A method, device and electronic equipment for identifying a nuisance call

Abstract

The embodiment of the present application provides a kind of identification method, device and electronic equipment of harassment telephone, the method includes: according to the opposite end user group of to-be-measured user, from the each network node of first network model of pre-constructed first network node, the first connection relationship between each first network node and the first close coefficient of each first connection relationship are extracted, to obtain the second network model corresponding to the to-be-measured user;Wherein, the first network model is used to indicate the connection relationship and close coefficient between each network node, and the network node includes user node and attribute node;The close degree of the opposite end user group is calculated according to the second network model;According to the close degree, determine whether the to-be-measured user is harassment telephone or not.By the embodiment of the present application, it is realized that harassment telephone and normal user are more accurately distinguished, and the identification success rate of harassment telephone is improved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method, apparatus, and electronic device for identifying nuisance calls. Background Technology

[0002] Existing technologies for identifying nuisance calls mostly employ traditional data mining algorithms based on pattern recognition. These algorithms analyze the behavioral characteristics of mobile phone users, using nuisance callers and mobile phone number users as positive and negative samples to train a model and form a classification algorithm.

[0003] With the rise of the mobile internet industry, some services have seen concentrated calling activity, such as express delivery and food delivery. This concentrated calling activity is a normal need for users and is not considered nuisance. Similarly, administrative departments such as schools and governments also have legitimate calling needs, such as communication between the Party and the public, communities, and students, and for epidemic prevention and control. Their normal calling activity, like nuisance calls, exhibits the same concentrated characteristics. If a school teacher calls more than 50 different mobile phone numbers a day to contact parents and inform students about work, and these calls are short and the communication between parents is relatively loose, traditional methods simply classify all non-mobile phone numbers as nuisance calls, and also consider these frequently called normal numbers as nuisance calls. This makes it difficult to distinguish these calls from customers, severely impacting users' service usage.

[0004] Traditional methods cannot accurately distinguish between mass calls from legitimate users and nuisance calls, leading to numerous misidentifications and incorrect handling of nuisance calls. Summary of the Invention

[0005] The purpose of this invention is to provide a method, device, and electronic device for identifying nuisance calls, in order to solve the problem that the inability to accurately distinguish between normal users making concentrated calls and nuisance calls leads to a large number of incorrect identifications and incorrect handling of nuisance calls.

[0006] To solve the above-mentioned technical problems, the embodiments of the present invention are implemented as follows:

[0007] In a first aspect, embodiments of the present invention provide a method for identifying harassing phone calls, comprising:

[0008] Based on the peer user group of the user under test, the first network node, the first connection relationship between each first network node, and the first tightness coefficient of each first connection relationship are extracted from each network node of the pre-constructed first network model to obtain the second network model corresponding to the user under test; wherein, the first network model is used to indicate the connection relationship and tightness coefficient between each network node, the network node includes user node and attribute node, the first network node includes the first user node corresponding to each peer user in the peer user group, and the second user node and attribute node that have connection relationship with each first user node;

[0009] The density of the peer user group is calculated based on the second network model;

[0010] The degree of density is used to determine whether the user being tested is making a harassing call.

[0011] Secondly, embodiments of the present invention provide a device for identifying harassing phone calls, comprising:

[0012] The model unit is used to extract, based on the peer user group of the user under test, a first network node, a first connection relationship between the first network nodes, and a first tightness coefficient of each first connection relationship from each network node of a pre-constructed first network model to obtain a second network model corresponding to the user under test; wherein, the first network model is used to indicate the connection relationship and tightness coefficient between each network node, the network node includes user nodes and attribute nodes, the first network node includes a first user node corresponding to each peer user in the peer user group, and a second user node and attribute node that have connection relationships with each first user node;

[0013] A calculation unit is used to calculate the density of the peer user group based on the second network model;

[0014] The judgment unit is used to determine whether the user under test is making a harassing call based on the density.

[0015] Thirdly, embodiments of the present invention provide an electronic device, including a processor, a communication interface, a memory, and a communication bus; wherein the processor, the communication interface, and the memory communicate with each other via the bus; the memory is used to store computer programs; and the processor is used to execute the programs stored in the memory to implement the steps of the harassing call identification method as described in the first aspect.

[0016] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the harassing call identification method as described in the first aspect.

[0017] As can be seen from the technical solutions provided by the above embodiments of the present invention, the embodiments of the present invention extract first network nodes, first connection relationships between first network nodes, and first tightness coefficients of each first connection relationship from each network node of a pre-constructed first network model based on the peer user group of the user to be tested, to obtain a second network model corresponding to the user to be tested; calculate the tightness of the user to be tested based on the second network model; and determine whether the user to be tested is a nuisance caller based on the tightness. Through the embodiments of the present invention, a more accurate distinction is made between nuisance calls and normal users, improving the success rate of nuisance call identification. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A schematic flowchart of a method for identifying harassing calls provided in an embodiment of the present invention;

[0020] Figure 2 A schematic diagram of the structure of a first network model provided in an embodiment of the present invention;

[0021] Figure 3 A schematic diagram of a method for constructing a second network model provided in an embodiment of the present invention;

[0022] Figure 4 Another flowchart illustrating the method for identifying harassing calls provided in an embodiment of the present invention;

[0023] Figure 5 This is a schematic diagram of a second network model provided in an embodiment of the present invention;

[0024] Figure 6 A schematic diagram of the structure of a third network model provided in an embodiment of the present invention;

[0025] Figure 7 A schematic diagram of the module composition of the harassment call identification device provided in an embodiment of the present invention;

[0026] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0027] This invention provides a method, apparatus, and electronic device for identifying harassing phone calls.

[0028] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0029] like Figure 1 As shown, this embodiment of the invention provides a method for identifying harassing phone calls. The execution subject of this method can be a server, which can be a standalone server or a server cluster composed of multiple servers. Furthermore, the server can be a server capable of network operation processing, such as a server configuring network resources. Specifically, the method may include the following steps:

[0030] Step S110: Based on the peer user group of the user to be tested, extract the first network node, the first connection relationship between each first network node, and the first tightness coefficient of each first connection relationship from each network node of the pre-constructed first network model to obtain the second network model corresponding to the user to be tested; wherein, the first network model is used to indicate the connection relationship and tightness coefficient between each network node, the network node includes user node and attribute node, the first network node includes the first user node corresponding to each peer user in the peer user group, and the second user node and attribute node that have connection relationship with each first user node.

[0031] A first network model is pre-constructed based on relevant information from attribute tags of all network end-users and big data. The first network model includes: each network node, the connection relationships between network nodes, and the density coefficients. The construction method of the first network model can be configured according to actual needs; this embodiment only provides one implementation method.

[0032] Each network node is constructed based on the acquired terminal users and the set attribute tags. The network nodes include user nodes corresponding to each terminal user and attribute nodes corresponding to each attribute tag.

[0033] It should be understood that the terminal users can include all terminal users on the operator's network and cross-network numbers, etc.

[0034] The attribute tags can include various social and behavioral attributes. When selecting attribute tags, one can choose the business association with the telecommunications operator, or one can choose the attribute that needs to be confirmed to have a certain potential social association, such as: ID card residence, workplace identified by big data model, residence identified by big data model, home broadband community, belonging to a corporate customer, participating in corporate customer customized activities, ordering customized packages, ordering a certain school communication network, real estate preferences, financial preferences, etc.

[0035] Based on the association information between each terminal user, the connection relationship between each user node is determined, and based on the association information between each terminal user and each attribute tag, the connection relationship between each user node and each attribute node is determined.

[0036] It should be understood that the association information between end users can include the relationships that occur between end users during the business processes of telecommunications operators, including direct and indirect relationships. Direct relationships refer to the relationships generated by direct interactions between end users, such as calls, SMS, and MMS, like SMS and MMS relationships, which are the closest relationships. Business relationships between end users, i.e., indirect relationships, include family networks, account merging, primary and secondary SIM cards, data sharing, transfers, and points rewards, which can clearly identify the social connections between two user nodes. Through these business relationships, the connection relationships between various end nodes can be constructed.

[0037] It should be understood that the association information between each end user and attribute tags can include the attribute tags possessed by each end user and their corresponding tag values, thereby constructing aggregated attributes. Aggregated attributes are beneficial for expanding their social connections. Taking school notifications to parents as an example, the relationships between parents, in addition to the relationships linked to students through school communication platforms and family networks, also largely involve relationships with parents living near the school. These relationships need to be reflected using the aggregated attributes of regions and grids.

[0038] For example, big data analysis identifies a user's residence as Zilai Garden Community, Wanxing Street, Nankai District, Tianjin. This can be categorized into three attribute tags: the complete and accurate address (Zilai Garden Community, Wanxing Street, Nankai District, Tianjin), the address aggregated to the grid (Wanxing Street, Nankai District, Tianjin), and the address aggregated to the region (Nankai District, Tianjin). These form three attribute nodes, and the user node corresponding to that user establishes connections to these three attribute nodes.

[0039] Based on the evaluation principles of each connection relationship, the tightness coefficient of each connection relationship is determined. A hierarchical scoring algorithm can be used to determine the tightness coefficient of each connection relationship to identify the degree of tightness of that connection relationship. In one implementation, the evaluation principles for each connection relationship may be: direct relationships are greater than indirect relationships, detailed address relationships are greater than convergent relationships, and the higher the frequency of call relationships, the higher the tightness coefficient. The frequency of call relationships can be evaluated based on the number of calls and the call duration.

[0040] The first network model is constructed by summing the values ​​of each user node, attribute node, connection relationship, and the tightness coefficient of each connection relationship. For example... Figure 2 The network model shown includes user nodes A1, A2, and A3, and attribute nodes B1 and B2. Each connection between the network nodes represents the connection relationship between them, and each connection relationship has a corresponding tightness coefficient.

[0041] After analyzing the call information of an end user, if the server determines that there is centralized dialing behavior, it identifies that end user as a test user and extracts the peer user group of that test user. The peer user group can be composed of peer users who have been identified as having centralized dialing behavior with the test user.

[0042] From all network nodes in the first network model, user nodes and attribute nodes related to each peer user of the user under test are extracted as first network nodes. At the same time, the connection relationships and corresponding tightness coefficients between each first network node are extracted from the first network model as first connection relationships and corresponding first tightness coefficients to form a second network model. The second network model is equivalent to a sub-model in the first network model corresponding to the user under test.

[0043] The selection method and selection range of the first network node can be set according to actual needs. In one implementation, such as... Figure 3 As shown, the first network node includes user nodes corresponding to each peer user as the first user node, such as... Figure 3 The user nodes A4 and A5 shown are included. The first network node also includes second user nodes that have a first connection relationship with each of the first user nodes A4 and A5, in addition to the user node corresponding to the user under test. Figure 3 The user nodes A6 and A7 shown, and the attribute nodes that have a first connection relationship with the first user nodes A4 and A5, are as follows: Figure 3 The attribute nodes B3, B4, and B5 are shown. Since the user under test is in communication with all peer users, the user node corresponding to the user under test must be removed from the first network node.

[0044] Step S120: Calculate the density of the peer user group based on the second network model;

[0045] Based on the first connection relationship between each first network node in the second network model and the first tightness coefficient of each first connection, the server calculates the tightness of the peer user group using a preset calculation method, that is, the tightness of the peer user group of the user to be tested.

[0046] Step S130: Determine whether the user under test is a nuisance caller based on the density. The density is used to determine whether the density of the other end users of the user under test is high enough. If the density of the other end users is low, the user under test can be determined to be a nuisance caller. If the density of the other end users is high, the user under test can be determined to be a normal user.

[0047] As can be seen from the technical solutions provided by the above embodiments of the present invention, the embodiments of the present invention extract first network nodes, first connection relationships between first network nodes, and first tightness coefficients of each first connection relationship from each network node of a pre-constructed first network model based on the peer user group of the user to be tested, to obtain a second network model corresponding to the user to be tested; calculate the tightness of the user to be tested based on the second network model; and determine whether the user to be tested is a nuisance caller based on the tightness. Through the embodiments of the present invention, a more accurate distinction is made between nuisance calls and normal users, improving the success rate of nuisance call identification.

[0048] Based on the above embodiments, further, as Figure 4 As shown, there are many ways to process step S120. Here is another optional processing method. For details, please refer to the processing of steps S121-S122 below.

[0049] Step S121: Based on the first connection relationship and the corresponding first tightness coefficient between each first network node in the second network model, obtain the second connection relationship and the second tightness coefficient between each first user node to obtain the third network model; wherein, the second connection relationship is determined based on the direct and indirect first connection relationship between each first user node.

[0050] Based on the first connection relationships and corresponding first tightness coefficients between each first user node and between each first user node and the first attribute node in the second network model, the second network model is simplified to determine whether there are second connection relationships between each first user node. If so, the second tightness coefficients between the first user nodes are calculated. The first user nodes, the second connection relationships between each first user node, and the second tightness coefficients are used to form a third network model corresponding to the user under test.

[0051] Furthermore, the specific processing method of step S121 above can be varied, such as... Figure 5 As shown, only one implementation method is given below.

[0052] Based on the first connection relationships between the first network nodes in the second network model, the first user node i and the first user node j corresponding to any two peer users in all peer user groups are selected one by one. The first-level link and the second-level link between the first user node i and the first user node j are determined, and the first user node i and the first user node j with a first-level link and / or a second-level link are determined to have a second connection relationship. Here, the first-level link is the first connection relationship between the first user node i and the first user node j, and the second-level link is the indirect first connection relationship between the first user node i and the second user node j with the first network node k as an intermediate node; for example, as... Figure 5 The first-level links between the first user node i and the first user node j are the first connection relationships L1, L2 and L3. The intermediate nodes of the second-level links between the first user node i and the second user node j are attribute node k1 and user node k2, respectively.

[0053] Based on the first connection relationships and corresponding first compactness coefficients among the first user node i, first user node j, and first network node k, the compactness coefficient R of each first-level link is obtained. ij The tightness coefficient R of each secondary link ikj .

[0054] There are various methods for calculating the compactness coefficient of the primary link. In one embodiment, the compactness coefficient R of the primary link is... ij For example, the first tightness coefficient of the corresponding first connection relationship, such as Figure 5 As shown, the first compactness coefficients of the first-level links L1, L2, and L3 are R, respectively. ij =7, R ij =6, R ij =9.

[0055] There are various methods for calculating the compactness coefficient of the secondary link. In one embodiment, the compactness coefficient R of the secondary link is... ikj It is obtained from the following calculation formula:

[0056] R ikj =min(∑R ik ,∑R kj )

[0057] Among them, R ik R is the first tightness coefficient of the first connection relationship between the first user node i and the first network node k. kjIt is the first tightness coefficient of the first connection relationship between the first network node k and the first user node j.

[0058] like Figure 5 As shown, for attribute node k1, the obtained R ik1j =min(R) ik1 ,R k1j ), where R ik1 and R k1j These are the first tightness coefficients for the first connectivity relationships L4 and L5, respectively. R is calculated. ik1j =min(R) ik1 ,R k1j )=min(4,3)=3.

[0059] For user node k2, the obtained R ik2j =min(∑R ik2 ,R k2j ), where R ik2 Including the first tightness coefficients of the first connectivity relationships L6, L7, and L8, R k2j R is the first tightness coefficient of the first connectivity relationship L9. ik2j =min(∑R ik2 ,R k2j ) = min((5+7+6),7) = 7.

[0060] Based on the compactness coefficient R of each primary link ij The tightness coefficient R of each secondary link ikj The second tightness coefficient T between the first user node i and the first user node j is obtained through the first calculation formula. ij .

[0061] In one implementation, the first calculation formula is expressed as follows:

[0062] T ij =∑R ij +∑R ikj .

[0063] like Figure 5 As shown, the second compactness coefficient T ij =∑R ij +(R ik1j +R ik2j )=(7+6+9)+(3+7)=32.

[0064] The third network model is obtained by combining all the first user nodes, the second connection relationships between each first user node, and the corresponding second tightness coefficients. For example, Figure 6 As shown, the second tightness coefficient among the first user nodes A1, A2, A3, and A4 is T. 13T 23 T 24 T 34 .

[0065] Step S122: Based on the second connection relationship and second tightness coefficient between each first user node in the third network model, the tightness of the peer user group is obtained through the tightness calculation method.

[0066] Furthermore, there are many different ways to process step S122 above. Only one implementation method is given below.

[0067] Based on the second connection relationship and the second tightness coefficient among the first user nodes in the third network model, the mean value L of the second tightness coefficient is calculated respectively. ss The proportion C of the compactness coefficient of the fully connected triangle in the third network model. ss Among them, there is a second connection relationship between the three first user nodes that make up the fully connected triangle;

[0068] Based on the mean value L of the second tightness coefficient ss The proportion C of the compactness coefficient of the fully connected triangle. ss The density of the user under test is obtained by using a density calculation method.

[0069] In one embodiment, the density calculation method is expressed as follows: Density = L ss ×C ss .

[0070] In one implementation, the mean value L of the second compactness coefficient ss It is obtained from the following calculation formula:

[0071]

[0072] Among them, T ij N is the second compactness coefficient between the first user node i and the first user node j. i This refers to the number of first user nodes in the third network model. For example... Figure 6 As shown, the calculation process for the mean of the second compactness coefficient is as follows:

[0073]

[0074] In one implementation, the proportion C of the compactness coefficients of the fully connected triangles in the third network model is... ss It is obtained from the following calculation formula:

[0075]

[0076] Wherein, the DCt D is the compactness coefficient of a fully connected triangle. nt The compactness coefficient is the sum of the second compactness coefficients among the three first user nodes that make up the fully connected triangle. The compactness coefficient of the non-fully connected triangle is the sum of the second compactness coefficients among the three first user nodes that make up the non-fully connected triangle.

[0077] For example Figure 6 As shown in the example, the fully connected triangle includes a triangle composed of A2, A3, and A4, and its corresponding compactness coefficient D Ct =T 23 +T 24 +T 34 =29.

[0078] Non-fully connected triangles include triangles composed of A1, A2, A3 and A1, A3, A4, with corresponding compactness coefficients D. nt =T 13 +T 23 =12 and D nt =T 13 +T 34 =21.

[0079] The proportion of compactness coefficients of fully connected triangles

[0080] Therefore, the peer user of the user under test is obtained.

[0081] Further, step S130 includes:

[0082] When the density of the user under test is higher than a preset threshold, the user under test is determined to be a normal user and there is no need to shut down the base.

[0083] If the density of the user under test is not higher than a preset threshold, the user under test is determined to be a nuisance caller and nuisance call processing is performed.

[0084] For example, if the set threshold is 0.6, then based on Figure 6 If the density obtained exceeds the threshold, the user under test can be determined to be a normal user.

[0085] The threshold needs to be set before the network model goes live. There are various methods for setting it. In one implementation, it needs to be calculated based on the actual results of historical labeled data. A certain number of sample data are selected and manually called out beforehand, labeled as nuisance calls and normal users. For each sample in these training data, a social network subgraph of the customer group is constructed, and the density coefficient is calculated. For the training data results with labeled positive and negative samples, a univariate logistic regression algorithm or an expert-based method can be used to determine the final threshold of the density coefficient. Since different strategies are adopted in actual nuisance call handling, for example, during specific holidays, there will be fewer nuisance calls and more normal calls. To avoid disturbing normal customers, it is necessary to improve accuracy and reduce recall. Therefore, in practical applications, it is recommended to use an expert-based method to balance accuracy and recall based on actual needs and determine the threshold to be used.

[0086] As can be seen from the technical solutions provided by the above embodiments of the present invention, the embodiments of the present invention simplify the second network model to obtain a third network model. Based on the second density coefficient of each first user node in the third network model, the density of the peer user group of the user under test is calculated using a preset calculation formula, which is used to determine whether the user under test is making a harassing call. Through the embodiments of the present invention, the distinction between harassing calls and normal users is achieved more accurately, improving the success rate of harassing call identification.

[0087] Corresponding to the harassment call identification method provided in the above embodiments, based on the same technical concept, this embodiment of the invention also provides a harassment call identification device. Figure 7 This is a schematic diagram of the module composition of a nuisance call identification device provided in an embodiment of the present invention. The nuisance call identification device is used to perform... Figures 1 to 6 The described method for identifying nuisance calls, such as... Figure 7 As shown, the harassment call identification device includes: a model unit 701, a calculation unit 702, and a judgment unit 703.

[0088] The model unit 701 is used to extract first network nodes, first connection relationships between first network nodes, and first tightness coefficients of each first connection relationship from each network node of a pre-constructed first network model based on the peer user group of the user to be tested, so as to obtain a second network model corresponding to the user to be tested; wherein, the first network model is used to indicate the connection relationships and tightness coefficients between each network node, the network node includes user nodes and attribute nodes, the first network node includes first user nodes corresponding to each peer user in the peer user group, and second user nodes and attribute nodes that have connection relationships with each first user node; the calculation unit 702 is used to calculate the tightness of the peer user group based on the second network model; the judgment unit 703 is used to determine whether the user to be tested is a harassing caller based on the tightness.

[0089] Furthermore, the model unit is also used for:

[0090] Each network node is constructed based on the acquired terminal users and the set attribute tags. The network nodes include user nodes corresponding to each terminal user and attribute nodes corresponding to each attribute tag.

[0091] Based on the association information between each terminal user, the connection relationship between each user node is determined, and based on the association information between each terminal user and each attribute tag, the connection relationship between each user node and each attribute node is determined.

[0092] Based on the evaluation principles of each connection relationship, the tightness coefficient of each connection relationship is determined;

[0093] The first network model is constructed by summing up the user nodes, attribute nodes, connection relationships, and the tightness coefficients of each connection relationship.

[0094] Furthermore, the first network node includes a first user node corresponding to each peer user, a second user node having a first connection relationship with each first user node (excluding the user node corresponding to the user under test), and an attribute node having a first connection relationship with the first user node.

[0095] As can be seen from the technical solutions provided by the above embodiments of the present invention, the embodiments of the present invention extract first network nodes, first connection relationships between first network nodes, and first tightness coefficients of each first connection relationship from each network node of a pre-constructed first network model based on the peer user group of the user to be tested, to obtain a second network model corresponding to the user to be tested; calculate the tightness of the user to be tested based on the second network model; and determine whether the user to be tested is a nuisance caller based on the tightness. Through the embodiments of the present invention, a more accurate distinction is made between nuisance calls and normal users, improving the success rate of nuisance call identification.

[0096] Furthermore, the computing unit is used for:

[0097] Based on the first connection relationship and the corresponding first tightness coefficient between each first network node in the second network model, the second connection relationship and the second tightness coefficient between each first user node are obtained to obtain the third network model; wherein, the second connection relationship is determined based on the direct and indirect first connection relationship between each first user node;

[0098] Based on the second connection relationship and second tightness coefficient between each first user node in the third network model, the tightness of the peer user group is obtained through the tightness calculation method.

[0099] Furthermore, the computing unit is used for:

[0100] Based on the first connection relationship between each first network node in the second network model, the first-level link and the second-level link between the first user node i and the first user node j are determined, and the first user node i and the first user node j with a first-level link and / or a second-level link are determined to have a second connection relationship; wherein, the first-level link is the first connection relationship between the first user node i and the first user node j, and the second-level link is the indirect first connection relationship between the first user node i and the second user node j with the first network node k as an intermediate node;

[0101] Based on the first connection relationships and corresponding first compactness coefficients among the first user node i, first user node j, and first network node k, the compactness coefficient R of each first-level link is obtained. ij The tightness coefficient R of each secondary link ikj ;

[0102] Based on the compactness coefficient R of each primary link ij The tightness coefficient R of each secondary link ikj The second tightness coefficient T between the first user node i and the first user node j is obtained through the first calculation formula. ij .

[0103] Furthermore, the first calculation formula is expressed as follows:

[0104] T ij =∑R ij +∑R ikj .

[0105] Furthermore, the compactness coefficient R of the first-level link ij This is the first tightness coefficient of the corresponding first connection relationship.

[0106] Furthermore, the compactness coefficient R of the secondary link ikjIt is obtained from the following calculation formula:

[0107] R ikj =min(∑R ik ,∑R kj )

[0108] Among them, R ik R is the first tightness coefficient of the first connection relationship between the first user node i and the first network node k. kj It is the first tightness coefficient of the first connection relationship between the first network node k and the first user node j.

[0109] Furthermore, the computing unit is used for:

[0110] Based on the second connection relationship and the second tightness coefficient among the first user nodes in the third network model, the mean value L of the second tightness coefficient is calculated respectively. ss The proportion C of the compactness coefficient of the fully connected triangle in the third network model. ss Among them, there is a second connection relationship between the three first user nodes that make up the fully connected triangle;

[0111] Based on the mean value L of the second tightness coefficient ss The proportion C of the compactness coefficient of the fully connected triangle. ss The density of the peer user group is obtained by using a density calculation method.

[0112] Furthermore, the density calculation method is expressed as follows: Density = L ss ×C ss .

[0113] Furthermore, the mean L of the second compactness coefficient ss It is obtained from the following calculation formula:

[0114]

[0115] Among them, T ij N is the second compactness coefficient between the first user node i and the first user node j. i The number of first user nodes in the third network model.

[0116] Furthermore, the proportion C of the compactness coefficient of the fully connected triangle in the third network model ss It is obtained from the following calculation formula:

[0117]

[0118] Wherein, the D Ct D is the compactness coefficient of a fully connected triangle. ntThe compactness coefficient is the sum of the second compactness coefficients among the three first user nodes that make up the fully connected triangle. The compactness coefficient of the non-fully connected triangle is the sum of the second compactness coefficients among the three first user nodes that make up the non-fully connected triangle.

[0119] Furthermore, the determination unit is used for:

[0120] When the density of the peer user group is higher than a preset threshold, the user to be tested is determined to be a normal user;

[0121] If the density of the peer user group is not higher than a preset threshold, the user to be tested is determined to be a harassing caller.

[0122] As can be seen from the technical solutions provided by the above embodiments of the present invention, the embodiments of the present invention simplify the second network model to obtain a third network model. Based on the second density coefficient of each first user node in the third network model, the density of the peer user group of the user under test is calculated using a preset calculation formula, which is used to determine whether the user under test is making a harassing call. Through the embodiments of the present invention, the distinction between harassing calls and normal users is achieved more accurately, improving the success rate of harassing call identification.

[0123] The harassment call identification device provided in this embodiment of the invention can realize the various processes in the embodiments corresponding to the above-mentioned harassment call identification method. To avoid repetition, it will not be described again here.

[0124] It should be noted that the harassment call identification device provided in this embodiment of the invention and the harassment call identification method provided in this embodiment of the invention are based on the same inventive concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned harassment call identification method, and the repeated parts will not be described again.

[0125] Corresponding to the harassment call identification method provided in the above embodiments, based on the same technical concept, this embodiment of the invention also provides an electronic device for executing the above-described harassment call identification method. Figure 8 To illustrate the structure of an electronic device according to various embodiments of the present invention, as shown in the schematic diagram... Figure 8As shown. Electronic devices can vary considerably due to differences in configuration or performance, and may include one or more processors 801 and memory 802. Memory 802 may store one or more application programs or data. Memory 802 may be temporary or persistent storage. The application programs stored in memory 802 may include one or more modules (not shown), each module may include a series of computer-executable instructions for the electronic device. Furthermore, processor 801 may be configured to communicate with memory 802 and execute the series of computer-executable instructions in memory 802 on the electronic device. The electronic device may also include one or more power supplies 803, one or more wired or wireless network interfaces 804, one or more input / output interfaces 805, and one or more keyboards 806.

[0126] Specifically, in this embodiment, the electronic device includes a processor, a communication interface, a memory, and a communication bus; wherein, the processor, the communication interface, and the memory communicate with each other via the bus; the memory is used to store computer programs; the processor is used to execute the programs stored in the memory to implement the following method steps:

[0127] Based on the peer user group of the user under test, the first network node, the first connection relationship between each first network node, and the first tightness coefficient of each first connection relationship are extracted from each network node of the pre-constructed first network model to obtain the second network model corresponding to the user under test; wherein, the first network model is used to indicate the connection relationship and tightness coefficient between each network node, the network node includes user node and attribute node, the first network node includes the first user node corresponding to each peer user in the peer user group, and the second user node and attribute node that have connection relationship with each first user node;

[0128] The density of the peer user group is calculated based on the second network model;

[0129] The degree of density is used to determine whether the user being tested is making a harassing call.

[0130] This application embodiment also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the following method steps:

[0131] Based on the peer user group of the user under test, the first network node, the first connection relationship between each first network node, and the first tightness coefficient of each first connection relationship are extracted from each network node of the pre-constructed first network model to obtain the second network model corresponding to the user under test; wherein, the first network model is used to indicate the connection relationship and tightness coefficient between each network node, the network node includes user node and attribute node, the first network node includes the first user node corresponding to each peer user in the peer user group, and the second user node and attribute node that have connection relationship with each first user node;

[0132] The density of the peer user group is calculated based on the second network model;

[0133] The degree of density is used to determine whether the user being tested is making a harassing call.

[0134] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0135] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0136] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0137] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0138] In a typical configuration, an electronic device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0139] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0140] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0141] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0142] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0143] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for identifying harassing phone calls, characterized in that, The method includes: Based on the peer user group of the user under test, the first network node, the first connection relationship between each first network node, and the first tightness coefficient of each first connection relationship are extracted from each network node of the pre-constructed first network model to obtain the second network model corresponding to the user under test; wherein, the first network model is used to indicate the connection relationship and tightness coefficient between each network node, the network node includes user node corresponding to each terminal user and attribute node corresponding to each attribute tag, the attribute tag includes residence, the first network node includes first user node corresponding to each peer user in the peer user group, and second user node and attribute node that have connection relationship with each first user node; The density of the peer user group is calculated based on the second network model; Based on the density, determine whether the user being tested is making a harassing call; The density of the peer user group calculated according to the second network model includes: Based on the first connection relationship and the corresponding first tightness coefficient between each first network node in the second network model, the second connection relationship and the second tightness coefficient between each first user node are obtained to obtain the third network model; wherein, the second connection relationship is determined based on the direct and indirect first connection relationship between each first user node; Based on the second connection relationship and second tightness coefficient between each first user node in the third network model, the tightness of the peer user group is obtained through the tightness calculation method.

2. The method according to claim 1, characterized in that, The step of obtaining the second connection relationship and second tightness coefficient between each first user node based on the first connection relationship and corresponding first tightness coefficient between each first network node in the second network model includes: Based on the first connection relationship between each first network node in the second network model, the first-level link and the second-level link between the first user node i and the first user node j are determined, and the first user node i and the first user node j with a first-level link and / or a second-level link are determined to have a second connection relationship; wherein, the first-level link is the first connection relationship between the first user node i and the first user node j, and the second-level link is the indirect first connection relationship between the first user node i and the second user node j with the first network node k as an intermediate node; Based on the first connection relationships and corresponding first compactness coefficients among the first user node i, first user node j, and first network node k, the compactness coefficients of each first-level link are obtained. And the tightness coefficient of each secondary link ; Based on the compactness coefficient of each primary link And the tightness coefficient of each secondary link The second tightness coefficient between the first user node i and the first user node j is obtained through the first calculation formula. .

3. The method according to claim 2, characterized in that, The first calculation formula is expressed as follows: Wherein, the compactness coefficient of the first-level link This is the first tightness coefficient of the corresponding first connection relationship; The compactness coefficient of the secondary link It is obtained from the following calculation formula: in, Let be the first tightness coefficient of the first connection relationship between the first user node i and the first network node k. It is the first tightness coefficient of the first connection relationship between the first network node k and the first user node j.

4. The method according to claim 1, characterized in that, The step of obtaining the density of the peer user group based on the second connection relationship and second density coefficient between each first user node in the third network model, through a density calculation method, includes: Based on the second connection relationship and the second tightness coefficient among the first user nodes in the third network model, the mean of the second tightness coefficient is calculated respectively. The proportion of the compactness coefficients of the fully connected triangles in the third network model. Among them, there is a second connection relationship between the three first user nodes that make up the fully connected triangle; Based on the mean of the second tightness coefficient The proportion of the compactness coefficient of the fully connected triangle. The density of the peer user group is obtained by using a density calculation method.

5. The method according to claim 4, characterized in that, The density calculation method is expressed as follows: ; Among them, the mean of the second tightness coefficient It is obtained from the following calculation formula: in, Let be the second compactness coefficient between the first user node i and the first user node j. The number of first user nodes in the third network model; The proportion of compactness coefficients of fully connected triangles in the third network model It is obtained from the following calculation formula: Among them, the The compactness coefficient of a fully connected triangle, the The compactness coefficient is the sum of the second compactness coefficients among the three first user nodes that make up the fully connected triangle. The compactness coefficient of the non-fully connected triangle is the sum of the second compactness coefficients among the three first user nodes that make up the non-fully connected triangle.

6. The method according to any one of claims 1-5, characterized in that, Before extracting the first network nodes, the first connection relationships between the first network nodes, and the first tightness coefficients of the first connection relationships from each network node of the pre-constructed first network model based on the peer user group of the user to be tested, in order to obtain the second network model corresponding to the user to be tested, the method further includes: Each network node is constructed based on the acquired terminal users and the set attribute tags. The network nodes include user nodes corresponding to each terminal user and attribute nodes corresponding to each attribute tag. Based on the association information between each terminal user, the connection relationship between each user node is determined, and based on the association information between each terminal user and each attribute tag, the connection relationship between each user node and each attribute node is determined. Based on the evaluation principles of each connection relationship, the tightness coefficient of each connection relationship is determined; The first network model is constructed by summing up the user nodes, attribute nodes, connection relationships, and the tightness coefficients of each connection relationship.

7. A device for identifying harassing phone calls, characterized in that, The device includes: The model unit is used to extract, based on the peer user group of the user under test, a first network node, a first connection relationship between the first network nodes, and a first tightness coefficient of each first connection relationship from each network node of a pre-constructed first network model to obtain a second network model corresponding to the user under test; wherein, the first network model is used to indicate the connection relationship and tightness coefficient between each network node, the network node includes a user node corresponding to each terminal user and an attribute node corresponding to each attribute label, the attribute label including place of residence, the first network node includes a first user node corresponding to each peer user in the peer user group, and a second user node and an attribute node that have a connection relationship with each first user node; A calculation unit is used to calculate the density of the peer user group based on the second network model; A judgment unit is used to determine whether the user under test is making a harassing call based on the density. In the computing unit, the density of the peer user group calculated according to the second network model includes: Based on the first connection relationship and the corresponding first tightness coefficient between each first network node in the second network model, the second connection relationship and the second tightness coefficient between each first user node are obtained to obtain the third network model; wherein, the second connection relationship is determined based on the direct and indirect first connection relationship between each first user node; Based on the second connection relationship and second tightness coefficient between each first user node in the third network model, the tightness of the peer user group is obtained through the tightness calculation method.

8. An electronic device, characterized in that, The device includes a processor, a communication interface, a memory, and a communication bus; wherein the processor, the communication interface, and the memory communicate with each other via the bus; the memory is used to store computer programs; and the processor is used to execute the programs stored in the memory to implement the steps of the harassing call identification method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the harassing call identification method as described in any one of claims 1-6.

Images