Illegal outbound marketing identification method, device and computing equipment

Abstract

The embodiment of the present invention relates to the field of mobile communication technology, and discloses a method, apparatus and computing device for identifying illegal outbound marketing calls, the method comprising: obtaining multiple features based on call signaling data, and applying a preset illegal outbound call identification model based on the multiple features to identify suspected illegal outbound numbers; associating the called number and call time in the call signaling data with the user number and service processing time in the user service subscription data to obtain suspected illegal agents and suspected illegal outbound numbers; confirming the illegal outbound numbers and suspected illegal agents based on the base station location information of the called number at the time of the call and the location information of the suspected illegal agent. Through the above-mentioned method, the embodiment of the present invention can quickly, timely and accurately identify outbound numbers and the agents involved, so that management personnel can conduct timely management and control of outbound calls and agents, and reduce complaints caused thereby.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of mobile communication technology, and specifically to a method, device and computing equipment for identifying illegal outbound marketing calls. Background Art

[0002] To expand their business, telecom companies currently recruit numerous agents. After these agents handle services for users, the telecom companies will settle fees with the agents based on the type of service. However, with the rapid development of electronic channels, many users prefer to conduct business online without leaving their homes. As a result, agent service processing has declined. To earn more fees, agents have resorted to a frenzied outbound marketing model, recommending and handling services for users. This outbound marketing method not only causes significant harassment to users but also generates numerous complaints. Therefore, it is necessary to block these marketing calls and investigate and punish the agents involved.

[0003] The current solution to the aforementioned problem is to verify whether outbound marketing is illegal based on user complaints and whether business has been processed in the Business & Operation Support System (BOSS). This method is time-consuming and difficult to handle. After a user complaint, it is necessary to audit the call records and work order records in the BOSS system to verify whether it is illegal outbound marketing. Moreover, only when a large number of complaints are caused can the illegal marketing calls and agents involved be investigated and dealt with, making it impossible to identify the channels (agents) involved in a timely, rapid, and accurate manner. Summary of the Invention

[0004] In view of the above problems, embodiments of the present invention provide a method, apparatus, and computing device for identifying illegal outbound marketing calls, which overcome the above problems or at least partially solve the above problems.

[0005] According to one aspect of an embodiment of the present invention, a method for identifying illegal outbound marketing calls is provided, the method comprising: obtaining multiple features based on call signaling data, and identifying suspected illegal outbound numbers by applying a preset illegal outbound call identification model based on the multiple features; associating the called number and call time in the call signaling data with the user number and service processing time in the user service subscription data to obtain suspected illegal agents and the suspected illegal outbound numbers; confirming the illegal outbound numbers and the suspected illegal agents based on the base station location information of the called number during the call and the location information of the suspected illegal agents.

[0006] In an optional manner, the method obtains multiple features based on call signaling data, and applies a preset illegal outbound call identification model based on the multiple features to identify suspected illegal outbound call numbers, including: selecting multiple features based on the call signaling data in combination with illegal outbound call business scenarios, and calculating feature values ​​of multiple features, the multiple features including: average daily number of calls, call dispersion, answering rate, caller-to-caller hang-up ratio, average call duration, and caller-to-call ratio; and applying a naive Bayes classification algorithm based on the multiple features to identify the suspected illegal outbound call numbers.

[0007] In an optional manner, the called number and call time in the call signaling data are respectively associated with the user number and service processing time in the user service subscription data to obtain the suspected illegal agent and the suspected illegal outbound number, including: constructing a first relationship vector between the suspected illegal outbound number and the user number and a second relationship vector between the service processing number and the user number based on the call signaling data and the user service subscription data, and identifying the suspected illegal agent based on the first relationship vector and the second relationship vector; and applying preset rules to further determine the suspected illegal agent and the suspected illegal outbound number based on the behavior of the called number in handling the service after the suspected illegal outbound number makes an outbound call.

[0008] In an optional manner, the first relationship vector between the suspected illegal outbound call number and the user number and the second relationship vector between the business handling number and the user number are constructed based on the call signaling data and the user service subscription data, including: obtaining the first relationship between the illegal outbound call number and the user number based on the call signaling data; obtaining the second relationship between the business handling number and the user number based on the user service subscription data; associating the first relationship and the second relationship based on the user number, and constructing and generating the first relationship vector and the second relationship vector.

[0009] In an optional manner, the identifying of suspected illegal agents based on the first relationship vector and the second relationship vector includes: applying the following relationship to calculate the first relationship vector l i =(l1,l2,l3,…,l n ) and the second relationship vector m i =(m1,m2,m3,…,m n )’s cosine similarity:

[0010]

[0011] The agent corresponding to the business handling number whose cosine similarity is not less than a preset value is matched as the suspected illegal agent.

[0012] In an optional manner, the suspected illegal outbound call number and the suspected illegal outbound call number are further determined by applying preset rules based on the behavior of the called number in handling business after the suspected illegal outbound call number makes a call, including: if the time interval between the call time of the suspected illegal outbound call number and the business handling time is less than a first time interval; and within a preset period, the number of users who handle business within the first time interval after the suspected illegal outbound call number makes an outbound call within the preset period is not less than a first value, then the suspected illegal outbound call number and the suspected illegal agent who handles the business are further determined.

[0013] In an optional manner, the confirmation of the illegal outbound call number and the illegal agent based on the first location information of the base station where the called number is located and the second location information of the suspected illegal agent according to the suspected illegal outbound call number and the suspected illegal agent includes: first latitude and longitude information of the base station where the called number is located obtained from the network side; second latitude and longitude information of the suspected illegal agent for business processing obtained from the BOSS system side; calculation of the distance between the called number when it was marketed and when the business was processed based on the first latitude and longitude information and the second latitude and longitude information; if the distance is greater than or equal to a preset distance, the suspected illegal outbound call number and the suspected illegal agent are respectively determined as an illegal outbound call number and an illegal agent.

[0014] According to another aspect of an embodiment of the present invention, a device for identifying illegal outbound marketing calls is provided, the device comprising: a first identification unit, for obtaining a plurality of features based on call signaling data, and identifying suspected illegal outbound numbers by applying a preset illegal outbound call identification model based on the plurality of features; a second identification unit, for associating the called number and call time in the call signaling data with the user number and service processing time in the user service subscription data, respectively, to obtain suspected illegal agents and the suspected illegal outbound numbers; a violation determination unit, for confirming illegal outbound numbers and illegal agents based on base station location information of the suspected illegal outbound numbers and the suspected illegal agents when the called number makes a call and location information of the suspected illegal agents.

[0015] According to another aspect of an embodiment of the present invention, there is provided a computing device, comprising: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus;

[0016] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute the steps of the above-mentioned illegal outbound marketing identification method.

[0017] According to another aspect of an embodiment of the present invention, a computer storage medium is provided, wherein the storage medium stores at least one executable instruction, and the executable instruction enables the processor to execute the steps of the above-mentioned illegal outbound marketing identification method.

[0018] The embodiment of the present invention obtains multiple features based on call signaling data, and applies a preset illegal outbound call identification model based on the multiple features to identify suspected illegal outbound call numbers; associates the called number and call time in the call signaling data with the user number and service processing time in the user service subscription data to obtain suspected illegal agents and the suspected illegal outbound call number; confirms the illegal outbound call number and the suspected illegal agent based on the base station location information of the called number when the call is made and the location information of the suspected illegal agent, so as to quickly, promptly and accurately identify the outbound call number and the agent involved, so that management personnel can conduct timely management and control of outbound calls and agents, thereby reducing complaints caused by this.

[0019] The above description is only an overview of the technical solutions of the embodiments of the present invention. In order to more clearly understand the technical means of the embodiments of the present invention, they can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the embodiments of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are specifically listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0021] Figure 1 A schematic diagram showing a flow chart of a method for identifying illegal outbound marketing calls provided by an embodiment of the present invention is shown;

[0022] Figure 2 A schematic diagram showing another method for identifying illegal outbound marketing calls provided by an embodiment of the present invention is shown;

[0023] Figure 3 A schematic diagram showing the structure of an illegal outbound marketing identification device provided by an embodiment of the present invention is shown;

[0024] Figure 4 A schematic structural diagram of a computing device provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0025] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0026] Figure 1 The flowchart of the method for identifying illegal outbound marketing provided by the embodiment of the present invention is shown in FIG. Figure 1 As shown, methods for identifying illegal outbound marketing include:

[0027] Step S11: obtaining multiple features based on the call signaling data, and applying a preset illegal outbound call identification model based on the multiple features to identify suspected illegal outbound call numbers.

[0028] In an embodiment of the present invention, call signaling data is collected from the network side, such as from a network management system. The key fields of the collected call signaling data are shown in Table 1, including calling number, called number, call time, etc.

[0029] Table 1 Call signaling data key fields

[0030] Serial number Field Name illustrate 1 Caller ID Agent outbound call number 2 called number Called user number 3 Talk time Call start time 4 Ringing duration Ringing duration 5 Call duration Call duration 6 Phone connected sign Is the call connected? 7 Caller hang-up sign Whether to hang up the calling number 8 Callee hangs up sign Whether the called number hangs up 9 Called number location The location of the base station to which the called number belongs

[0031] In step S11, multiple features are selected based on the call signaling data and the illegal outbound call business scenario, and the feature values ​​of the multiple features are calculated. The multiple features include: average number of calls per day, call dispersion, answer rate, caller-to-call hang-up ratio, average call duration and caller-to-call ratio; based on the multiple features, a naive Bayes classification algorithm is applied to identify the suspected illegal outbound call number.

[0032] The average daily number of calls is calculated as the number of times the calling number appears in signaling (including both connected and unconnected calls) divided by the number of days the calling number is in call mode. For example, if a calling number makes 100 calls on 20 days in a given month, the average daily number of calls is 100 / 20 = 5. Call dispersion is expressed as the number of calls made / the number of calls. The answer rate is expressed as the number of connected calls / the number of calls. The caller-to-caller hang-up ratio is expressed as the number of caller hang-ups / the number of callee hang-ups. The average call duration is expressed as the total call duration / the number of calls (connected). The caller-to-caller ratio is expressed as the number of outbound calls / (the number of outbound calls made by the outbound number + the number of outbound calls received by the outbound number).

[0033] In an embodiment of the present invention, before applying a preset illegal outbound call identification model based on multiple features to identify suspected illegal outbound call numbers, the illegal outbound call identification model is trained using existing data on illegal outbound call numbers that have been complained about, so as to select a suitable illegal outbound call identification model. Specifically, based on the existing data on illegal outbound call numbers that have been complained about, the illegal outbound call identification model is preprocessed and cleaned to form training data samples, resulting in the data sample format shown in Table 2. The illegal outbound call identification model training uses the Naive Bayesian classification algorithm, which is an algorithm that uses probabilistic statistical knowledge for classification. This Naive Bayesian classification algorithm can be applied to massive amounts of data, with high classification accuracy and speed.

[0034] Table 2 Illegal number data

[0035]

[0036] Naive Bayes classification is defined as follows:

[0037] (1) Let x = {a1, a2, a3, ..., a n} is an item to be classified, and each a is a characteristic attribute of x. Here, a1, a2, a3, ..., a n These are the 6 features obtained above.

[0038] (2) There is a category set c = {y1,y2,y3,…,y n}, calculate the prior probability p(y i ). Here, y1,y2,y3,…,y n Indicates that outbound call numbers can be divided into illegal outbound call numbers and non-illegal outbound call numbers, where 0 represents an illegal outbound call number and 1 represents a non-illegal outbound call number. Of course, in other embodiments of the present invention, 1 can also represent an illegal outbound call number and 0 can represent a non-illegal outbound call number.

[0039] (3) The conditional probability estimate p(a) for each feature attribute in each category i |y i ).

[0040] (4) Calculate p(y1|x), p(y2|x), ..., p(y n |x). According to Bayes' theorem, we can deduce as follows:

[0041] Since the denominator p(x) is constant for all categories, we only need to maximize the numerator. Since each feature attribute is conditionally independent, we have:

[0042]

[0043] (5) The maximum item in the prediction result is taken as the category of x max(p(x|y i )p(y i )). That is, the category of x is determined to be a category with a higher probability. If the probability that the calling number is an illegal outbound call number is higher in the embodiment of the present invention, the calling number is regarded as a suspected illegal outbound call number, otherwise, the calling number is determined to be a non-illegal outbound call number.

[0044] Step S12: Associating the called number and call time in the call signaling data with the user number and service processing time in the user service subscription data to obtain the suspected illegal agent and the suspected illegal outbound call number.

[0045] In an embodiment of the present invention, user service subscription data is collected from the Business & Operation Support System (BOSS) side, and a vector of suspected illegal outbound call numbers and user numbers and a vector of business processing work numbers and user numbers are constructed. The two vectors are measured for similarity, and agents with high similarity to illegal numbers are identified as illegal agents.

[0046] Specifically, a first relationship vector between the suspected illegal outbound call number and the user number and a second relationship vector between the business handling number and the user number are constructed based on the call signaling data and the user service subscription data, and the suspected illegal agent is identified based on the first relationship vector and the second relationship vector; based on the behavior of the called number in handling the business after the suspected illegal outbound call number makes an outbound call, the preset rules are applied to further determine the suspected illegal agent and the suspected illegal outbound number.

[0047] User service subscription data is obtained by collecting service subscription work order data from the BOSS system. The key data fields of the work order are shown in Table 3, including user number, service processing time, service processing work number, and service agency channel.

[0048] Table 3 User service subscription data

[0049] Serial number Field Name illustrate 1 User Number Mobile phone number for business operations 2 Business processing hours Business processing hours 3 Business Name Business Name 4 Business processing number The work number for users to handle business 5 Business agency channels Name of the agent who handles business for the user

[0050] In an embodiment of the present invention, the first relationship between the illegal outbound call number and the user number is obtained based on the call signaling data. For example, suspected illegal outbound call number 1: (customer number 1, customer number 2, customer number 4...), suspected illegal outbound call number 2: (customer number 1, customer number 3, customer number 5...), where the customer number is the user number. The second relationship between the business processing employee number and the user number is obtained based on the user service subscription data. For example, business processing employee number 1: (customer number 1, customer number 2, customer number 4...), business processing employee number 2: (customer number 4, customer number 5, customer number 7...). Then, the first relationship and the second relationship are associated based on the user number, and the first relationship vector and the second relationship vector are constructed and generated. The dimension generation rule of the vector is the union of the relationship between the illegal outbound call number and the user number and the relationship between the business processing employee number and the user number. For example, for the suspected illegal outbound call number N: (customer number 1, customer number 2, customer number 4) and the business handling number M: (customer number 1, customer number 2, customer number 3, customer number 5), the first relationship vector generated is the suspected illegal outbound call number N: (1, 1, 0, 1, 0), and the second relationship vector generated is the business handling number M: (1, 1, 1, 0, 1).

[0051] Assume that the first relationship vector is l i =(l1,l2,l3,…,l n ), the second relationship vector is m i =(m1,m2,m3,…,m n ), the first relationship vector l is calculated by applying the following relationship i =(l1,l2,l3,…,l n ) and the second relationship vector m i =(m1,m2,m3,…,m n )’s cosine similarity:

[0052]

[0053] The agent corresponding to the business ID whose cosine similarity is not less than a preset value is matched as the suspected illegal agent. The closer the cosine similarity value is to 1, the closer the similarity between the two vectors is, and the agent corresponding to the second vector is determined to be the suspected illegal agent. The preset value can be set according to actual circumstances.

[0054] In this embodiment of the present invention, the suspected illegal agent and suspected illegal outbound number are further determined based on the service handling behavior of the called number after the suspected illegal outbound number makes an outbound call. Specifically, the call signaling data is correlated with the user's service subscription data, that is, the "called number" in Table 1 is correlated with the "user number" in Table 3, and the "call time" in Table 1 is correlated with the "service handling time" in Table 3. The resulting correlated information data is shown in Table 4.

[0055] Table 4 Related information data

[0056]

[0057] As can be seen from Table 4, if the calling number (an outbound marketing number, such as 024667519XX) conducts an outbound marketing call, and the called number, after answering the outbound marketing call, conducts a service, the following rules are used to determine whether the user's service processing was caused by the outbound marketing. If the time interval between the call duration and the service processing time of the suspected illegal outbound call number is less than a first time interval; and within a preset period, the number of users who processed services within the first time interval after the suspected illegal outbound call number is not less than a first value, then the suspected illegal outbound call number and the suspected illegal agent who processed the service are further determined. For example, according to the data in Table 4, after the outbound call is connected, the user conducts a service within a first time interval (e.g., 30 minutes), that is, the service processing time minus the call duration <= 30 minutes, and within a preset period (e.g., 1 day), the number of users who processed services within the first time interval after the same outbound call number calls the user is greater than i (e.g., i>=10), then the outbound call number is further determined to be a suspected illegal outbound call number, and the agent who processed the service is a suspected illegal agent. The preset period, the first value and the first time interval can be set as needed and are not limited here. Preferably, the preset period is one day, the first time interval is 30 minutes, and the first value is 10.

[0058] In this way, through the above method, we can basically lock down the suspected illegal outbound call numbers used by the suspected illegal agents for outbound marketing, the work numbers used by the suspected illegal agents, the names of the suspected illegal agents and other information.

[0059] Step S13: confirming the illegal outbound call number and the illegal agent based on the first location information of the base station where the called number is located and the second location information of the suspected illegal agent.

[0060] In this embodiment of the present invention, first latitude and longitude information of the base station where the called number is located is obtained from the network; second latitude and longitude information of the suspected illegal agent is obtained from the BOSS system; the distance between the called number when the marketing was made and when the service was processed is calculated based on the first latitude and longitude information and the second latitude and longitude information; if the distance is greater than or equal to a preset distance, the suspected illegal outbound call number and the suspected illegal agent are respectively determined to be illegal outbound call number and illegal agent. The preset distance can be set as needed, preferably 2 kilometers.

[0061] Specifically, the agent location data shown in Table 5 on the access BOSS system side and the base station address data shown in Table 6 on the network side are associated to obtain the associated data shown in Table 7, which is used to calculate the call location of the called number and the distance from the location of the agent handling the business.

[0062] Table 5 Agent location data

[0063] Serial number Field Name illustrate 1 Channel Code Agent Code 2 Channel Name Agent Name 3 Channel longitude Agent Longitude 4 Channel Latitude Agent Latitude

[0064] Table 6 Base station address data

[0065] Serial number Field Name illustrate 1 Base station coding 2 Base station name 3 Base station lac 4 Base station ci 5 Base station longitude 6 Base station latitude

[0066] By associating the agent location data in Table 5 with the base station address data in Table 6, the base station location information when the called number is making a call and the agent location information when the service is being processed can be obtained.

[0067] Table 7 Related data

[0068]

[0069] Through the latitude and longitude information of the base station where the called number is located and the latitude and longitude information of the business processing channel of the called number, the distance j between the called number when it is marketed and when the business is processed can be calculated. By setting a preset distance for distance j (for example, j>2 kilometers), the judgment of the outbound call number, agent work number, and agent name can be further strengthened.

[0070] The complete illegal outbound marketing identification method of the embodiment of the present invention is as follows Figure 2 As shown, including:

[0071] Step S200: Acquire call signaling data.

[0072] Specifically, the call signaling data of 2G / 3G / 4G users is obtained from the network management system on the network side, and multiple features are selected based on the call signaling data.

[0073] Step S201: Apply a harassing outbound call recognition model to identify illegal outbound call numbers.

[0074] The harassing outbound call identification model used in this embodiment of the present invention is a Naive Bayesian classification algorithm. The Naive Bayesian classification algorithm is applied to perform a series of calculations based on the selected features to identify suspected illegal outbound call numbers. The specific calculation method is the same as that in the aforementioned step S11 and will not be repeated here.

[0075] Step S202: Determine whether the call meets the characteristics of illegal outbound calls. If yes, proceed to step S203; if not, terminate directly.

[0076] According to the calculation result in step S201, it is determined whether the corresponding calling number meets the illegal outbound call characteristics. If it meets the illegal outbound call characteristics, the process is directly terminated.

[0077] Step S203: Determine the number as a suspected illegal outbound call number.

[0078] If it is determined according to the calculation result in step S201 that the illegal outbound call feature is met, the corresponding calling number is determined to be a suspected illegal outbound call number.

[0079] Step S204: Obtain user service subscription data from the BOSS system.

[0080] Step S205: associate the called number with the service handling number, and the call time with the service handling time.

[0081] Based on call signaling data and user service subscription data, a vector is constructed between the suspected illegal outbound call number and the user number, and a vector between the service processing ID and the user number. A similarity measurement is performed on the two vectors, and agents with high similarity to the illegal number are identified as illegal agents. The suspected illegal agents and suspected illegal outbound call numbers are further determined based on the service processing behavior of the called number after the suspected illegal outbound call number makes an outbound call. For more details, see the method in step S12 above.

[0082] Step S206: Determine whether the outbound call number or agent is illegal. If yes, proceed to step S207; if not, terminate the process directly.

[0083] According to the correlation result in step S205, it is determined whether the outbound call number and the agent are in violation of the rules.

[0084] Step S207: Determine the suspected illegal outbound call number, the called number base station information, and the agent information.

[0085] The suspected illegal outbound call number, the called number base station information, and the agent information are determined according to the judgment result of step S206.

[0086] Step S208: Obtain the base station latitude and longitude information from the network management system.

[0087] Specifically, the latitude and longitude information of the base station where the suspected illegal number is located when making an outbound call, that is, the first location information, is obtained from the network management system.

[0088] Step S209: Obtain agent location information from the BOSS system.

[0089] Obtain the latitude and longitude information of the suspected illegal agent from the BOSS system, that is, the second location information.

[0090] Step S210: Calculate the distance between the user's call base station location information and the service agent's location.

[0091] Step S211: Determine whether the distance between the two is greater than a preset distance. If yes, proceed to step S212; if not, terminate directly.

[0092] Step S212: Confirm the illegal outbound call number and illegal agent.

[0093] The embodiment of the present invention combines key data from the network side and the BOSS side, obtains call signaling data from the network side, establishes an illegal outbound call identification model by setting key thresholds for each indicator rule, obtains suspected illegal outbound call phone numbers, and associates them with the BOSS side business processing work orders. A cluster analysis method is used to obtain information related to outbound call agents, and by calculating the distance between the location of the called number at the time of the call and the agent handling the business, the illegal outbound call number and agent are finally judged. The outbound call number and the agent involved can be identified quickly, promptly and accurately, so that management personnel can conduct timely management and control of outbound calls and agents, thereby reducing complaints caused by this.

[0094] The embodiment of the present invention obtains multiple features based on call signaling data, and applies a preset illegal outbound call identification model based on the multiple features to identify suspected illegal outbound call numbers; associates the called number and call time in the call signaling data with the user number and service processing time in the user service subscription data to obtain suspected illegal agents and the suspected illegal outbound call number; confirms the illegal outbound call number and the suspected illegal agent based on the base station location information of the called number when the call is made and the location information of the suspected illegal agent, so as to quickly, promptly and accurately identify the outbound call number and the agent involved, so that management personnel can conduct timely management and control of outbound calls and agents, thereby reducing complaints caused by this.

[0095] Figure 3 The following is a schematic diagram showing the structure of the device for identifying illegal outbound marketing according to an embodiment of the present invention. Figure 3 As shown, the illegal outbound marketing identification device includes: a first identification unit 301, a second identification unit 302 and a violation determination unit 303.

[0096] The first identification unit 301 is used to obtain multiple features based on call signaling data, and apply a preset illegal outbound call identification model based on the multiple features to identify suspected illegal outbound call numbers; the second identification unit 302 is used to associate the called number and call time in the call signaling data with the user number and service processing time in the user service subscription data to obtain suspected illegal agents and the suspected illegal outbound call number; the violation determination unit 303 is used to confirm the illegal outbound call number and the suspected illegal agent based on the base station location information of the called number when the call was made and the location information of the suspected illegal agent.

[0097] In an optional manner, the first identification unit 301 is used to: select multiple features based on the call signaling data in combination with the illegal outbound call business scenario, and calculate the feature values ​​of multiple features, the multiple features including: average number of calls per day, call dispersion, answer rate, caller-to-caller hang-up ratio, average call duration and caller-to-call ratio; apply the naive Bayes classification algorithm based on multiple features to identify the suspected illegal outbound call number.

[0098] In an optional manner, the second identification unit 302 is used to: construct a first relationship vector between the suspected illegal outbound call number and the user number and a second relationship vector between the business handling number and the user number based on the call signaling data and the user service subscription data, and identify the suspected illegal agent based on the first relationship vector and the second relationship vector; and further determine the suspected illegal agent and the suspected illegal outbound number based on the behavior of the called number in handling the business after the suspected illegal outbound call number makes an outbound call by applying preset rules.

[0099] In an optional manner, the second identification unit 302 is used to: obtain a first relationship between the illegal outbound call number and the user number based on the call signaling data; obtain a second relationship between the business processing number and the user number based on the user service subscription data; associate the first relationship and the second relationship based on the user number, and construct and generate the first relationship vector and the second relationship vector.

[0100] In an optional manner, the second recognition unit 302 is configured to calculate the first relationship vector l by applying the following relational expression: i =(l1,l2,l3,…,l n ) and the second relationship vector m i =(m1,m2,m3,…,m n )’s cosine similarity:

[0101]

[0102] The agent corresponding to the business handling number whose cosine similarity is not less than a preset value is matched as the suspected illegal agent.

[0103] In an optional manner, the second identification unit 302 is used for: if the time interval between the call time of the suspected illegal outbound call number and the service processing time is less than the first time interval; and within a preset period, the number of users who handle services within the first time interval after the suspected illegal outbound call number makes an outbound call within the preset period is not less than a first value, then the suspected illegal outbound call number and the suspected illegal agent who handles the service are further determined.

[0104] In an optional manner, the violation determination unit 303 is used to: obtain the first latitude and longitude information of the base station where the called number is located from the network side; obtain the second latitude and longitude information of the suspected illegal agent who handles the business from the BOSS system side; calculate the distance between the called number when it was marketed and when the business was handled based on the first latitude and longitude information and the second latitude and longitude information; if the distance is greater than or equal to the preset distance, the suspected illegal outbound number and the suspected illegal agent are respectively determined as an illegal outbound number and an illegal agent.

[0105] The embodiment of the present invention obtains multiple features based on call signaling data, and applies a preset illegal outbound call identification model based on the multiple features to identify suspected illegal outbound call numbers; associates the called number and call time in the call signaling data with the user number and service processing time in the user service subscription data to obtain suspected illegal agents and the suspected illegal outbound call number; confirms the illegal outbound call number and the suspected illegal agent based on the base station location information of the called number when the call is made and the location information of the suspected illegal agent, so as to quickly, promptly and accurately identify the outbound call number and the agent involved, so that management personnel can conduct timely management and control of outbound calls and agents, thereby reducing complaints caused by this.

[0106] An embodiment of the present invention provides a non-volatile computer storage medium, which stores at least one executable instruction. The computer executable instruction can execute the illegal outbound marketing identification method in any of the above method embodiments.

[0107] The executable instructions can be used to cause the processor to perform the following operations:

[0108] Acquire multiple features based on call signaling data, and apply a preset illegal outbound call identification model based on the multiple features to identify suspected illegal outbound call numbers;

[0109] Associating the called number and call time in the call signaling data with the user number and service processing time in the user service subscription data to obtain the suspected illegal agent and the suspected illegal outbound call number;

[0110] The illegal outbound call number and the illegal agent are confirmed based on the first location information of the base station where the called number is located and the second location information of the suspected illegal agent.

[0111] In an optional manner, the executable instruction causes the processor to perform the following operations:

[0112] Based on the call signaling data and the illegal outbound call business scenario, a plurality of features are selected and characteristic values ​​of the plurality of features are calculated, wherein the plurality of features include: average number of calls per day, call dispersion, answer rate, caller-to-caller hang-up ratio, average call duration, and caller-to-call ratio;

[0113] A naive Bayesian classification algorithm is applied based on the multiple features to identify the suspected illegal outbound call number.

[0114] In an optional manner, the executable instruction causes the processor to perform the following operations:

[0115] Constructing a first relationship vector between the suspected illegal outbound call number and the user number and a second relationship vector between the service handling number and the user number based on the call signaling data and the user service subscription data, and identifying the suspected illegal agent based on the first relationship vector and the second relationship vector;

[0116] According to the behavior of the called number in handling services after the suspected illegal outbound call number makes an outbound call, the preset rules are applied to further determine the suspected illegal agent and the suspected illegal outbound call number.

[0117] In an optional manner, the executable instruction causes the processor to perform the following operations:

[0118] Acquire a first relationship between the illegal outbound call number and the user number according to the call signaling data;

[0119] Acquire a second relationship between the service handling ID and the user number based on the user service subscription data;

[0120] The first relationship and the second relationship are associated according to the user number, and the first relationship vector and the second relationship vector are constructed and generated.

[0121] In an optional manner, the executable instruction causes the processor to perform the following operations:

[0122] Apply the following relationship to calculate the first relationship vector l i =(l1,l2,l3,…,l n ) and the second relationship vector m i =(m1,m2,m3,…,m n )’s cosine similarity:

[0123]

[0124] The agent corresponding to the business handling number whose cosine similarity is not less than a preset value is matched as the suspected illegal agent.

[0125] In an optional manner, the executable instruction causes the processor to perform the following operations:

[0126] If the time interval between the call time of the outbound call of the suspected illegal outbound call number and the service processing time is less than the first time interval; and within the preset period, the number of users who handle services within the first time interval after the outbound call of the suspected illegal outbound call number within the preset period is not less than the first value, the suspected illegal outbound call number and the suspected illegal agent who handles the service are further determined.

[0127] In an optional manner, the executable instruction causes the processor to perform the following operations:

[0128] The first latitude and longitude information of the base station where the called number is located obtained from the network side;

[0129] The second latitude and longitude information of the suspected illegal agent obtained from the BOSS system;

[0130] Calculate the distance between the called number when the marketing is carried out and when the service is handled based on the first longitude and latitude information and the second longitude and latitude information;

[0131] If the distance is greater than or equal to the preset distance, the suspected illegal outbound call number and the suspected illegal agent are determined as an illegal outbound call number and an illegal agent respectively.

[0132] The embodiment of the present invention obtains multiple features based on call signaling data, and applies a preset illegal outbound call identification model based on the multiple features to identify suspected illegal outbound call numbers; associates the called number and call time in the call signaling data with the user number and service processing time in the user service subscription data to obtain suspected illegal agents and the suspected illegal outbound call number; confirms the illegal outbound call number and the suspected illegal agent based on the base station location information of the called number when the call is made and the location information of the suspected illegal agent, so as to quickly, promptly and accurately identify the outbound call number and the agent involved, so that management personnel can conduct timely management and control of outbound calls and agents, thereby reducing complaints caused by this.

[0133] An embodiment of the present invention provides a computer program product, which includes a computer program stored on a computer storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the illegal outbound marketing identification method in any of the above-mentioned method embodiments.

[0134] Figure 4 The schematic diagram of the structure of the computing device provided by the embodiment of the present invention is shown. The specific embodiment of the present invention does not limit the specific implementation of the device.

[0135] like Figure 4 As shown, the computing device may include: a processor 402 , a communications interface 404 , a memory 406 , and a communication bus 408 .

[0136] Processor 402, communication interface 404, and memory 406 communicate with each other via communication bus 408. Communication interface 404 is used to communicate with other devices, such as clients or other server network elements. Processor 402 is used to execute program 410, which may specifically perform the steps described in the aforementioned embodiment of the method for identifying illegal outbound marketing calls.

[0137] Specifically, the program 410 may include program codes, which include computer operation instructions.

[0138] Processor 402 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement an embodiment of the present invention. The one or more processors included in the device may be processors of the same type, such as one or more CPUs, or may be processors of different types, such as one or more CPUs and one or more ASICs.

[0139] The memory 406 is used to store the program 410. The memory 406 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0140] The program 410 may be specifically configured to cause the processor 402 to perform the following operations:

[0141] Acquire multiple features based on call signaling data, and apply a preset illegal outbound call identification model based on the multiple features to identify suspected illegal outbound call numbers;

[0142] Associating the called number and call time in the call signaling data with the user number and service processing time in the user service subscription data to obtain the suspected illegal agent and the suspected illegal outbound call number;

[0143] The illegal outbound call number and the illegal agent are confirmed based on the first location information of the base station where the called number is located and the second location information of the suspected illegal agent.

[0144] In an optional manner, the program 410 enables the processor to perform the following operations:

[0145] Based on the call signaling data and the illegal outbound call business scenario, a plurality of features are selected and characteristic values ​​of the plurality of features are calculated, wherein the plurality of features include: average number of calls per day, call dispersion, answer rate, caller-to-caller hang-up ratio, average call duration, and caller-to-call ratio;

[0146] A naive Bayesian classification algorithm is applied based on the multiple features to identify the suspected illegal outbound call number.

[0147] In an optional manner, the program 410 enables the processor to perform the following operations:

[0148] Constructing a first relationship vector between the suspected illegal outbound call number and the user number and a second relationship vector between the service handling number and the user number based on the call signaling data and the user service subscription data, and identifying the suspected illegal agent based on the first relationship vector and the second relationship vector;

[0149] According to the behavior of the called number in handling services after the suspected illegal outbound call number makes an outbound call, the preset rules are applied to further determine the suspected illegal agent and the suspected illegal outbound call number.

[0150] In an optional manner, the program 410 enables the processor to perform the following operations:

[0151] Acquire a first relationship between the illegal outbound call number and the user number according to the call signaling data;

[0152] Acquire a second relationship between the service handling ID and the user number based on the user service subscription data;

[0153] The first relationship and the second relationship are associated according to the user number, and the first relationship vector and the second relationship vector are constructed and generated.

[0154] In an optional manner, the program 410 enables the processor to perform the following operations:

[0155] Apply the following relationship to calculate the first relationship vector l i =(l1,l2,l3,…,l n ) and the second relationship vector m i =(m1,m2,m3,…,m n )’s cosine similarity:

[0156]

[0157] The agent corresponding to the business handling number whose cosine similarity is not less than a preset value is matched as the suspected illegal agent.

[0158] In an optional manner, the program 410 enables the processor to perform the following operations:

[0159] If the time interval between the call time of the outbound call of the suspected illegal outbound call number and the service processing time is less than the first time interval; and within the preset period, the number of users who handle services within the first time interval after the outbound call of the suspected illegal outbound call number within the preset period is not less than the first value, the suspected illegal outbound call number and the suspected illegal agent who handles the service are further determined.

[0160] In an optional manner, the program 410 enables the processor to perform the following operations:

[0161] The first latitude and longitude information of the base station where the called number is located obtained from the network side;

[0162] The second latitude and longitude information of the suspected illegal agent obtained from the BOSS system;

[0163] Calculate the distance between the called number when the marketing is carried out and when the service is handled based on the first longitude and latitude information and the second longitude and latitude information;

[0164] If the distance is greater than or equal to the preset distance, the suspected illegal outbound call number and the suspected illegal agent are determined as an illegal outbound call number and an illegal agent respectively.

[0165] The embodiment of the present invention obtains multiple features based on call signaling data, and applies a preset illegal outbound call identification model based on the multiple features to identify suspected illegal outbound call numbers; associates the called number and call time in the call signaling data with the user number and service processing time in the user service subscription data to obtain suspected illegal agents and the suspected illegal outbound call number; confirms the illegal outbound call number and the suspected illegal agent based on the base station location information of the called number when the call is made and the location information of the suspected illegal agent, so as to quickly, promptly and accurately identify the outbound call number and the agent involved, so that management personnel can conduct timely management and control of outbound calls and agents, thereby reducing complaints caused by this.

[0166] The algorithm or demonstration provided herein are not inherently relevant to any particular computer, virtual system or other equipment. Various general-purpose systems may also be used together with the teachings based on this. According to the above description, it is apparent that the structure required for constructing this type of system. In addition, the embodiment of the present invention is not directed to any specific programming language yet. It should be understood that various programming languages ​​can be utilized to realize the content of the present invention described herein, and the above description of specific languages ​​is for the purpose of disclosing the best mode of the present invention.

[0167] In the description provided herein, numerous specific details are described. However, it is understood that embodiments of the present invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques are not shown in detail so as not to obscure the understanding of this description.

[0168] Similarly, it should be understood that in order to streamline the present invention and facilitate understanding of one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, various features of the embodiments of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this method of disclosure should not be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim.

[0169] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and in addition may be divided into multiple submodules or subunits or subcomponents. All features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed herein may be combined in any combination, except that at least some of such features and / or processes or units are mutually exclusive. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0170] It should be noted that the above embodiments illustrate rather than limit the invention, and that alternative embodiments may be devised by a person skilled in the art without departing from the scope of the appended claims. In the claims, any reference signs placed between brackets should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising several different elements and by means of appropriately programmed computers. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third etc. does not indicate any order. These words may be interpreted as names. The steps in the above embodiments should not be understood as limiting the order of execution unless otherwise specified.

Claims

1. A method for identifying illegal outbound marketing, characterized in that: The method comprises: Acquire multiple features based on call signaling data, and apply a preset illegal outbound call identification model based on the multiple features to identify suspected illegal outbound call numbers; A first relationship vector between the suspected illegal outbound call number and the user number and a second relationship vector between the service handling ID and the user number are constructed based on the call signaling data and the user service subscription data, and a suspected illegal agent is identified based on the first relationship vector and the second relationship vector; and a preset rule is applied based on the behavior of the called number handling services after the suspected illegal outbound call number makes an outbound call and further determines the suspected illegal agent and the suspected illegal outbound call number; The illegal outbound call number and the illegal agent are confirmed based on the first location information of the base station where the called number is located and the second location information of the suspected illegal agent.

2. The method according to claim 1, characterized in that The method of obtaining multiple features based on the call signaling data and applying a preset illegal outbound call identification model based on the multiple features to identify suspected illegal outbound call numbers includes: Based on the call signaling data and the illegal outbound call business scenario, a plurality of features are selected and characteristic values ​​of the plurality of features are calculated, wherein the plurality of features include: average number of calls per day, call dispersion, answer rate, caller-to-caller hang-up ratio, average call duration, and caller-to-call ratio; A naive Bayesian classification algorithm is applied based on the multiple features to identify the suspected illegal outbound call number.

3. The method according to claim 1, characterized in that The step of constructing a first relationship vector between the suspected illegal outbound call number and the user number and a second relationship vector between the service handling number and the user number based on the call signaling data and the user service subscription data includes: Acquire a first relationship between the illegal outbound call number and the user number according to the call signaling data; Acquire a second relationship between the service handling ID and the user number based on the user service subscription data; The first relationship and the second relationship are associated according to the user number, and the first relationship vector and the second relationship vector are constructed and generated.

4. The method according to claim 1, wherein The step of identifying a suspected illegal agent based on the first relationship vector and the second relationship vector includes: Apply the following relationship to calculate the first relationship vector l i =(l1,l2,l3,…,l n ) and the second relationship vector m i =(m1,m2,m3,…,m n )’s cosine similarity: The agent corresponding to the business handling number whose cosine similarity is not less than a preset value is matched as the suspected illegal agent.

5. The method according to claim 1, wherein The method further determines the suspected illegal agent and the suspected illegal outbound number based on the behavior of the called number in handling services after the call from the suspected illegal outbound number, using preset rules, including: If the time interval between the call time of the outbound call of the suspected illegal outbound call number and the service processing time is less than the first time interval; and within the preset period, the number of users who handle services within the first time interval after the outbound call of the suspected illegal outbound call number within the preset period is not less than the first value, the suspected illegal outbound call number and the suspected illegal agent who handles the service are further determined.

6. The method according to claim 1, characterized in that The step of confirming the illegal outbound call number and the illegal agent based on the first location information of the base station where the called number is located and the second location information of the suspected illegal agent includes: The first latitude and longitude information of the base station where the called number is located obtained from the network side; The second latitude and longitude information of the suspected illegal agent obtained from the BOSS system; Calculate the distance between the called number when the marketing is carried out and when the service is handled based on the first longitude and latitude information and the second longitude and latitude information; If the distance is greater than or equal to the preset distance, the suspected illegal outbound call number and the suspected illegal agent are determined as an illegal outbound call number and an illegal agent respectively.

7. A device for identifying illegal outbound marketing, characterized in that: The device comprises: A first identification unit is configured to obtain a plurality of features based on the call signaling data, and identify suspected illegal outbound call numbers using a preset illegal outbound call identification model based on the plurality of features; A second identification unit is configured to construct a first relationship vector between the suspected illegal outbound call number and the user number and a second relationship vector between the service handling number and the user number based on the call signaling data and the user service subscription data, and identify the suspected illegal agent based on the first relationship vector and the second relationship vector; and further determine the suspected illegal agent and the suspected illegal outbound number based on the behavior of the called number handling the service after the suspected illegal outbound call number makes an outbound call by applying a preset rule; The violation determination unit is used to confirm the illegal outbound call number and the illegal agent according to the base station location information of the suspected illegal outbound call number and the suspected illegal agent when the called number is used for communication and the location information of the suspected illegal agent.

8. A computing device comprising: A processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute the steps of the illegal outbound marketing identification method according to any one of claims 1-6.

9. A computer storage medium, wherein the storage medium stores at least one executable instruction, wherein the executable instruction enables a processor to execute the steps of the method for identifying illegal outbound marketing according to any one of claims 1-6.

Images