User identification method and device, equipment, storage medium and program product

Abstract

The invention discloses a user identification method and device, equipment, a storage medium and a program product. According to the specific technical scheme, the method comprises the following steps: acquiring related information of a to-be-detected number, wherein the related information comprises position information, associated number information and communication information of the to-be-detected number; calculating the probability that the to-be-detected number is a marked number according to the position information of the to-be-detected number, the associated number information, the communication information message content information and the internet access data information; under the condition that the probability is greater than a preset probability threshold value, judging whether the communication information comprises preset mark content or not; and determining a user corresponding to the to-be-detected number as a marked user under the condition that the communication information comprises the preset marked content.

Description

Technical Field

[0001] The present application belongs to the field of communication information security technology, and in particular, relates to a method, device, equipment, storage medium and program product for user identification. Background Art

[0002] With the rapid development of information technology, the information exchanged between devices can include text, pictures, audio, video, etc. Based on the increasingly frequent information exchange, the devices of black and gray industry users in the telecommunications network may send information containing illegal content to the devices of ordinary users.

[0003] In the related art, numbers in the telecommunications network are identified through predefined identification rules to identify illegal and gray market users. However, since the identification rules are manually defined and there is no guarantee that the identification rules will be updated in a timely manner, the accuracy of identifying illegal and gray market users is low. Since illegal and gray market users cannot be accurately identified, they pose a threat to the life or property safety of other users. Summary of the invention

[0004] The embodiments of the present application provide a method, apparatus, device, storage medium and program product for user identification, which can accurately identify illegal and gray market users and prevent illegal and gray market users from threatening the life or property safety of other users.

[0005] In a first aspect, an embodiment of the present application provides a method for user identification, including:

[0006] Acquire relevant information of the number to be detected, the relevant information including location information, associated number information and communication information of the number to be detected;

[0007] Calculating the probability that the number to be detected is a marked number according to the location information of the number to be detected, the associated number information and the communication information;

[0008] When the probability is greater than a preset probability threshold, determining whether the communication information includes preset marking content;

[0009] In a case where the communication information includes the preset marking content, the user corresponding to the number to be detected is determined as a marked user.

[0010] In a possible implementation, the associated number information includes the number of associated numbers associated with the number to be detected and the location information of the associated numbers; and the calculating the probability that the number to be detected is a marked number according to the location information of the number to be detected, the associated number information, the message content information and the Internet access data information includes:

[0011] When the number of associated numbers is greater than a first preset threshold, determining a product of the number of associated numbers and a first preset probability as a first risk score;

[0012] Calculating a first distance between the number to be detected and the associated number according to the location information of the number to be detected and the location information of the associated number, and obtaining a second risk score when the first distance is greater than a second preset threshold;

[0013] Calculating a second distance between the position of the number to be detected and the preset mark position according to the position information of the number to be detected, and obtaining a third risk score when the second distance is less than a third preset threshold;

[0014] determining a communication risk score based on the communication information;

[0015] Calculating the sum of the first risk score, the second risk score, the third risk score and the communication risk score to obtain a sum value to be detected;

[0016] The sum to be detected is used as the probability that the number to be detected is a marked number.

[0017] In a possible implementation manner, the communication information includes Internet access data information; and calculating the communication risk score according to the communication information includes:

[0018] In a case where the online data information and the preset marked online data information are the same, a fourth risk score is obtained, and the fourth risk score is used as the communication risk score.

[0019] In a possible implementation manner, the communication information further includes message content information; and calculating the communication risk score according to the communication information includes:

[0020] According to the message content information, count the sending frequency and the sending quantity of the number to be detected, the number of receiving numbers that receive the message of the number to be detected, and the number of messages with the same content;

[0021] Respectively comparing the number of messages with a fourth preset threshold, the sending frequency with a fifth preset threshold, the sending quantity with a sixth preset threshold, the number of receiving numbers with the seventh preset threshold, counting the number of indicators greater than or equal to the corresponding preset thresholds among the number of messages, the sending frequency, the sending quantity, and the number of receiving numbers, and calculating the product of the number of indicators and a second preset probability to obtain a fifth risk score;

[0022] A sum of the fifth risk score and the fourth risk score is calculated to obtain the communication risk score.

[0023] In a possible implementation, the calculating the probability that the number to be detected is a marked number according to the location information of the number to be detected, the associated number information, the message content information and the Internet access data information includes:

[0024] The number to be detected, the location information of the number to be detected, the associated number information and the communication information are input as input data into a user identification model, and the user identification model predicts a prediction result based on the input data;

[0025] Among them, the prediction result includes the probability that the number to be detected is a marked number, and the user recognition model is trained using sample-related information of the marked number and the marked user corresponding to the marked number, and the sample-related information includes sample location information and sample communication information of the marked user number.

[0026] In a possible implementation, before inputting the number to be detected, the location information of the number to be detected, the associated number information, the message content information, and the online data information into a number recognition model to obtain a prediction result output by the number recognition model, the method further includes:

[0027] Obtaining information related to the sample;

[0028] Inputting the sample related information into a neural network model, and the neural network model outputs a sample prediction result;

[0029] Using the marked number as label data, calculating the loss function of the label data and the sample prediction result;

[0030] The parameters of the neural network model are optimized according to the loss function to obtain the user recognition model.

[0031] In a possible implementation, optimizing the parameters of the neural network model according to the loss function to obtain the user identification model includes:

[0032] Optimizing the parameters of the neural network model according to the loss function to obtain the target model;

[0033] The model performance of the target model is evaluated using a verification data set, and when the recognition accuracy of the target model is greater than or equal to a preset accuracy threshold, the target model is used as the user recognition model.

[0034] In a possible implementation, the evaluating the model performance of the target model by using the validation data set, and using the target model as the user recognition model when the recognition accuracy of the target model is greater than or equal to a preset accuracy threshold, includes:

[0035] Acquire verification related information, wherein the verification related information includes verification location information and verification communication information of the user number to be verified;

[0036] Inputting the verification related information into the target model to obtain a recognition result output by the target model;

[0037] Verifying the recognition result according to the tag data of the user number to be verified;

[0038] Counting the first target number of the tag data of the user number to be verified that is consistent with the recognition result, and the second target number of the user number to be verified;

[0039] Calculating a ratio of the first target quantity to the second target quantity to obtain a recognition accuracy rate;

[0040] When the recognition accuracy is greater than or equal to a preset accuracy threshold, the target model is used as the user recognition model.

[0041] In a possible implementation manner, the determining whether the communication information includes preset marking content includes:

[0042] Determining target type information corresponding to the communication information;

[0043] According to the correspondence between the type information and the identification method, obtaining the target identification method corresponding to the target type information;

[0044] The preset marking content in the communication information is identified according to the target identification method.

[0045] In a possible implementation, the communication information includes text message type information, picture message type information, and audio message type information; and identifying preset marked content in the communication information according to the target recognition method includes:

[0046] In the case where the communication information is the text message type information, matching the text content in the communication information according to the preset keywords;

[0047] In the case where the communication information is the picture message type information, obtaining picture content in the communication information, performing image feature extraction on the picture content, obtaining target features, and searching for the target features in a mark feature library;

[0048] In the case that the communication information is of the audio message type, speech-to-text processing is performed on the communication information to obtain text information, and the text information is matched according to the preset keywords.

[0049] In a second aspect, an embodiment of the present application provides a user identification device, including:

[0050] An acquisition module, used to acquire relevant information of the number to be detected, wherein the relevant information includes location information, associated number information and communication information of the number to be detected;

[0051] A calculation module, used for calculating the probability that the number to be detected is a marked number according to the location information of the number to be detected, the associated number information and the communication information;

[0052] A judgment module, used for judging whether the communication information includes preset marking content when the probability is greater than a preset probability threshold;

[0053] The determination module is used to determine the user corresponding to the number to be detected as a marked user when the communication information includes the preset marking content.

[0054] In a third aspect, an embodiment of the present application provides an electronic device, the device comprising: a processor and a memory storing computer program instructions;

[0055] When the processor executes the computer program instructions, the method according to any one of the first aspects is implemented.

[0056] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described in any one of the first aspects is implemented.

[0057] In a fifth aspect, an embodiment of the present application provides a computer program product, the computer product comprising a computer program, and when the computer program is executed by a processor, the method as described in any one of the first aspects is implemented.

[0058] A method, device, equipment, storage medium and program product for user identification provided by an embodiment of the present application. After obtaining relevant information of a number to be detected, the probability that the user corresponding to the number to be detected is a marked user is calculated according to the relevant information. Wherein, the relevant information includes location information, associated number information, and communication information of the number to be detected. The number to be detected is calculated based on data of multiple dimensions such as comprehensive relevant information, marked position and marked Internet access data, and the probability that the number to be detected is a marked number is obtained, and then when the probability is greater than a preset probability threshold, the user corresponding to the number to be detected is determined to be a marked user. In this way, the probability that the number to be detected is a marked number is calculated by multiple information in the relevant information, and it can be judged from multiple dimensions whether the number to be detected is the number of a black and gray production user, thereby improving the recognition accuracy of black and gray production users, and on the basis of identifying black and gray production users, the threat of black and gray production users to the life or property safety of other users is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0060] Figure 1 It is a flowchart of a method for user identification provided by an embodiment of the present application;

[0061] Figure 2 is an exemplary schematic diagram of a user identification method provided in an embodiment of the present application;

[0062] Figure 3 It is a schematic diagram of a system structure of a method for user identification provided by an embodiment of the present application using an embodiment of the present application;

[0063] Figure 4 is a structural diagram of a user identification device provided in an embodiment of the present application;

[0064] Figure 5 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0065] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.

[0066] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0067] In order to solve the problems of the prior art, the embodiments of the present application provide a method, apparatus, device, storage medium and program product for user identification. Figure 1 The user identification method provided by the embodiment of the present application is introduced, such as Figure 1 As shown, the method includes:

[0068] S101, obtaining relevant information of the number to be detected.

[0069] The relevant information includes the location information, associated number information and communication information of the number to be detected. The associated number refers to the number under the same user name, except for the number of the user to be detected. The associated number information may include the location information and quantity information of the associated number.

[0070] Specifically, the electronic device can obtain the location information of the number to be detected in real time through the base station, the associated number information can be obtained from a third-party database, and the communication information can include the interaction information between the number to be detected and the outside world, such as Internet access data or message information sent by the user to other users through the number to be detected. Among them, the communication information can be obtained from the third-party database.

[0071] S102: Calculate the probability that the number to be detected is a marked number based on the location information, associated number information and communication information of the number to be detected.

[0072] It can be understood that the location information, associated number information and communication information of the number to be detected can be used to conduct risk assessment on the user corresponding to the number to be detected from the distance dimension, the relationship dimension between the number to be detected and the associated number, and the user's interactive behavior dimension, respectively, so as to judge the probability that the number to be detected is a marked number based on the risk assessment result.

[0073] S103: When the probability is greater than a preset probability threshold, determine whether the communication information includes preset marking content.

[0074] The preset probability threshold is set based on experience, and the preset marking content refers to the content corresponding to the black and gray production behavior. For example, a user sends a text message containing a blacklist link or marketing advertisement through the number to be detected. The embodiment of the present application does not specifically limit the preset marking content.

[0075] S104: When the communication information includes preset marking content, determine the user corresponding to the number to be detected as a marked user.

[0076] Adopting the above method, after obtaining the relevant information of the number to be detected, the probability that the user corresponding to the number to be detected is a marked user is calculated according to the relevant information. Wherein, the relevant information includes the position information, associated number information, and communication information of the number to be detected. The data of multiple dimensions such as comprehensive relevant information, mark position and mark Internet access data are calculated for the number to be detected, and the probability that the number to be detected is a marked number is obtained, and then when the probability is greater than a preset probability threshold, the user corresponding to the number to be detected is determined to be a marked user. In this way, the probability that the number to be detected is a marked number is calculated by multiple information in the relevant information, and whether the number to be detected is the number of a black and gray production user can be judged from multiple dimensions, thereby improving the recognition accuracy of the black and gray production user, and on the basis of identifying the black and gray production user, the threat of the black and gray production user to the life or property safety of other users is reduced.

[0077] In one embodiment, for the above S102, the probability that the number to be detected is a marked number is calculated according to the location information, associated number information and communication information of the number to be detected, and there are two specific implementation methods, including:

[0078] Implementation method 1: The electronic device uses the location information, associated number information and communication information of the number to be detected to perform risk scoring on the number to be detected. Each risk score is used to measure the probability that the number to be detected is a black or gray product number in the dimension corresponding to the risk score.

[0079] Implementation method 2: Input the location information, associated number information and communication information of the number to be detected into the trained user recognition model, and use the user recognition model to output the probability that the number to be detected is a marked number.

[0080] For implementation mode 1, specifically, the associated number information includes the number of associated numbers associated with the number to be detected and the location information of the associated numbers. Then S102, according to the location information, associated number information and communication information of the number to be detected, calculate the probability that the number to be detected is a marked number. The specific implementation process may include steps 1 to 6:

[0081] Step 1: When the number of associated numbers is greater than a first preset threshold, determine the product of the number of associated numbers and a first preset probability as a first risk score.

[0082] The first preset threshold and the first preset probability are set based on experience.

[0083] It is understandable that the more numbers associated with the same user name, the higher the probability that the user is a black and gray industry user.

[0084] Step 2: Calculate a first distance between the number to be detected and the associated number based on the location information of the number to be detected and the location information of the associated number, and obtain a second risk score when the first distance is greater than a second preset threshold.

[0085] The second preset threshold and the second risk score are set based on experience.

[0086] It should be noted that if the first distance between the associated numbers under the same user name is greater than the preset second preset threshold, it means that the number is not used by the user himself, and therefore, the second risk score is obtained.

[0087] Step 3: Calculate the second distance between the position of the number to be detected and the preset mark position according to the position information of the number to be detected, and obtain the third risk score when the second distance is less than the third preset threshold.

[0088] The third preset threshold and the third risk score are set based on experience.

[0089] The preset marked position is the position of the marked black and gray industry user. The electronic device can obtain the above preset marked position from a preset black and gray industry user information storage database. The black and gray industry user information storage database stores information such as the location information and online data information of the identified black and gray industry users.

[0090] Step 4: Determine a communication risk score based on the communication information.

[0091] Step 5: Calculate the sum of the first risk score, the second risk score, the third risk score and the communication risk score to obtain a sum value to be detected.

[0092] Step 6: Use the sum to be detected as the probability that the number to be detected is a marked number.

[0093] Using the method for calculating the probability provided by the above-mentioned implementation 1, when the number of associated numbers is greater than the first preset threshold value, it means that the number opened by the user is more, and the more associated numbers under the same user name, the higher the probability that the user is a black and gray production user. In the case where the distance between the number to be detected and the associated number is greater than the second preset threshold value, it means that the distance between the associated number of the user and the number to be detected is far, and the user may use the associated number to perform illegal operations. In the case where the distance between the number to be detected and the preset mark position is less than the third preset threshold value, it means that the possibility that the number to be detected is a black and gray production number is high. Then calculate the sum of the first risk score, the second risk score, the third risk score and the communication risk score, and the probability that the number to be detected is a black and gray production number can be measured by combining multiple dimensions, and accordingly, the probability can be used as the probability that the corresponding user of the number to be detected is a black and gray production user. In this way, the probability of the user being a black and gray production user can be judged from multiple dimensions to improve the accuracy of the calculation result.

[0094] In one embodiment, for the above step 4, the specific implementation of determining the communication risk score according to the communication information may be:

[0095] If the communication information includes online data information, when the online data information is the same as the preset marked online data information, a fourth risk score is obtained, and the fourth risk score is used as the communication risk score.

[0096] The electronic device can obtain the online data of the black and gray industry users from the black and gray industry user information storage database, and use the online data of the black and gray industry users as the preset marked online data. The online data can be the Internet Protocol (IP) for online access.

[0097] In this way, by comparing the preset marked online data and the online data information in the communication information, it is determined whether there is any illegal online behavior in the online behavior of the number to be detected. If there is any illegal online behavior, the fourth risk score can be obtained, thereby improving the accuracy of the prediction result.

[0098] In some embodiments of the present application, the above communication information includes Internet access data information, and the communication information also includes message content information. Step 4, determining a communication risk score according to the communication information, can be specifically implemented as follows:

[0099] According to the message content information, the sending frequency and number of the number to be detected, the number of receiving numbers that receive messages from the number to be detected, and the number of messages with the same content are counted. The number of messages is compared with the fourth preset threshold, the sending frequency is compared with the fifth preset threshold, the sending number is compared with the sixth preset threshold, and the number of receiving numbers is compared with the seventh preset threshold. The number of indicators greater than or equal to the corresponding preset thresholds in the number of messages, the sending frequency, the sending number, and the number of receiving numbers is counted, and the product of the number of indicators and the second preset probability is calculated to obtain the fifth risk score. The sum of the fifth risk score and the fourth risk score is calculated to obtain the communication risk score.

[0100] The electronic device can count the sending behavior data of the number to be detected within the preset time period according to the message content information within the preset time period, wherein the sending behavior data is the above-mentioned sending frequency, sending quantity, number of receiving numbers and number of messages with the same content.

[0101] The electronic device can also collect statistics on the sending behavior data of the number to be detected from the time when the number is successfully registered to the current time.

[0102] Among them, the fourth preset threshold, the fifth preset threshold, the sixth preset threshold, the seventh preset threshold and the second preset probability are set based on experience.

[0103] In one example, the fourth preset threshold is S1, the fifth preset threshold is S2, the sixth preset threshold is S3, the seventh preset threshold is S4, the second preset probability is P2, and the fourth risk score is H4. The electronic device compares the number of messages with the fourth preset threshold S1, the sending frequency with the fifth preset threshold S2, the number of messages sent with the sixth preset threshold S3, and the number of received numbers with the seventh preset threshold S4. Among them, if the number of messages is greater than the fourth preset threshold S1, the number of messages sent is greater than the sixth preset threshold S3, and the number of received numbers is greater than the seventh preset threshold S4, then the number of indicators is 3. When the online data information and the preset marked online data information are the same, the communication risk score is calculated as 3*P2+H4.

[0104] Using the method provided by the embodiment of the present application, when the communication information includes the above-mentioned Internet data information and message content information, the sending frequency, sending quantity, receiving number number and the number of messages with the same content of the number to be detected are counted, and then the number of indicators greater than or equal to the corresponding preset threshold value in the above indicators is judged in turn. Among them, the sending frequency, sending quantity, receiving number number and the number of messages with the same content can be used to judge whether there is a violation from the message corresponding to the user, and the message content can directly reflect whether the user has violated the rules. Therefore, the fifth risk score calculated based on the number of indicators can more accurately reflect the possibility that the number to be detected is a marked number. Further, the calculated communication risk score can improve the accuracy of the prediction result that the user is a marked user.

[0105] The following combination Figure 2 The above is the user identification method provided in implementation mode 1, such as Figure 2 As shown, the method includes:

[0106] S201. Score the probability of illegal activities based on the number under the user name.

[0107] Among them, the number under the user name is the associated number under the user name corresponding to the number to be detected, and the black industry probability score is the above-mentioned first risk score.

[0108] Specifically, the number of numbers associated with the user name is counted, and the product of the number of associated numbers and the first preset probability is calculated to obtain a first risk score.

[0109] S202. Score the probability of illegal activities based on the positional differences between different numbers under the name.

[0110] Among them, the black industry probability score is the above-mentioned second risk score. The specific method of obtaining the second risk score refers to the relevant description in the above embodiment, which will not be repeated here.

[0111] S203, scoring the probability of illegal production based on the difference between the illegal production high-risk location and the marked illegal production high-risk location.

[0112] Among them, the marked high-risk location of the black industry is the above-mentioned preset marked location information, and the black industry probability score is the above-mentioned third risk score.

[0113] S204. Compare with the marked black market user IP data and perform black market probability scoring.

[0114] Among them, the IP data of users marked as black market users is the above-mentioned preset marked Internet access data information, and the black market probability score is the above-mentioned fourth risk score.

[0115] S205. Scoring the probability of black production based on the information sending behavior.

[0116] Among them, the sending behavior refers to the above-mentioned sending frequency, sending quantity, number of receiving numbers and the number of messages with the same content. The black industry probability score is the above-mentioned fifth risk score.

[0117] S206: If the user information contains illegal content and the user's black market probability score is greater than a threshold, the number is determined to be a black market number.

[0118] The illegal content is the above-mentioned preset marked content, the threshold is the above-mentioned preset probability threshold, and the black market probability score is the probability that the above-mentioned number to be detected is the marked number. The specific calculation method refers to the relevant description in the above-mentioned embodiment, which will not be repeated here.

[0119] Using the method provided by the embodiment of the present application, when the number of associated numbers is greater than the first preset threshold, it means that the number opened by the user is more, and the more the number opened, the higher the possibility that the user is a black and gray production user. In the case where the distance between the number to be detected and the associated number is greater than the second preset threshold, it means that the distance between the associated number of the user and the number to be detected is far, and the user may use the associated number to perform illegal operations. In the case where the distance between the number to be detected and the preset mark position is less than the third preset threshold, it means that the possibility that the number to be detected is a black and gray production number is high. Then calculate the sum of the first risk score, the second risk score, the third risk score, the fourth risk score and the fifth risk score, and the probability that the number to be detected is a black and gray production number can be measured by combining multiple dimensions, and accordingly, the probability can be used as the probability that the corresponding user of the number to be detected is a black and gray production user. In this way, the probability that the user is a black and gray production user can be judged from multiple dimensions to improve the accuracy of the calculation result.

[0120] For implementation method 2: the location information, associated number information and communication information of the number to be detected are input into the trained user identification model, and the specific implementation process of using the user identification model to output the probability that the user corresponding to the number to be detected is the marked user is as follows:

[0121] The number to be detected, location information of the number to be detected, associated number information and communication information are input into the user identification model as input data, and the user identification model predicts a prediction result based on the input data.

[0122] The prediction result includes the probability that the number to be detected is a marked number. The user identification model is trained using sample-related information of the marked number and the marked number. The sample-related information includes sample location information and sample communication information of the marked number. The marked number is the number corresponding to the black and gray market user.

[0123] In some embodiments of the present application, based on the probability that the number to be detected is a marked number predicted by the user identification model, it is determined whether the user corresponding to the number to be detected is a marked user.

[0124] Accordingly, when the user identification model outputs that the number to be detected is a marked number, the message content of the sent message corresponding to the number to be detected is identified, and when the preset marked content is identified, the user corresponding to the number to be detected is determined to be a marked user.

[0125] In addition, after the marking number is predicted using the user identification model, the relevant information corresponding to the marking number is stored in the above-mentioned black and gray industry user information storage database, so as to facilitate the subsequent calibration of the user identification model using the data in the black and gray industry user information storage database.

[0126] It should be noted that before using the user identification model to identify the tag number, a neural network model is pre-trained to obtain the above user identification model.

[0127] Specifically, the method for training a user identification model includes the following steps:

[0128] Obtain sample-related information. Input sample-related information into the neural network model, and the neural network model outputs the sample prediction result. Use the label number as the label data and calculate the loss function of the label data and the sample prediction result. Optimize the parameters of the neural network model according to the loss function to obtain the user identification model.

[0129] It should be noted that in order to improve the accuracy of the user identification model in identifying black and gray market numbers, the location information of the associated numbers, the movement trajectory of the marked numbers and the sending behavior data can also be input into the neural network model, so that the neural network model can adjust the model weights and biases based on the above-mentioned features such as the location information of the associated numbers, the movement trajectory of the marked numbers and the sending behavior data.

[0130] By using the method provided in the embodiment of the present application, the above sample-related information is input into the neural network model, the neural network model can learn the characteristics of the marked number, and then use the marked number as label data, and calculate the loss function, so as to continuously optimize the parameters of the neural network model using the loss function to obtain the user identification model. Subsequently, the user identification model is used to classify the number to be detected, and the probability that the number to be detected is a marked user is output, thereby improving the recognition efficiency of the number to be detected and improving the accuracy of the prediction results.

[0131] In addition, the training data can be divided into a training sample data set and a verification sample data set, wherein the training sample data set includes the above sample related information, and the verification sample data set is used to evaluate the performance indicators of the user recognition model, such as the accuracy and recall rate of the model.

[0132] Specifically, the above-mentioned optimization of the parameters of the neural network model according to the loss function to obtain the user identification model can be specifically implemented as follows:

[0133] The parameters of the neural network model are optimized according to the loss function to obtain the target model.

[0134] The model performance of the target model is evaluated using the validation data set. When the recognition accuracy of the target model is greater than or equal to a preset accuracy threshold, the target model is used as the user recognition model.

[0135] The above-mentioned use of the validation data set to evaluate the model performance of the target model, when the recognition accuracy of the target model is greater than or equal to the preset accuracy threshold, uses the target model as the user recognition model, which can be performed according to the following steps:

[0136] Step A: Obtain verification related information.

[0137] The verification related information includes the verification location information and verification communication information of the user number to be verified.

[0138] Step B: input the verification related information into the target model to obtain the recognition result output by the target model.

[0139] Step C: verify the recognition result according to the tag data of the user number to be verified.

[0140] Specifically, the recognition result of the target model is the probability that the user number to be verified is a marked number. If the probability is greater than or equal to 50%, the user number to be verified is determined to be a marked number; if the probability is less than 50%, the user number to be verified is determined to be an unmarked number. The electronic device determines whether the user number to be verified is a marked number based on the recognition result output by the target model, thereby comparing the label data with the result determined by the electronic device based on the output of the target model to complete the verification of the recognition result.

[0141] Step D: Count the first target number of user numbers to be verified whose tag data is consistent with the identification result, and the second target number of user numbers to be verified.

[0142] Step E: Calculate the ratio of the first target quantity to the second target quantity to obtain the recognition accuracy.

[0143] Step F: When the recognition accuracy is greater than or equal to a preset accuracy threshold, the target model is used as a user recognition model.

[0144] By using the method provided in the embodiment of the present application, the verification-related information in the verification data set is input into the target model to obtain the recognition result output by the target model, and then the recognition result is verified using the label data in the verification data set. When the recognition accuracy of the target model calculated is greater than or equal to the preset accuracy threshold, the target model can be used as the user recognition model. In this way, the recognition accuracy of the target model is calculated using the verification data set, and the performance of the model is evaluated using the recognition accuracy, thereby ensuring the accuracy of the model output results.

[0145] Based on the above implementation method 2, combined with Figure 3 Introduce a system that uses user identification models to identify black and gray market users, such as Figure 3 As shown, the system includes a data preparation module, a data storage module, a model training module, a data preprocessing module and a data processing module.

[0146] After the number to be detected is input into the system, the data preparation module obtains all data of the number to be detected, wherein all data of the number to be detected includes movement trajectory data of the number to be detected, number data under the user name, IP data and user sending behavior data.

[0147] Specifically, the data preparation module responds to the input number to be detected, obtains the location information of the number to be detected through base station positioning, and obtains the movement trajectory of the number to be detected from the user location information storage database in the data storage module.

[0148] The data preparation module obtains the associated number information from the user number storage database in the data storage module, the associated number information including the number of associated numbers under the user name and the location information of the associated numbers;

[0149] The data preparation module obtains IP data from the IP data storage database in the data storage module, and the IP data is the above-mentioned preset marked online data. That is, the data preparation module obtains the preset marked online data from the IP data storage database in the data storage module;

[0150] The data preparation module obtains the user sending behavior data from the message data storage database in the data storage module, wherein the sending behavior data is the above-mentioned sending frequency, sending quantity, number of messages with the same content, and number of receiving numbers.

[0151] After receiving all the data of the number to be detected, the data preprocessing module performs data preprocessing on all the data of the number to be detected, wherein the data preprocessing includes data cleaning and feature extraction.

[0152] Specifically, feature extraction refers to extracting relevant information of the number to be detected required for subsequent model reasoning according to the user identification method provided in the embodiment of the present application. The specific relevant information is referred to the relevant description in the above embodiment and will not be repeated here.

[0153] The relevant information of the number to be detected obtained through data preprocessing is input into the above-mentioned user identification model for model inference, and the user identification model outputs the probability that the number to be detected is a marked number.

[0154] Before model inference, a neural network model is pre-trained to obtain a user recognition model.

[0155] Specifically, the model training module obtains sample-related information from the data storage module for data preparation, then performs word segmentation on the obtained data to obtain the word segmentation processing result, extracts features from the word segmentation processing result, and uses the extracted features to pre-train the neural network model. During the model training process, the loss function of the model is calculated using a pre-defined loss function, and the model is iteratively trained. After the user recognition model is trained, in order to improve the accuracy of the model prediction results, the model is evaluated and tuned using the verification sample data set to optimize the model parameters.

[0156] After the user identification model outputs the probability that the number to be detected is a marked number, the data processing module obtains the sent message content in the communication information, and then identifies the illegal content of the sent message content.

[0157] The data processing module includes a text recognition unit, an image recognition unit, an audio recognition unit and a video recognition unit. After determining the message type corresponding to the content of the sent message, the illegal content of the sent message is identified according to the preset correspondence between the message type and the identification method.

[0158] Specifically, when the probability output by the user identification model is greater than a preset probability threshold, the target type information corresponding to the communication information is determined, and then the target identification method corresponding to the target type information is obtained based on the correspondence between the type information and the identification method. The preset mark content in the communication information is identified according to the target identification method, so as to determine whether the preset mark content is included in the sent message.

[0159] The electronic device pre-sets the correspondence between the type information and the identification method. For example, the message type may be a text message, an audio message, and a video message, and the electronic device may respectively set the identification methods corresponding to the text message, the audio message, and the video message.

[0160] In this way, different identification methods are used to process the communication information for different types of information, so that the preset marking content in the communication information can be identified more accurately, ensuring the accuracy of the identification result.

[0161] Wherein, in the case where the communication information includes text message type information, picture message type information, and audio message type information, the above-mentioned identifying the preset marking content in the communication information according to the target recognition method can be implemented as follows:

[0162] In the case where the communication information is of text message type information, the text content in the communication information is matched according to the preset keywords.

[0163] The preset keywords are determined based on the preset tag content.

[0164] In the case where the communication information is of picture message type, the picture content in the communication information is obtained, image features are extracted from the picture content, target features are obtained, and the target features are searched in a tag feature library.

[0165] The mark feature library stores image features including preset mark content pictures.

[0166] In the case where the communication information is of an audio message type, speech-to-text processing is performed on the communication information to obtain text information, and the text information is matched according to preset keywords.

[0167] Specifically, for text message types, the electronic device performs keyword and semantic recognition on the content of the text message, where the keyword can be a word in the preset tag content; for picture message types, an optical character recognition (OCR) method is used to identify the picture content in the picture message, and then an image feature extraction algorithm is used to extract features in the picture content, and the extracted features are compared with features in a violation sample feature library to determine whether there is any violation content in the picture message; for audio message types, the audio message can be processed by voice-to-text conversion, and then the text content is processed using the recognition method corresponding to the text message type; for video message types, the electronic device extracts frames from the video in the video message, uses the extracted video frames as picture messages, and processes the video frames according to the processing method of the picture message type. The embodiment of the present application does not limit the method of video frame extraction.

[0168] If the probability that the number to be detected is a marked number output by the model is greater than the preset probability threshold, and the data processing module identifies that the sent message contains illegal content, the result is output, which includes the identification result of the number to be detected, the identification result of the user corresponding to the number to be detected, and the relevant information of the number to be detected. The output result is stored in the black industry number database, which is the above-mentioned black and gray industry user information storage database.

[0169] Based on the same concept, the embodiment of the present application also provides a user identification device, such as Figure 4 As shown, the device comprises:

[0170] An acquisition module 401 is used to acquire relevant information of the number to be detected, wherein the relevant information includes location information, associated number information and communication information of the number to be detected;

[0171] A calculation module 402, configured to calculate the probability that the number to be detected is a marked number according to the location information of the number to be detected, the associated number information and the communication information;

[0172] A judgment module 403 is used to judge whether the communication information includes preset marking content when the probability is greater than a preset probability threshold;

[0173] The determination module 404 is configured to determine the user corresponding to the to-be-detected number as a marked user when the communication information includes the preset marked content.

[0174] In a possible implementation, the associated number information includes the number of associated numbers associated with the number to be detected and location information of the associated numbers; the calculation module 402 is specifically configured to:

[0175] When the number of associated numbers is greater than a first preset threshold, determining a product of the number of associated numbers and a first preset probability as a first risk score;

[0176] Calculating a first distance between the number to be detected and the associated number according to the location information of the number to be detected and the location information of the associated number, and obtaining a second risk score when the first distance is greater than a second preset threshold;

[0177] Calculating a second distance between the position of the number to be detected and the preset mark position according to the position information of the number to be detected, and obtaining a third risk score when the second distance is less than a third preset threshold;

[0178] determining a communication risk score based on the communication information;

[0179] Calculating the sum of the first risk score, the second risk score, the third risk score and the communication risk score to obtain a sum value to be detected;

[0180] The sum to be detected is used as the probability that the number to be detected is a marked number.

[0181] In a possible implementation, the communication information includes Internet access data information; the calculation module 402 is specifically configured to:

[0182] In a case where the online data information and the preset marked online data information are the same, a fourth risk score is obtained, and the fourth risk score is used as the communication risk score.

[0183] In a possible implementation, the communication information further includes message content information; the calculation module 402 is specifically configured to:

[0184] According to the message content information, count the sending frequency and the sending quantity of the number to be detected, the number of receiving numbers that receive the message of the number to be detected, and the number of messages with the same content;

[0185] Respectively comparing the number of messages with a fourth preset threshold, the sending frequency with a fifth preset threshold, the sending quantity with a sixth preset threshold, the number of receiving numbers with the seventh preset threshold, counting the number of indicators greater than or equal to the corresponding preset thresholds among the number of messages, the sending frequency, the sending quantity, and the number of receiving numbers, and calculating the product of the number of indicators and a second preset probability to obtain a fifth risk score;

[0186] A sum of the fifth risk score and the fourth risk score is calculated to obtain the communication risk score.

[0187] In a possible implementation, the calculation module 402 is specifically configured to:

[0188] The number to be detected, the location information of the number to be detected, the associated number information and the communication information are input as input data into a user identification model, and the user identification model predicts a prediction result based on the input data;

[0189] Among them, the prediction result includes the probability that the number to be detected is a marked number, and the user recognition model is obtained by using sample related information of the marked number and the marked number for training, and the sample related information includes sample location information and sample communication information of the marked number.

[0190] In a possible implementation, the device further includes:

[0191] The acquisition module 401 is also used to acquire the sample related information;

[0192] A training module, used for inputting the sample-related information into a neural network model, and the neural network model outputs a sample prediction result;

[0193] The calculation module 402 is further used to use the marked number as label data and calculate the loss function of the label data and the sample prediction result;

[0194] An optimization module is used to optimize the parameters of the neural network model according to the loss function to obtain the user recognition model.

[0195] In a possible implementation, the optimization module is specifically used to:

[0196] Optimizing the parameters of the neural network model according to the loss function to obtain the target model;

[0197] The model performance of the target model is evaluated using a verification data set, and when the recognition accuracy of the target model is greater than or equal to a preset accuracy threshold, the target model is used as the user recognition model.

[0198] In a possible implementation, the optimization module is specifically used to:

[0199] Acquire verification related information, wherein the verification related information includes verification location information and verification communication information of the user number to be verified;

[0200] Inputting the verification related information into the target model to obtain a recognition result output by the target model;

[0201] Verifying the recognition result according to the tag data of the user number to be verified;

[0202] Counting the first target number of the tag data of the user number to be verified that is consistent with the recognition result, and the second target number of the user number to be verified;

[0203] Calculating a ratio of the first target quantity to the second target quantity to obtain a recognition accuracy rate;

[0204] When the recognition accuracy is greater than or equal to a preset accuracy threshold, the target model is used as the user recognition model.

[0205] In a possible implementation, the determination module 403 is specifically configured to:

[0206] Determining target type information corresponding to the communication information;

[0207] According to the correspondence between the type information and the identification method, obtaining the target identification method corresponding to the target type information;

[0208] The preset marking content in the communication information is identified according to the target identification method.

[0209] In a possible implementation, the communication information includes text message type information, picture message type information, and audio message type information; the determination module 403 is specifically configured to:

[0210] In the case where the communication information is the text message type information, matching the text content in the communication information according to the preset keywords;

[0211] In the case where the communication information is the picture message type information, obtaining picture content in the communication information, performing image feature extraction on the picture content, obtaining target features, and searching for the target features in a mark feature library;

[0212] In the case that the communication information is of the audio message type, speech-to-text processing is performed on the communication information to obtain text information, and the text information is matched according to the preset keywords.

[0213] It should be noted that the user identification device is a device corresponding to the above-mentioned user identification method, and all implementation methods in the above-mentioned method embodiment are applicable to the embodiment of the device and can achieve the same technical effect.

[0214] Figure 5 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application is shown.

[0215] The terminal device may include a processor 501 and a memory 502 storing computer program instructions.

[0216] Specifically, the processor 701 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0217] The memory 502 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 502 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 502 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 502 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 502 is a non-volatile solid-state memory.

[0218] In certain embodiments, the memory 502 may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk storage medium device, an optical storage medium device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Thus, in general, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) 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 with reference to the method according to an aspect of the present disclosure.

[0219] The processor 501 implements any one of the user identification methods in the above embodiments by reading and executing computer program instructions stored in the memory 502 .

[0220] In one example, the terminal device may further include a communication interface 503 and a bus 504. Figure 5As shown, the processor 501, the memory 502, and the communication interface 503 are connected via a bus 504 and communicate with each other.

[0221] The communication interface 503 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0222] Bus 504 includes hardware, software or both, and the parts of electronic equipment are coupled to each other.For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front side bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 504 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the application considers any suitable bus or interconnection.

[0223] In addition, in combination with the user identification method in the above embodiment, the embodiment of the present application can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when the computer program instructions are executed by a processor, any one of the user identification methods in the above embodiment is implemented.

[0224] In combination with the user identification method in the above embodiments, an embodiment of the present application may provide a computer program product, the computer product including a computer program, and when the computer program is executed by a processor, any one of the user identification methods in the above embodiments is implemented.

[0225] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0226] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in 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, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0227] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. 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 embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0228] Aspects of the present disclosure are described above with reference to the flowchart and / or block diagram of the method, device (system) and computer program product according to the embodiment of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram 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 device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the function / action specified in one or more boxes of the flowchart and / or block diagram. 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 can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0229] The above are only specific implementation methods of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.

Claims

1. A method for user identification, characterized in that: include: Acquire relevant information of the number to be detected, the relevant information including location information, associated number information and communication information of the number to be detected; Calculating the probability that the number to be detected is a marked number according to the location information of the number to be detected, the associated number information and the communication information; When the probability is greater than a preset probability threshold, determining whether the communication information includes preset marking content; In a case where the communication information includes the preset marking content, the user corresponding to the number to be detected is determined as a marked user.

2. The method according to claim 1, characterized in that The associated number information includes the number of associated numbers associated with the number to be detected and the location information of the associated numbers; the calculating the probability that the number to be detected is a marked number according to the location information of the number to be detected, the associated number information and the communication information includes: When the number of associated numbers is greater than a first preset threshold, determining a product of the number of associated numbers and a first preset probability as a first risk score; Calculating a first distance between the number to be detected and the associated number according to the location information of the number to be detected and the location information of the associated number, and obtaining a second risk score when the first distance is greater than a second preset threshold; Calculate, according to the position information of the number to be detected, a second distance between the position of the number to be detected and the preset mark position, and obtain a third risk score when the second distance is less than a third preset threshold; determining a communication risk score based on the communication information; Calculating the sum of the first risk score, the second risk score, the third risk score, and the communication risk score to obtain a sum value to be detected; The sum to be detected is used as the probability that the number to be detected is a marked number.

3. The method according to claim 2, characterized in that The communication information includes online data information; and determining the communication risk score according to the communication information includes: In a case where the online data information and the preset marked online data information are the same, a fourth risk score is obtained, and the fourth risk score is used as the communication risk score.

4. The method according to claim 3, characterized in that The communication information also includes message content information; and determining the communication risk score according to the communication information includes: According to the message content information, count the sending frequency and the sending quantity of the number to be detected, the number of receiving numbers that receive the message of the number to be detected, and the number of messages with the same content; Respectively comparing the number of messages with a fourth preset threshold, the sending frequency with a fifth preset threshold, the sending quantity with a sixth preset threshold, and the number of receiving numbers with a seventh preset threshold, counting the number of indicators greater than or equal to the corresponding preset thresholds among the number of messages, the sending frequency, the sending quantity, and the number of receiving numbers, and calculating the product of the number of indicators and a second preset probability to obtain a fifth risk score; A sum of the fifth risk score and the fourth risk score is calculated to obtain the communication risk score.

5. The method according to claim 1, characterized in that The calculating the probability that the number to be detected is a marked number according to the location information of the number to be detected, the associated number information and the communication information includes: The number to be detected, the location information of the number to be detected, the associated number information and the communication information are input into a user identification model as input data, and the user identification model predicts a prediction result based on the input data; Among them, the prediction result includes the probability that the number to be detected is a marked number, and the user recognition model is obtained by using sample related information of the marked number and the marked number for training, and the sample related information includes sample location information and sample communication information of the marked number.

6. The method according to claim 5, characterized in that Before the number to be detected, the location information of the number to be detected, the associated number information and the communication information are input as input data into a user identification model, and the user identification model predicts a prediction result based on the input data, the method further includes: Obtaining information related to the sample; Inputting the sample related information into a neural network model, and the neural network model outputs a sample prediction result; Using the marked number as label data, calculating the loss function of the label data and the sample prediction result; The parameters of the neural network model are optimized according to the loss function to obtain the user recognition model.

7. The method according to claim 6, characterized in that The step of optimizing the parameters of the neural network model according to the loss function to obtain the user identification model includes: Optimizing the parameters of the neural network model according to the loss function to obtain a target model; The model performance of the target model is evaluated using a verification data set, and when the recognition accuracy of the target model is greater than or equal to a preset accuracy threshold, the target model is used as the user recognition model.

8. The method according to claim 7, characterized in that The using the verification data set to evaluate the model performance of the target model, and when the recognition accuracy of the target model is greater than or equal to a preset accuracy threshold, using the target model as the user recognition model, includes: Acquire verification related information, the verification related information including verification location information and verification communication information of the user number to be verified; Inputting the verification related information into the target model to obtain a recognition result output by the target model; Verifying the recognition result according to the tag data of the user number to be verified; Counting the first target number of the tag data of the user number to be verified that is consistent with the recognition result, and the second target number of the user number to be verified; Calculating a ratio of the first target quantity to the second target quantity to obtain a recognition accuracy rate; When the recognition accuracy is greater than or equal to a preset accuracy threshold, the target model is used as the user recognition model.

9. The method according to claim 1, characterized in that: The determining whether the communication information includes preset marking content includes: Determining target type information corresponding to the communication information; According to the correspondence between the type information and the identification method, obtaining the target identification method corresponding to the target type information; The preset marking content in the communication information is identified according to the target identification method.

10. The method according to claim 9, characterized in that The communication information includes text message type information, picture message type information, and audio message type information; and identifying the preset mark content in the communication information according to the target recognition method includes: In the case where the communication information is the text message type information, matching the text content in the communication information according to the preset keywords; In the case where the communication information is the picture message type information, obtaining picture content in the communication information, performing image feature extraction on the picture content, obtaining target features, and searching for the target features in a mark feature library; In the case that the communication information is of the audio message type, speech-to-text processing is performed on the communication information to obtain text information, and the text information is matched according to the preset keywords.

Images