methods described in this application involve different actions depending on the risk situation, such as suspending service, issuing risk warnings, or sending offline confirmation notices. This allows for different risk management approaches based on different circumstances.

[0182] As one implementation of this application, the IoT card classification device 400 further includes:

[0183] The acquisition module is used to acquire the actual risk category of the IoT card.

[0184] The adjustment module is used to adjust the weight corresponding to the preset abnormal call transfer indicator according to the actual risk category and the determined risk category when the actual risk category is inconsistent with the determined risk category.

[0185] The method described in this application, after determining the risk category, obtains the actual risk category of the IoT card. If the actual risk category differs from the determined risk category, the weight of the preset abnormal call forwarding indicator is adjusted accordingly. This allows for the adjustment of relevant parameters based on the actual risk level, making the results closer to the actual situation.

[0186] As one implementation of this application, the IoT card classification device 400 further includes:

[0187] The extraction module is used to extract the first user identifier from the first card data.

[0188] The extraction module is also used to extract IoT data features corresponding to the first user identifier from the IoT system to obtain IoT information.

[0189] The extraction module is also used to extract the second user identifier from the second card data.

[0190] The extraction module is also used to extract human network data features corresponding to the second user identifier from the human network system to obtain human network information.

[0191] The combination module is used to combine the IoT information and the human network information to obtain a wide data table.

[0192] The method described in this application extracts IoT data features corresponding to a first user identifier from the IoT system, and similarly extracts human network data features corresponding to a second user identifier from the human network system, thereby obtaining IoT information and human network information. Finally, the IoT information and the human network information are combined to obtain a wide data table. This method enables data fusion between two different systems, breaking down data silos while avoiding data security issues.

[0193] Each module in the IoT card classification device provided in this application embodiment can implement each step in the above-described IoT card classification method and achieve the corresponding effect. For the sake of brevity, it will not be described in detail here.

[0194] Figure 5 A schematic diagram of the structure of the classification hardware of the Internet of Things card provided in the embodiments of this application is shown.

[0195] The classification device for IoT cards may include a processor 501 and a memory 502 storing computer program instructions.

[0196] Specifically, the processor 501 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0197] Memory 502 may include mass storage for data or instructions. For example, and not limitingly, memory 502 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 502 may include removable or non-removable (or fixed) media. Where appropriate, memory 502 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 502 is non-volatile solid-state memory.

[0198] The memory may include read-only memory (ROM), random access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, and electrical, optical, or other physical / tangible memory storage devices. Therefore, typically, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described in the classification method of IoT cards according to any embodiment of this disclosure.

[0199] The processor 501 reads and executes computer program instructions stored in the memory 502 to implement any of the IoT card classification methods in the above embodiments.

[0200] In one example, the IoT card sorting device may also include a communication interface 503 and a bus 510. Wherein, as Figure 5 As shown, the processor 501, memory 502, and communication interface 503 are connected through bus 510 and complete communication with each other.

[0201] The communication interface 503 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0202] Bus 510 includes hardware, software, or both, that couples components of an online data traffic metering device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 510 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0203] Furthermore, in conjunction with the IoT card classification methods in the above embodiments, this application embodiment can provide a computer storage medium for implementation. This computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the IoT card classification methods in the above embodiments.

[0204] However, it should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. Furthermore, for the sake of brevity, detailed descriptions of known methods and techniques are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0205] The functional blocks shown in the above-described structural diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0206] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0207] The aspects of this application have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by dedicated hardware performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0208] This application may be implemented in other specific forms without departing from its spirit and essential characteristics. For example, the algorithm described in a particular embodiment may be modified without departing from the basic spirit of this application. Therefore, the present embodiments are to be regarded as exemplary rather than limiting in all respects, and the scope of this application is defined by the appended claims rather than the foregoing description, and all changes falling within the meaning and scope of the claims and their equivalents are thus included within the scope of this application.

[0209] Those skilled in the art will understand that the above embodiments are exemplary and not restrictive. Different technical features appearing in different embodiments can be combined to achieve beneficial effects. Based on a study of the drawings, specification, and claims, those skilled in the art should be able to understand and implement other variations of the disclosed embodiments. In the claims, the term "comprising" does not exclude other means or steps; the indefinite article "a" does not exclude a plurality; the terms "first" and "second" are used to identify names and not to indicate any particular order. No reference numerals in the claims should be construed as limiting the scope of protection. The functionality of multiple parts appearing in the claims can be implemented by a single hardware or software module. The appearance of certain technical features in different dependent claims does not mean that these technical features cannot be combined to achieve beneficial effects.

Claims

1. A method for classifying Internet of Things (IoT) cards, characterized in that, The method includes: Obtain the first card data of the IoT card; Query the call card number whitelist database corresponding to the IoT card based on the IoT card number; The second card data is determined based on the number information in the call card number whitelist database; the second card data includes outgoing call data and user trajectory location information corresponding to the communication card number. Extract the first user identifier from the first card data; extract the IoT data features corresponding to the first user identifier from the IoT system to obtain IoT information; Extract the second user identifier from the second card data; extract the human network data features corresponding to the second user identifier from the human network system to obtain human network information; By combining the IoT information and the human network information, a wide data table is obtained; Based on the caller's data and the user's location information corresponding to the communication card number, the social circle information of the communication card number is categorized to obtain social circle categories. These categories serve as social circle indicators in a preset abnormal call forwarding index to measure the degree of fraud risk when the IoT card communicates with the communication card number via call forwarding. The social circle categories include family relationship categories, work relationship categories, and stranger categories. The risk of forwarding calls to family relationship category numbers is lower than the risk of forwarding calls to stranger category numbers; the indicator weight for forwarding calls to family relationship category numbers is lower than the indicator weight for forwarding calls to stranger category numbers. Based on preset abnormal call forwarding indicators, target data corresponding to the preset abnormal call forwarding indicators are determined from the data wide table; the preset abnormal call forwarding indicators include at least social circle indicators; the target data includes at least social circle categories. The risk category of the IoT card is determined based on the preset abnormal call transfer index, the index weight corresponding to the social circle category, and the target data.

2. The classification method for IoT cards according to claim 1, characterized in that, The calling call data includes the calling call time. Based on the calling call data and the user trajectory location information corresponding to the communication card number, the social circle information of the communication card number is classified to obtain social circle categories, including: Extract the caller's time from the data on the second SIM card; Search the second card data for the target user's trajectory location information that matches the caller's call time; Based on the caller's call time and the matching target user's location information, the called numbers of the communication card number are classified to obtain the social circle category.

3. The classification method for IoT cards according to claim 2, characterized in that, The method of classifying the called numbers of the communication card number according to the calling time and the matching target user trajectory location information to obtain social circle categories includes: From the target user's location trajectory, filter out the user's work location trajectory information corresponding to the caller's call time being on a weekday; Based on the user's work trajectory location information, determine the work location of the user with the communication card number; The time when the user of the aforementioned communication card number is at their workplace is defined as the working period. For each called number, count the first number of proactive calls between the communication card number and the called card number from the calling call data, and the second number of proactive calls between the communication card number and the called number at the work location and during the work hours. For each called number, calculate the ratio of the second number to the first number; The social circle category corresponding to the called number whose ratio is greater than a preset threshold is determined as the working relationship category.

4. The classification method for IoT cards according to claim 2, characterized in that, The method of classifying the called numbers of the communication card number according to the calling time and the matching target user trajectory location information to obtain social circle categories includes: From the target user's location trajectory, filter out the location trajectory information of users whose call time was a rest day; Based on the user's rest trajectory location information, determine the rest location of the user with the communication card number; The time when the user with the aforementioned communication card number is at a rest location is defined as the rest period. For each called number, the number of the first proactive calls between the communication card number and the called card number are counted from the calling call data, as well as the number of the third proactive calls between the communication card number and the called number at the rest location and rest time. For each called number, calculate the ratio of the third digit to the first digit; Called numbers with a ratio greater than a preset threshold are identified as belonging to the social circle category of family and friends.

5. The classification method for IoT cards according to claim 4, characterized in that, After determining the risk category of the IoT card based on the preset abnormal call transfer index, index weight, and target data, the method further includes: If the risk category is Level 1, service to the first number will be suspended. The first number is the number corresponding to the data of the first card. When the risk category is classified as the second risk level, a risk warning will be issued. In the case of risk category three, an offline confirmation notification is issued, wherein the levels of the first risk, the second risk, and the third risk increase sequentially.

6. The classification method for IoT cards according to claim 4, characterized in that, After determining the risk category of the IoT card based on the preset abnormal call transfer index, index weight, and target data, the method further includes: Obtain the actual risk category of the IoT card; If the actual risk category is inconsistent with the determined risk category, the weight corresponding to the preset abnormal call transfer index shall be adjusted according to the actual risk category and the risk category.

7. A sorting device for Internet of Things (IoT) cards, characterized in that, The device includes: The acquisition module is used to acquire the first card data of the IoT card; The query module is used to query the call card number whitelist database corresponding to the IoT card based on the IoT card number; The second card data determination module is used to determine the second card data based on the number information in the call card number whitelist database; the second card data includes outgoing call data and user trajectory location information corresponding to the communication card number. The first feature extraction module is used to extract a first user identifier from the first card data; and to extract IoT data features corresponding to the first user identifier from the IoT system to obtain IoT information. The second feature extraction module extracts the second user identifier from the second card data; and extracts the human network data features corresponding to the second user identifier from the human network system to obtain human network information. The feature fusion module is used to combine the IoT information and the human network information to obtain a wide data table; The classification module is used to classify the social circle information of the communication card number based on the calling call data and the user trajectory location information corresponding to the communication card number, thereby obtaining social circle categories. These social circle categories serve as social circle indicators in a preset abnormal call forwarding index to measure the degree of fraud risk when the IoT card communicates with the communication card number via call forwarding. The social circle categories include family relationship categories, work relationship categories, and stranger categories. The risk level of call forwarding to family relationship category numbers is lower than that of call forwarding to stranger category numbers; the indicator weight for call forwarding to family relationship category numbers is lower than that for call forwarding to stranger category numbers. The determination module is used to determine target data corresponding to the preset abnormal call transfer indicators from the data wide table based on the preset abnormal call transfer indicators; the preset abnormal call transfer indicators include at least social circle indicators; the target data includes at least social circle categories; The determination module is also used to determine the risk category of the IoT card based on the preset abnormal call transfer index, the index weight corresponding to the social circle category, and the target data.

8. A sorting device for Internet of Things (IoT) cards, characterized in that, The device includes: a processor and a memory storing computer program instructions; the processor reads and executes the computer program instructions to implement the classification method for IoT cards as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program instructions, which, when executed by a processor, implement the IoT card classification method as described in any one of claims 1-6.

10. A computer program product, characterized in that, When the instructions in the computer program product are executed by the processor of the electronic device, the electronic device causes the electronic device to perform the IoT card classification method as described in any one of claims 1-6.

Images