Call restriction method, device, system, electronic device and storage medium

Abstract

The present application discloses a call restriction method, device, system, electronic device and storage medium, wherein the method comprises: when establishing a call connection with a suspected abnormal number, playing a first voice message, wherein the first voice message is used to inquire about the historical call intention of the suspected abnormal number; when obtaining a second voice message of the suspected abnormal number, determining the abnormal type information of the suspected abnormal number according to the second voice message; determining the target call connection rate of the suspected abnormal number according to the abnormal type information of the suspected abnormal number; and restricting calls to the suspected abnormal number so that the call connection rate of the suspected abnormal number does not exceed the target call connection rate. According to the embodiments of the present application, the present application can improve the timeliness of call restriction on abnormal numbers.

Description

Technical Field

[0001] The present application belongs to the field of communication technology, and in particular, relates to a call restriction method, device, system, electronic device and storage medium. Background Art

[0002] With the development of communication technology, people are receiving unknown calls more and more frequently. Among these unknown calls, there are many abnormal calls for the purpose of fraud, sales, harassment, etc., which bring great interference to people's lives. At present, operators can obtain suspected abnormal numbers based on user reporting information, and analyze the calling behavior of suspected abnormal numbers according to existing strategies. If the suspected abnormal number is determined to be an abnormal number, the abnormal number will be stored in the Internet tag library to restrict calls to the abnormal number. However, the cycle from users receiving calls, to reporting suspected abnormal numbers, to operators analyzing suspected abnormal numbers, and finally to restricting calls is long, resulting in poor timeliness of call restrictions on abnormal numbers. Summary of the invention

[0003] Embodiments of the present application provide a call restriction method, device, system, electronic device, and storage medium, which can improve the timeliness of call restriction for abnormal numbers.

[0004] In a first aspect, an embodiment of the present application provides a call restriction method, including:

[0005] When a call connection is established with a suspected abnormal number, a first voice message is played, where the first voice message is used to inquire about the historical call intention of the suspected abnormal number;

[0006] In the case where the second voice information of the suspected abnormal number is obtained, determining the abnormal type information of the suspected abnormal number according to the second voice information;

[0007] Determining a target call connection rate of the suspected abnormal number according to the abnormal type information of the suspected abnormal number;

[0008] Call restriction is performed on the suspected abnormal number so that the call connection rate of the suspected abnormal number does not exceed the target call connection rate.

[0009] In a second aspect, an embodiment of the present application provides a call restriction device, including:

[0010] A playing module, configured to play a first voice message when a call connection is established with a suspected abnormal number, wherein the first voice message is used to inquire about the historical call intention of the suspected abnormal number;

[0011] A first determining module is used to determine the abnormal type information of the suspected abnormal number according to the second voice information when the second voice information of the suspected abnormal number is obtained;

[0012] A second determination module is used to determine a target call connection rate of the suspected abnormal number according to the abnormal type information of the suspected abnormal number;

[0013] The restriction module is used to restrict calls to the suspected abnormal number so that the call connection rate of the suspected abnormal number does not exceed the target call connection rate.

[0014] In a third aspect, an embodiment of the present application provides a call restriction system, including an outbound call management device, a media bridging device, a voice control device, and a call control device;

[0015] The outbound call management device is used to: establish a call connection with a suspected abnormal number;

[0016] The media bridging device is used to: play a first voice message, obtain a second voice message of the suspected abnormal number, and forward the second voice message to the voice control device, wherein the first voice message is used to inquire about the historical call intention of the suspected abnormal number;

[0017] The voice control device is used to: determine the abnormal type information of the suspected abnormal number according to the second voice information, and send the abnormal type information of the suspected abnormal number to the call control device;

[0018] The call control device is used to control the call connection rate of the suspected abnormal number according to the abnormal type information of the suspected abnormal number.

[0019] In a fourth aspect, an embodiment of the present application provides an electronic device, comprising: a processor, and a memory storing computer program instructions;

[0020] The processor reads and executes the computer program instructions to implement the call restriction method described in the first aspect.

[0021] In a fifth aspect, an embodiment of the present application provides a computer storage medium on which computer program instructions are stored. When the computer program instructions are executed by a processor, the call restriction method as described in the first aspect is implemented.

[0022] The call restriction method, device, system, electronic device and storage medium of the embodiment of the present application establish a call connection with the suspected abnormal number and perform voice interaction with the suspected abnormal number to obtain the historical call intention of the suspected abnormal number, thereby determining the abnormal type of the suspected abnormal number, and then restricting the calls to the suspected abnormal number according to the abnormal type of the suspected abnormal number. In this way, on the one hand, the attributes of the suspected abnormal number can be determined in a relatively short time, and the abnormal number can be restricted immediately, which can improve the timeliness of the call restriction on the abnormal number; on the other hand, different strategies can be used to restrict calls to abnormal numbers of different abnormal types, thereby improving the flexibility and adaptability of call restriction on abnormal numbers. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] 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.

[0024] Figure 1 is a schematic diagram of a call restriction system provided by some embodiments of the present application;

[0025] Figure 2 is a flowchart of a call restriction method provided in some embodiments of the present application;

[0026] Figure 3 is a flowchart of a call restriction method provided in some other embodiments of the present application;

[0027] Figure 4 is a schematic diagram of the structure of a call restriction device provided in some embodiments of the present application;

[0028] Figure 5 It is a schematic diagram of the hardware structure of the call restriction device provided in some embodiments of the present application. DETAILED DESCRIPTION

[0029] 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 configured to explain the present application and are not configured 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.

[0030] 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 presence of other identical elements in the process, method, article or device including the elements.

[0031] In order to solve the problems in the prior art, embodiments of the present application provide a call restriction method, device, system, electronic device, and computer storage medium.

[0032] The following first introduces the call restriction system provided in the embodiment of the present application.

[0033] Figure 1 A schematic diagram of a call restriction system 100 provided in an embodiment of the present application is shown.

[0034] like Figure 1 As shown, the call restriction system 100 includes an outbound call management device 101, a media bridge device 102, a voice control device 103 and a call control device 104;

[0035] The outbound call management device 101 is used to establish a call connection with a suspected abnormal number;

[0036] The media bridge device 102 is used to play the first voice information, obtain the second voice information of the suspected abnormal number, and forward the second voice information to the voice control device, wherein the first voice information is used to inquire about the historical call intention of the suspected abnormal number;

[0037] The voice control device 103 is used to determine the abnormal type information of the suspected abnormal number according to the second voice information, and send the abnormal type information of the suspected abnormal number to the call control device;

[0038] The call control device 104 is used to control the call connection rate of the suspected abnormal number according to the abnormal type information of the suspected abnormal number.

[0039] The call restriction system 100 provided in the embodiment of the present application can establish a communication connection with the operator communication network A. Specifically, after obtaining the suspected abnormal number, the outbound call management device 101 can establish an outbound call signaling with the operator communication network A, thereby establishing a call connection with the suspected abnormal number. The media bridging device 102 can be connected to the media gateway of the operator communication network A, and after the outbound call management device 101 establishes a call connection with the suspected abnormal number, a voice message for inquiring about the historical call intention of the suspected abnormal number is played to the suspected abnormal number, and the voice message of the suspected abnormal number is received in real time.

[0040] The media bridge device 102 may also forward the received voice information from the suspected abnormal number to the voice control device 103. The voice control device 103 may analyze the received voice information, determine the abnormal type information of the suspected abnormal number, and send the determination result to the call control device 104.

[0041] The voice interaction between the call restriction system 100 and the suspected abnormal number may be one round or multiple rounds. During the multiple rounds of dialogue, the voice control device 103 may generate the next round of voice interaction information based on the analysis results of the voice information. The voice control device 103 may also use the pre-trained meta-information extraction model to determine whether the other party has finished speaking, and control the media bridge device 102 to play the next round of voice information after the other party has finished speaking.

[0042] The call control device 104 can predict the number of calls of the suspected abnormal number in the next time period based on the type information of the suspected abnormal number, and determine the target call connection rate of the suspected abnormal number based on the preset maximum call volume of the corresponding abnormal type in the next time period. In addition, the call control device 104 can restrict calls to the suspected abnormal number based on the abnormal type information of the abnormal number. Call restriction can be achieved by limiting the call connection rate of the abnormal number to not exceed the target call connection rate.

[0043] The call restriction system provided in the embodiment of the present application establishes a call connection with a suspected abnormal number and performs voice interaction with the suspected abnormal number to obtain the historical call intention of the suspected abnormal number, thereby determining the abnormal type of the suspected abnormal number, and then restricting the calls to the suspected abnormal number according to the abnormal type of the suspected abnormal number. In this way, on the one hand, the attributes of the suspected abnormal number can be determined in a relatively short time, and the abnormal number can be restricted immediately, which can improve the timeliness of the call restriction on the abnormal number; on the other hand, different strategies can be used to restrict calls to abnormal numbers of different abnormal types, thereby improving the flexibility and adaptability of the call restriction on abnormal numbers.

[0044] The call restriction system provided in the embodiment of the present application can be used to execute the call restriction method of the present application. The call restriction method of the present application is introduced below.

[0045] Figure 2 A flow chart of a call restriction method provided in an embodiment of the present application is shown.

[0046] like Figure 2 As shown, the method may include the following steps:

[0047] Step 201: When a call connection is established with a suspected abnormal number, a first voice message is played, where the first voice message is used to inquire about the historical call intention of the suspected abnormal number;

[0048] Step 202: when the second voice information of the suspected abnormal number is obtained, determining the abnormal type information of the suspected abnormal number according to the second voice information;

[0049] Step 203: determining a target call connection rate of the suspected abnormal number according to the abnormal type information of the suspected abnormal number;

[0050] Step 204: performing call restriction on the suspected abnormal number so that the call connection rate of the suspected abnormal number does not exceed the target call connection rate.

[0051] The suspected abnormal numbers involved in this application may be numbers with the following characteristics:

[0052] Suspected abnormal numbers call multiple numbers in a unit of time with a high frequency and are concentrated in a certain time period; the called numbers are scattered and the correlation between the called numbers is small; the suspected abnormal numbers and the called numbers are usually weakly correlated, that is, there are few historical call relationships, and the number of calls initiated by the suspected abnormal number as a caller is usually much larger than the number of calls made by the suspected abnormal number as a called party; the call duration of calls made by the suspected abnormal number is usually short, and the probability of the called party hanging up is high; in addition, suspected abnormal numbers can also include numbers reported by users.

[0053] In order to better understand the call restriction method provided by this application, Figure 1 The call restriction system of the present application introduces the call restriction method of the embodiment of the present application:

[0054] In step 201, the outbound call management device may initiate a call instruction after obtaining the suspected abnormal number and establish a call connection with the suspected abnormal number.

[0055] The first voice information is generated by the voice control device according to the conversation scenario and the voice information from the media bridge device. The voice control device can judge the voice information and the conversation scenario through natural language processing. The voice control device controls the media bridge device to play the first voice information. The first voice information includes information for inquiring about the historical call intention of the suspected abnormal number, and the historical call intention of the suspected abnormal number may include sales promotion, advertising, fraud, harassment, etc.

[0056] In some embodiments, the outbound call management device obtains a suspected abnormal number and initiates a call instruction to the suspected abnormal number, and establishes a call connection with the suspected abnormal number. After establishing a call connection with the suspected abnormal number, the voice control device will control the media bridging device to play a voice message and conduct voice interaction with the suspected abnormal number. The voice interaction may be one round or multiple rounds depending on the conversation scenario. The voice control device can control the timing of playing the first voice message. The first voice message can be played at the beginning of the call, or it can be played at an appropriate time according to the scenario of the voice interaction. This application does not impose any restrictions on this. For example, before sending the first voice message, the voice control device controls the media bridging device to play a voice message to inquire whether the suspected abnormal number has made a call during the historical call time, or to inquire about the identity of the historical caller, etc. After one or more rounds of conversation, the first voice message is played according to the interaction scenario.

[0057] In step 202, the second voice message is sent by the suspected abnormal number, which is the response information of the suspected abnormal number to the first voice message played by the media bridge device, and may include the historical call intention information of the suspected abnormal number. For example: voice messages with keywords or meanings such as telephone sales, advertising, harassment, etc. In addition, the second voice message may also be a voice message that has no response for a long time, is chaotic, disordered, or meaningless. According to the response of the suspected abnormal number, the voice control device can control the media bridge device to send the first voice message again or multiple times.

[0058] The voice control device performs semantic recognition on the second voice information, and determines the abnormal type of the abnormal number according to the information included in the recognition result.

[0059] Semantic analysis is divided into word-level speech analysis and sentence-level speech analysis. The basic task of word-level semantic analysis is to perform semantic disambiguation and determine the specific meaning of polysemous words in the context. The basic task of sentence-level speech analysis is semantic role labeling, which analyzes the structure of sentences as units, studies the relationship between each component and the predicate in the sentence with the predicate as the center, and describes their relationship with semantic roles. Semantic analysis can identify the structure of sentences and the meaning of each word in the sentence to determine the meaning of the text contained in the speech information, and convert natural language that humans can understand into language that computers can understand.

[0060] In the embodiment of the present application, the voice control device can determine the type information of the suspected abnormal number through the meaning of the words and sentences obtained by semantic analysis. For example: the voice control device can determine the type of the suspected abnormal number and mark it according to the keyword information such as "insurance", "house purchase", "car purchase", "card application" obtained by semantic recognition; the voice control device can also determine the type of abnormal number and mark it according to the meaning information of sentences such as "Do you want to travel abroad in the near future?", "Does your child need to enroll in a cram school?", etc.

[0061] The voice control device determines the abnormal type of the suspected abnormal number based on the semantic analysis result of the voice information of the suspected abnormal number, thereby determining that the suspected abnormal number is an abnormal number. The abnormal type of the suspected abnormal number may include sales promotion, advertising, fraud, harassment and other types.

[0062] In the embodiment of the present application, during the voice interaction with the suspected abnormal number, the media bridge device obtains the second voice information from the suspected abnormal number and sends it to the voice control device. The voice control device performs semantic analysis on the received voice information, and determines the abnormal type of the suspected abnormal number according to the semantic content of the second voice information obtained by the semantic analysis. The voice control device can also determine the abnormal type of the abnormal number according to the situation that the call connection is established but there is no response for a long time and chaotic, disordered or meaningless voice.

[0063] In step 203, the voice control device may send abnormal number type information of the suspected abnormal number to the call restriction device, and the call restriction device may determine a target call connection rate of the suspected abnormal number according to a preset rule.

[0064] The call connection rate refers to the percentage of the number of calls connected by the telephone exchange to the total number of calls made by the user. The target call connection rate of the suspected abnormal number is set by the call control device according to the abnormal type information of the suspected abnormal number.

[0065] The target call connection rate of suspected abnormal numbers can be set according to the abnormal type of the abnormal number; it can also be set according to the comparison of the historical call frequency of the abnormal number in its abnormal type with other numbers of the same abnormal type; it can also be set according to the proportion of the abnormal number type in all abnormal number types and other conditions, and this application does not impose any restrictions on this. The target call connection rate of different types of abnormal numbers can be managed by managing the number of call connections in the next time period. For example: for telephone sales type numbers, the number of call connections in the next time period can be set to 2, the number of call connections for advertising type numbers in the next time period can be set to 1, and the number of call connections for harassment numbers in the next time period can be set to 0, etc. The number of call connections can be set according to the actual business situation of the abnormal number, and this application does not impose any restrictions on this.

[0066] In step 204, the call control device may restrict calls to the suspected abnormal number according to the target call connection rate corresponding to the abnormal type of the suspected abnormal number, so that the call connection rate of the suspected abnormal number does not exceed the target call connection rate.

[0067] The call restriction method in the embodiment of the present application establishes a call connection with the suspected abnormal number and performs voice interaction with the suspected abnormal number to obtain the historical call intention of the suspected abnormal number, thereby determining the abnormal type of the suspected abnormal number, and then restricting the calls to the suspected abnormal number according to the abnormal type of the suspected abnormal number. In this way, on the one hand, the attributes of the suspected abnormal number can be determined in a relatively short time, and the abnormal number can be restricted immediately, which can improve the timeliness of the call restriction on the abnormal number; on the other hand, different strategies can be used to restrict calls to abnormal numbers of different abnormal types, thereby improving the flexibility and adaptability of the call restriction on abnormal numbers.

[0068] Optionally, after establishing a call connection with the suspected abnormal number and before playing the first voice message, the method further includes:

[0069] Converting the acquired voice information of the suspected abnormal number into a first frequency spectrum;

[0070] Using a pre-trained meta-information extraction model, identifying a meta-information tag of the first spectrum;

[0071] The first voice information is determined according to the meta information tag.

[0072] In this implementation, the media bridge device receives voice information in the current conversation with the suspected abnormal number and sends the voice information to the voice control device. The voice control device converts the voice information from the media bridge device into a first spectrum. The first spectrum can be, for example, a Log domain spectrum. The first spectrum can be used as input data of a pre-trained meta-information extraction model. The meta-information extraction model is used to extract meta-information tags of the voice information.

[0073] The voice control device can determine the first voice information content that matches the current conversation scenario based on the meta-information tag value. In addition, the voice control device can also select the voice information that matches the current conversation scenario according to a pre-set rule based on the meta-information tag and the current number of voice interaction rounds, and send the voice information to the media bridge device. After receiving the voice information from the voice control device, the media bridge device can play it to the suspected abnormal number, realize continuous voice interaction, and enhance the personalization of voice interaction.

[0074] Before determining the first voice information, the voice control device will pre-generate and train a meta-information extraction model. The main function of the meta-information extraction model is to extract meta-information tags of voice information. The meta-information tags may include information such as the tone, emotion, sentence structure, etc. of the voice information in the voice sample, and may also include information such as the gender, region, age, etc. of the speaker.

[0075] The generation process of the meta-information extraction model is as follows:

[0076] First, based on the human hearing and brain recognition mechanism, an audio signal feature extraction algorithm is designed, and then the audio signal features and tone and emotion labels are modeled based on the long short-term memory network (LSTM). The detailed process is as follows:

[0077] First, the Log domain spectrum converted from the speech information is pre-emphasized: the spectrum is filtered using the following algorithm, where x is the speech sample spectrum.

[0078] H(x)=1-μx -1 (1)

[0079] Then the filtered spectrum is framed and windowed: 512 sampling points are taken every 30ms for windowing. The window length is determined by the following algorithm, where n is the nth sampling interval and L is the frame length.

[0080]

[0081] Perform Fourier transform on the energy distribution on the spectrum. The algorithm is as follows, where N is the number of Fourier transform points, and x(n) is the audio signal in the nth window.

[0082]

[0083] The Mel filter bank is used to map the audio energy information to the Mel scale to conform to the human hearing habits. The mapping formula is as follows, where f is the frequency.

[0084]

[0085] The filtered audio signal is decorrelated and dimensionally reduced using discrete cosine transform to generate audio signal features that can be used for modeling. The algorithm is:

[0086]

[0087] C(n) is the nth order eigenvector. The value of n ranges from 1 to 12.

[0088] The LSTM model based on the recurrent neural network (RNN) uses the processed audio signal features for modeling, where:

[0089] RNN node model:

[0090] h i =σ(W xh x i +W hh h t-1 +b h ) (6)

[0091] LSTM node model:

[0092] c t =f t c t-1 +i t tanh(W wc x t +W hc h t-1 +b c ) (7)

[0093] Connection layer model of RNN and LSTM:

[0094] h t =o t tanh(c t ) (8)

[0095] The audio feature vector and labels such as tone and emotion are then input into the model established by the above process and trained to obtain a prediction model.

[0096] Finally, the above algorithm steps and prediction models are encapsulated as a meta-information extraction module.

[0097] In the process of establishing the meta-information extraction model, in order to weaken the specific semantic content features and enable the meta-information extraction model to realize the judgment of emotion and tone independently of the specific semantic content, the encoder-decoder model is used to optimize the meta-information extraction model.

[0098] First, a general speech recognition algorithm is used to convert each speech sample into a text paragraph and perform word segmentation to generate semantic features such as word vectors.

[0099] Secondly, define the semantic content function as f(x), the meta-information extraction function as g(x), and construct the reconstruction functions F and G corresponding to f and g. Assuming the overall reconstruction function S, the following model can be established based on extraction and reconstruction:

[0100] f(x)=c (9)

[0101] g(x)=m (10)

[0102] S(F(f(x)),G(g(x)))=x (11)

[0103] Then, the above model is trained using a deep neural network, and after parameter fitting, g is obtained as the optimized meta-information extraction function. The model module corresponding to this function can recognize the meta-information of speech samples while excluding the influence of semantic content as much as possible.

[0104] In an embodiment of the present application, the voice control device extracts meta-information tags through a pre-trained meta-information extraction model, and selects appropriate conversation voice information and first voice information according to the meta-information tags and the conversation scenario. The voice control device uses the meta-information extraction model to weaken the semantic content features, and can judge emotions and tones independently of specific semantic content, thereby improving the interactivity with suspected abnormal numbers. The call restriction system enhances the personalization of voice interaction information by obtaining meta-information tags of voice information, which is conducive to obtaining more expected voice information and can better cope with various application scenarios.

[0105] Optionally, determining a target call connection rate of the suspected abnormal number according to the abnormal type information of the suspected abnormal number includes:

[0106] Using a pre-trained call count prediction model, predict the number of calls made by the suspected abnormal number in the next time period;

[0107] The target call connection rate of the suspected abnormal number in the next time period is determined according to the predicted number of calls and the preset maximum call volume of the abnormal type number to which the suspected abnormal number belongs in the next time period.

[0108] In this embodiment, after the voice control device determines the abnormal type of the suspected abnormal number, the call control device will use a pre-trained call count prediction model to predict the number of calls of the suspected abnormal number in the next time period, and determine the target call connection rate of the suspected abnormal number based on the maximum call volume pre-set for the abnormal type.

[0109] The call restriction device predicts the number of calls from the suspected abnormal number in the next time period, and the length of the next time period can be flexibly set or adjusted as needed.

[0110] The maximum call volume of a suspected abnormal number of a certain abnormal type in the next period refers to the maximum number of calls connected by the suspected abnormal number in the next period. The target call connection rate is the ratio of the maximum call volume allowed by the suspected abnormal number type in the next period to the predicted number of calls.

[0111] In this implementation, the call restriction device can adaptively adjust the target call connection rate of abnormal numbers according to pre-set rules for real-time application scenarios, thereby enhancing the flexibility and adaptive capability of call restriction.

[0112] Optionally, the call frequency prediction model is trained by the following steps: obtaining the number of complaints of various abnormal numbers in the first preset period and the call characteristic indicators of the second preset period, generating a characteristic matrix of various suspected abnormal numbers, wherein the call characteristic indicators include at least one of call frequency, discreteness, call interval, connection rate, and call duration;

[0113] Based on the LASSO regression model and the feature matrix of each type of abnormal number, the call number prediction model is trained.

[0114] As an example, for each type of suspected abnormal number, the number of complaints received the previous day and the call frequency, dispersion, average call interval, connection rate, average call duration and other indicators of the previous hour are obtained and calculated to generate a number feature matrix. The call count prediction model is obtained by training based on the LASSO regression model and the abnormal number feature matrix.

[0115] The LASSO (Least absolute shrinkage and selection operator) method is a compression estimation method based on the idea of ​​reducing the variable set. It constructs a penalty function to compress the coefficients of the variables and make some regression coefficients zero, thereby achieving the purpose of variable selection.

[0116] Optionally, the call number prediction model is trained based on the LASSO regression model and the feature matrix of each type of abnormal number, including:

[0117] Configure the following cost function:

[0118]

[0119] Among them, M is the number of various abnormal numbers, p is the number of call feature indicators, and x ij is the feature matrix, w j is the characteristic coefficient;

[0120] The calling frequencies of various abnormal numbers in the second preset time period are used as dependent variables, and the calling characteristic indicators other than the calling frequencies are used as independent variables for training;

[0121] The model when the cost function takes the lowest value is used as the call number prediction model.

[0122] In this implementation, based on the LASSO regression model, the feature consistency of the abnormal numbers is comprehensively considered and the cost function is designed. For each type of abnormality, the call frequency in the previous period is used as the dependent variable, and other feature indicators are used as independent variables to train the model, and the model with the lowest cost function is taken as the optimal model. Finally, the call frequency prediction model of each type of abnormal number is obtained.

[0123] In order to better understand the call restriction method according to the embodiment of the present application, a specific example of the call restriction method is provided below.

[0124] like Figure 3 As shown, the call restriction method includes:

[0125] Step 301: the outbound call management device obtains a suspected abnormal number and establishes a call connection with the suspected abnormal number;

[0126] Step 302: The voice control device instructs the media bridge device to play a voice message to inquire whether the suspected abnormal number has made a historical call and inquire about its historical call intention;

[0127] Step 303: The operator communication network sends the voice information of the suspected abnormal number to the media bridge device;

[0128] Step 304: the media bridge device sends the received voice information to the voice control device;

[0129] Step 305: the voice control device converts the voice information into a first spectrum, and uses a pre-trained meta-information extraction model to identify a meta-information tag of the first spectrum;

[0130] Step 306: The voice control device selects a voice file that matches the current dialogue scenario according to the meta-information tag and the current number of speech rounds, and sends the voice file to the media bridge device;

[0131] Step 307: After receiving the voice file, the media bridge device plays the voice information;

[0132] Step 308: After collecting the historical call intentions of the suspected abnormal number, the voice control device determines the abnormal type information of the suspected abnormal number according to the historical call intentions of the suspected abnormal number;

[0133] Step 309: The call control device determines a target call connection rate of the suspected abnormal number according to the abnormal type of the suspected abnormal number, and performs call restriction on the suspected abnormal number so that the call connection rate of the suspected abnormal number does not exceed the target call connection rate.

[0134] In summary, the call restriction method in the embodiment of the present application establishes a call connection with the suspected abnormal number and performs voice interaction with the suspected abnormal number to obtain the historical call intention of the suspected abnormal number, thereby determining the abnormal type of the suspected abnormal number, and then restricting the call to the suspected abnormal number according to the abnormal type of the suspected abnormal number. In this way, on the one hand, the attributes of the suspected abnormal number can be determined in a relatively short time, and the call restriction of the abnormal number can be performed immediately, which can improve the timeliness of the call restriction on the abnormal number; on the other hand, different strategies can be used to restrict calls to abnormal numbers of different abnormal types, thereby improving the flexibility and adaptability of call restriction on abnormal numbers.

[0135] Figure 4 Schematic diagram of the structure of the call restriction device provided in the embodiment of the present application. Figure 4 As shown, the call restriction device 400 includes:

[0136] A playing module 401 is used to play a first voice message when a call connection is established with a suspected abnormal number, wherein the first voice message is used to inquire about the historical call intention of the suspected abnormal number;

[0137] A first determining module 402 is configured to determine abnormal type information of the suspected abnormal number according to the second voice information when the second voice information of the suspected abnormal number is obtained;

[0138] The second determination module 403 is used to determine the target call connection rate of the suspected abnormal number according to the abnormal type information of the suspected abnormal number;

[0139] The restriction module 404 is used to restrict calls to the suspected abnormal number so that the call connection rate of the suspected abnormal number does not exceed the target call connection rate.

[0140] Optionally, the call restriction device 400 further includes:

[0141] A conversion module, used for converting the acquired voice information of the suspected abnormal number into a first frequency spectrum;

[0142] an identification module, configured to identify the meta-information tag of the first spectrum using a pre-trained meta-information extraction model;

[0143] The third determining module is used to determine the first voice information according to the meta information tag.

[0144] Optionally, the second determining module 403 includes:

[0145] A prediction unit, configured to use a pre-trained call count prediction model to predict the number of calls made by the suspected abnormal number in the next time period;

[0146] The determination unit is used to determine the target call connection rate of the suspected abnormal number in the next time period according to the predicted number of calls and the preset maximum call volume of the number of the abnormal type to which the suspected abnormal number belongs in the next time period.

[0147] Optionally, the call restriction device 400 further includes a training module for training the call number prediction model, and the training module includes:

[0148] An acquisition unit, used to acquire the number of complaints received by various abnormal numbers in a first preset period and call characteristic indicators in a second preset period, and generate a characteristic matrix of various suspected abnormal numbers, wherein the call characteristic indicators include at least one of call frequency, discreteness, call interval, connection rate, and call duration;

[0149] The training unit is used to train the call number prediction model based on the LASSO regression model and the feature matrix of each type of abnormal number.

[0150] Optionally, the training unit is specifically used for:

[0151] Configure the following cost function:

[0152]

[0153] Among them, M is the number of various abnormal numbers, p is the number of call feature indicators, and x ij is the feature matrix, w j is the characteristic coefficient;

[0154] The calling frequencies of various abnormal numbers in the second preset time period are used as dependent variables, and the calling characteristic indicators other than the calling frequencies are used as independent variables for training;

[0155] The model when the cost function takes the lowest value is used as the call number prediction model.

[0156] Figure 4 Each module / unit in the device shown has the function of realizing Figures 2 to 3 The functions of each step in the process can achieve the corresponding technical effects, which will not be described in detail here for the sake of brevity.

[0157] Figure 5 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application is shown.

[0158] like Figure 5 As shown, the electronic device may include a processor 501 and a memory 502 storing computer program instructions.

[0159] Specifically, the processor 501 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.

[0160] Specifically, the processor 501 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.

[0161] 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 one example, the memory 302 may include a removable or non-removable (or fixed) medium, or the memory 502 is a non-volatile solid-state memory. The memory 502 may be inside or outside the integrated gateway disaster recovery device.

[0162] In one example, the memory 502 may be a read-only memory (ROM). In one example, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or a flash memory, or a combination of two or more of these.

[0163] 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. Therefore, generally, 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 call restriction method according to the embodiment of the present application.

[0164] The processor 501 reads and executes the computer program instructions stored in the memory 502 to implement any one of the call restriction methods in the above embodiments, and achieves Figure 2 The corresponding technical effects achieved by executing the methods / steps in the example shown are not repeated here for the sake of brevity.

[0165] In one example, the electronic device may further include a communication interface 503 and a bus 510. Figure 5 As shown, the processor 501, the memory 502, and the communication interface 503 are connected via a bus 510 and communicate with each other.

[0166] 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.

[0167] Bus 510 includes hardware, software or both, and the components of online data flow billing equipment are coupled to each other. For example, but not limitation, bus may include accelerated graphics port (Accelerated Graphics Port, AGP) or other graphics bus, enhanced industry standard architecture (Extended Industry Standard Architecture, EISA) bus, front side bus (Front Side Bus, FSB), Hyper Transport (Hyper Transport, HT) interconnection, industry standard architecture (Industry 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 510 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the present application considers any suitable bus or interconnection.

[0168] It should be noted that the electronic devices in the embodiments of the present application include the mobile electronic devices and non-mobile electronic devices mentioned above.

[0169] The electronic device can execute the call restriction method in the embodiment of the present application, thereby realizing the combination Figures 2 to 3 Describes the call restriction method.

[0170] In addition, in combination with the call restriction 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 call restriction methods in the above embodiment is implemented.

[0171] 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.

[0172] The functional blocks shown in the above-described 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 (Application Specific Integrated Circuit, ASIC), appropriate firmware, plug-in, function card, etc. When implemented in software, the elements of the present application are programs or code segments 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 communication link by a data signal carried in a carrier. "Machine-readable medium" may 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 (Radio Frequency, RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0173] 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.

[0174] 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.

[0175] The above is only a specific implementation 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 call restriction method, characterized in that: include: When a call connection is established with a suspected abnormal number, a first voice message is played, where the first voice message is used to inquire about the historical call intention of the suspected abnormal number; In the case where the second voice information of the suspected abnormal number is obtained, determining the abnormal type information of the suspected abnormal number according to the second voice information; Determining a target call connection rate of the suspected abnormal number according to the abnormal type information of the suspected abnormal number; When the target call connection rate exceeds the target call connection rate, restricting calls to the suspected abnormal number so that the call connection rate of the suspected abnormal number does not exceed the target call connection rate; The step of determining a target call connection rate of the suspected abnormal number according to the abnormal type information of the suspected abnormal number includes: Using a pre-trained call count prediction model, predict the number of calls made by the suspected abnormal number in the next time period; Determine a target call connection rate of the suspected abnormal number in the next time period according to the predicted number of calls and a preset maximum call volume of the number of the abnormal type to which the suspected abnormal number belongs in the next time period; The call number prediction model is trained by the following steps: Obtain the number of complaints received by various abnormal numbers in the first preset period and the call characteristic indicators in the second preset period, and generate a characteristic matrix of various abnormal numbers, wherein the call characteristic indicators include at least one of call frequency, dispersion, call interval, connection rate, and call duration; Based on the LASSO regression model and the feature matrix of each type of abnormal number, the call number prediction model is trained; The call number prediction model is trained based on the LASSO regression model and the feature matrix of each type of abnormal number, including: Configure the following cost function: Among them, M is the number of various abnormal numbers, p is the number of call feature indicators, and x ij is the feature matrix representing the jth eigenvalue of the ith abnormal number, w j is the jth characteristic coefficient, y i is the actual number of calls to the ith abnormal number, is the predicted number of calls for the ith abnormal number; The calling frequencies of various abnormal numbers in the second preset time period are used as dependent variables, and the calling characteristic indicators other than the calling frequencies are used as independent variables for training; The model when the cost function takes the lowest value is used as the call number prediction model.

2. The method according to claim 1, characterized in that After establishing a call connection with the suspected abnormal number and before playing the first voice message, the method further includes: Converting the acquired voice information of the suspected abnormal number into a first frequency spectrum; Using a pre-trained meta-information extraction model, identifying a meta-information tag of the first spectrum; The first voice information is determined according to the meta information tag.

3. A call restriction device, characterized in that: include: A playing module, configured to play a first voice message when a call connection is established with a suspected abnormal number, wherein the first voice message is used to inquire about the historical call intention of the suspected abnormal number; A first determining module is used to determine the abnormal type information of the suspected abnormal number according to the second voice information when the second voice information of the suspected abnormal number is obtained; A second determination module is used to determine a target call connection rate of the suspected abnormal number according to the abnormal type information of the suspected abnormal number; A restriction module, configured to restrict calls to the suspected abnormal number when the target call connection rate exceeds the target call connection rate, so that the call connection rate of the suspected abnormal number does not exceed the target call connection rate; The second determination module is specifically used to use a pre-trained call count prediction model to predict the number of calls of the suspected abnormal number in the next time period; according to the predicted number of calls and the preset maximum call volume of the number of abnormal types to which the suspected abnormal number belongs in the next time period, determine the target call connection rate of the suspected abnormal number in the next time period; wherein the call count prediction model is trained by the following steps: obtaining the number of complaints received by various abnormal numbers in the first preset time period and the call feature indicators of the second preset time period, generating a feature matrix of various abnormal numbers, the call feature indicators including at least one of call frequency, discreteness, call interval, connection rate, and call duration; training the call count prediction model based on the LASSO regression model and the feature matrix of the various abnormal numbers; the call count prediction model is trained based on the LASSO regression model and the feature matrix of the various abnormal numbers, including: configuring the following cost function: Among them, M is the number of various abnormal numbers, p is the number of call feature indicators, and x ij is the feature matrix representing the jth eigenvalue of the ith abnormal number, w j is the jth characteristic coefficient, y i is the actual number of calls to the ith abnormal number, is the predicted number of calls for the i-th abnormal number; the call frequencies of various abnormal numbers in the second preset time period are taken as dependent variables, and the call feature indicators other than the call frequencies are trained as independent variables; the model when the cost function takes the lowest value is taken as the call number prediction model.

4. A call restriction system, characterized in that: It includes an outbound call management device, a media bridge device, a voice control device and a call control device; The outbound call management device is used to: establish a call connection with a suspected abnormal number; The media bridging device is used to: play a first voice message, obtain a second voice message of the suspected abnormal number, and forward the second voice message to the voice control device, wherein the first voice message is used to inquire about the historical call intention of the suspected abnormal number; The voice control device is used to: determine the abnormal type information of the suspected abnormal number according to the second voice information, and send the abnormal type information of the suspected abnormal number to the call control device; The call control device is used to: control the call connection rate of the suspected abnormal number according to the abnormal type information of the suspected abnormal number; The call control device is pre-configured with a call count prediction model; the call control device is specifically used to: use the call count prediction model to predict the number of calls of the suspected abnormal number in the next time period; determine the target call connection rate of the suspected abnormal number in the next time period according to the predicted number of calls and the preset maximum call volume of the number of the abnormal type to which the suspected abnormal number belongs in the next time period; wherein the call count prediction model is trained by the following steps: obtaining the number of complaints received by various abnormal numbers in the first preset time period and the call feature indicators of the second preset time period, generating a feature matrix of various abnormal numbers, the call feature indicators including at least one of call frequency, discreteness, call interval, connection rate, and call duration; training the call count prediction model based on the LASSO regression model and the feature matrix of the various abnormal numbers; the training of the call count prediction model based on the LASSO regression model and the feature matrix of the various abnormal numbers includes: configuring the following cost function: Among them, M is the number of various abnormal numbers, p is the number of call feature indicators, and x ij is the feature matrix representing the jth eigenvalue of the ith abnormal number, w j is the jth characteristic coefficient, y i is the actual number of calls to the ith abnormal number, is the predicted number of calls for the i-th abnormal number; the call frequencies of various abnormal numbers in the second preset time period are taken as dependent variables, and the call feature indicators other than the call frequencies are trained as independent variables; the model when the cost function takes the lowest value is taken as the call number prediction model; and the call restrictions are imposed on the suspected abnormal numbers.

5. The system according to claim 4, characterized in that The voice control device is also used for: Performing model training on the voice information forwarded by the media bridging device to generate a meta-information extraction model; Determine the first speech information using the meta-information extraction model; The determined first voice information is sent to the media bridge device.

6. An electronic device, characterized in that: The invention comprises: a processor, and a memory storing computer program instructions; The processor reads and executes the computer program instructions to implement the call restriction method as claimed in claim 1 or 2.

7. A computer storage medium, characterized in that: The computer storage medium stores computer program instructions, which implement the call restriction method as claimed in claim 1 or 2 when executed by the processor.

Images