Method and device for identifying one-person multi-card user

Abstract

The invention relates to a method and device for identifying a one-person multi-card user, and the method comprises the steps: obtaining the fingerprint data of a user number according to a preset fingerprint variable, obtaining the analysis time window of the fingerprint variable according to the operator attribute, and obtaining the analysis time window of the user number according to the analysis time window; therefore, the fingerprint data under the analysis time window are combined to obtain unique fingerprint information of user numbers corresponding to different operator attributes, the unique fingerprint information is compared pairwise, the overlap ratio is calculated, and a one-person multi-card user is obtained according to the overlap ratio. Therefore, when the unique fingerprint information of the user number is constructed, not only is multi-dimensional comprehensive consideration carried out in space, but also the unique fingerprint information is constructed according to the operator attributes in time, so that the integrity of the constructed unique fingerprint information is ensured, and the user number has corresponding unique fingerprint information under different operator attributes; and the accuracy of the unique fingerprint information is improved, so that the accuracy of one-person multi-card user identification is improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method and device for identifying a user with multiple cards. Background Art

[0002] When one person applies for multiple card numbers under the same or different telecommunications operators, it is called multiple cards. Traditional methods of identifying multiple cards mainly include identification based on social circles and identification based on location data. Although traditional methods can identify multiple cards, there is a problem of inaccurate identification. Summary of the invention

[0003] The technical problem to be solved by the present invention is: the present invention provides a method and device for identifying a user with multiple cards, thereby improving the accuracy of identification.

[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0005] In a first aspect, the present invention provides a method for identifying a user with multiple cards, comprising:

[0006] Acquire fingerprint data of the user number according to preset fingerprint variables, wherein the fingerprint data includes operator attributes, social circle data and location data;

[0007] Acquire an analysis time window of the fingerprint variable according to the operator attribute, combine the fingerprint data under the analysis time window according to a preset construction rule, obtain unique fingerprint information of user numbers corresponding to different operator attributes, and aggregate the unique fingerprint information of all user numbers to obtain a fingerprint database;

[0008] The unique fingerprint information in the fingerprint database is compared in pairs in sequence, the overlap degree is calculated, and the multi-card user is obtained according to the overlap degree.

[0009] The beneficial effects of the present invention are as follows: by acquiring the operator attributes, social circle data and location data of the user number, the unique fingerprint information of the user number is constructed, so that the fingerprint information of the user number is comprehensively considered in multiple dimensions in space, and at the same time, the analysis time window of the fingerprint variable is obtained in time according to the operator attributes, and the fingerprint data under the analysis time window is combined to construct the unique fingerprint information of the user number corresponding to different operator attributes. When constructing the unique fingerprint information, it is constructed based on the fingerprint data. The comprehensiveness of the fingerprint data ensures the integrity of the constructed unique fingerprint information, and the user number has corresponding unique fingerprint information under different operator attributes, which improves the accuracy of the unique fingerprint information, thereby ensuring the accuracy of the overlap obtained when the unique fingerprint information is compared two by two in sequence, that is, the accuracy of identifying users with multiple cards is improved.

[0010] Optionally, acquiring the analysis time window of the fingerprint variable according to the operator attribute, combining the fingerprint data under the analysis time window according to a preset construction rule, and obtaining unique fingerprint information of the user number corresponding to different operator attributes includes:

[0011] When the operator attribute is itself, the fingerprint data includes the user account opening certificate number, terminal data and content preference data, and the first contact number, the out-of-province number, the out-of-province number, the first family number and the first colleague number within the first preset time period are obtained from the social circle data; the historical terminal data within the second preset time period and the commonly used terminal data within the third preset time period are obtained from the terminal data; the first working base station, the first residential base station, the first social circle location, the roaming location within the fourth preset time period and the signaling track within the fifth preset time period are obtained from the location data; the APP usage preference data and the traffic content preference data within the sixth preset time period are obtained from the content preference data;

[0012] Combine the user's account opening certificate number, first contact number, in-province number, out-of-province number, first family number, first colleague number, historical terminal data, frequently used terminal data, first work base station, first residential base station, first contact circle location, roaming location, signaling trace, APP usage preference data and traffic content preference data to obtain the unique fingerprint information of the user number when the operator attribute is itself;

[0013] When the operator attribute is a competitor, the second contact number, the second family number and the second colleague number within the seventh preset time period are obtained from the social circle data, the second work base station, the second residence base station and the second social circle location are obtained from the location data, and the second contact number, the second family number, the second colleague number, the second work base station, the second residence base station and the second social circle location are combined to obtain unique fingerprint information of the user number when the operator attribute is a competitor.

[0014] According to the above description, the operator attributes are divided into itself and competitors, so that different methods are used according to different operator attributes to construct the unique fingerprint information of the user number, ensuring the rationality and diversity of the constructed unique fingerprint information. When the operator attribute is itself, the user's account opening document number, terminal data and content preference data are taken into account to improve the accuracy and comprehensiveness of the unique fingerprint information.

[0015] Optionally, the acquiring a first contact number within a first preset time period from the circle of contacts data comprises:

[0016] Obtaining all first contact numbers ranked in the top N places and all first contact numbers ranked in the top M places for nighttime calls within a first preset time period from the contact circle data, and filtering out all first contact numbers of stable contacts from the contact circle data according to the filtering rule;

[0017] The step of acquiring the second contact number within the seventh preset time period from the circle of contacts data includes:

[0018] Obtaining the first second contact number ranked in the top 1, all second contact numbers ranked in the top Q, and all second contact numbers ranked in the top P for night calls from the social circle data within the seventh preset time period, and filtering out all second contact numbers of stable contacts from the social circle data according to the screening rule;

[0019] The screening rule is that the contact time within the eighth preset time period is not less than the first threshold.

[0020] According to the above description, when the operator attributes are different, the rules for obtaining the contact numbers from the contact circle data are different, which ensures the rationality of the obtained contact numbers. At the same time, the diversity of the acquisition rules also improves the accuracy and completeness of the obtained contact numbers.

[0021] Optionally, the unique fingerprint information is dynamically updated as the fingerprint data changes.

[0022] According to the above description, the unique fingerprint information is not static. As the fingerprint data changes, the unique fingerprint information will be dynamically updated to ensure the real-time nature of the unique fingerprint information.

[0023] Optionally, the step of sequentially comparing the unique fingerprint information in the fingerprint database in pairs, calculating the overlap degree, and obtaining the multi-card user according to the overlap degree includes:

[0024] Classifying the unique fingerprint information in the fingerprint database according to the type of the operator attribute to obtain classified unique fingerprint information;

[0025] The classified unique fingerprint information is cross-combined to form D different types of comparison scenarios. In each different type of comparison scenario and each same type of comparison scenario, the unique fingerprint information in the fingerprint database is respectively compared two by two in sequence, and the overlap degree is calculated. According to the overlap degree, the multi-card user is obtained.

[0026] According to the above description, when identifying one person with multiple cards, not only the unique fingerprint information under the same operator attribute is compared, but also the unique fingerprint information classified according to the type of operator attribute is cross-combined to form D different types of comparison scenarios. One person with multiple cards is identified in each different type of comparison scenario and each same type of comparison scenario. The diversified comparison scenarios improve the comprehensiveness and accuracy of one person with multiple cards identification.

[0027] Optionally, in each of the different types of comparison scenarios and each of the same type of comparison scenarios, the unique fingerprint information in the fingerprint database is respectively compared in pairs, and the overlap degree is calculated, and the multi-card user is obtained according to the overlap degree, including:

[0028] When calculating the coincidence degree, the corresponding method is used for calculation according to the type of the fingerprint variable. When the type of the fingerprint variable is attribute type, the coincidence degree is calculated by the precise comparison method. When the type of the fingerprint variable is sequence type, the coincidence degree is calculated by the overlap comparison method. When the type of the fingerprint variable is position type, the coincidence degree is calculated by the distance comparison method.

[0029] In the same type of comparison scenario, the first conversion rule is used to convert the overlap degree into an overlap score value. In different types of comparison scenarios, the second conversion rule is used to convert the overlap degree into an overlap score value, and a one-person multi-card user is obtained according to the overlap score value.

[0030] According to the above description, when calculating the overlap by pairwise comparison of unique fingerprint information, a diversified calculation method is adopted according to the type of fingerprint variable, breaking the single calculation method, improving the adaptability and accuracy of the overlap calculation, and the conversion rules for converting the overlap into the overlap score value are also different for the same comparison scenario and different comparison scenarios, ensuring the rationality of the conversion, and then ensuring the accuracy of the identification of one person with multiple cards.

[0031] Optionally, when the type of the fingerprint variable is a sequence type, calculating the coincidence degree by using an overlap comparison method includes:

[0032] When the sequence class is social circle data, in the same type of comparison scenario, the social circle data is directly overlapped and the calculated first overlap is used as the coincidence degree. In different types of comparison scenarios, if the operator attributes are respectively itself and competitors, candidate numbers are screened out from the user numbers of the operator being itself and the operator being competitors according to the first overlap rule, and then the overlap comparison method is used to calculate the second overlap between the unique fingerprint information corresponding to the candidate numbers, and the second overlap is used as the coincidence degree. If the operator attributes are all competitors, suspected numbers are screened out from the user numbers of the operator being competitors according to the second overlap rule, and then the overlap comparison method is used to calculate the third overlap between the unique fingerprint information corresponding to the suspected numbers, and the third overlap is used as the coincidence degree.

[0033] According to the above description, when calculating the overlap of social circle data, not only are the calculation methods for the same type of comparison scenarios and different types of comparison scenarios different, but the different types of comparison scenarios are further divided into comparisons between oneself and competitors, and comparisons between competitors. Multiple overlap rules are used to screen out candidate numbers and suspected numbers, and then the overlapping comparison method is used to calculate the second overlap between the unique fingerprint information corresponding to the candidate numbers, and the third overlap between the unique fingerprint information corresponding to the suspected numbers, thereby improving the calculation efficiency and the accuracy of the calculated overlap.

[0034] Optionally, obtaining a user with multiple cards according to the overlap score includes:

[0035] In the same type comparison scenario, if the overlap score value exceeds the first score threshold, and the user numbers of the two unique fingerprint information corresponding to the first score threshold are both comprehensive cards, then the user numbers of the two unique fingerprint information are the same user, and a multi-card user is obtained, and the comprehensive card is a voice card and a traffic card;

[0036] Or, in the same type comparison scenario, if the overlap score value exceeds the second score threshold, and the user numbers of the two unique fingerprint information corresponding to the second score threshold are a voice card and a traffic card respectively, then the user numbers of the two unique fingerprint information are the same user, and a multi-card user is obtained;

[0037] In different types of comparison scenarios, if the overlap score value exceeds the third score threshold, the user numbers of the two unique fingerprint information are the same user, and a one-person-multiple-card user is obtained.

[0038] According to the above description, when facing the same type of comparison scenario, the user with multiple cards is not identified simply based on the overlap score value. Instead, the overlap score value and whether the user number is a comprehensive card are taken into consideration, and the two judgment situations are further divided to improve the accuracy of identifying users with multiple cards in the same type of comparison scenario.

[0039] Optionally, obtaining a user with multiple cards according to the overlap degree further includes:

[0040] Obtaining the usage of all users' cards, classifying the cards according to the usage, and obtaining corresponding grade information;

[0041] Calculate the usage ratio of each card of the same user in the usage ratio of all cards, use the usage ratio as the user's network age contribution, and calculate the consumption amount, call minutes and data traffic value of each card of the same user within a preset period;

[0042] A multi-dimensional evaluation index is established from the grade information, the network age contribution, the consumption amount, the number of call minutes and the data traffic value through the hierarchical analysis method, and the importance score of each card corresponding to each user is calculated. The card with the highest importance score is screened out according to the importance score. If there are multiple cards with the highest importance score, they are sorted in descending order according to the consumption amount or the number of call minutes or the data traffic value to obtain the cards in descending order. The card with the highest importance score that is ranked first is used as the main card. If there are not multiple cards with the highest importance scores, the card with the highest importance score is directly used as the main card.

[0043] According to the above description, after identifying multiple users, the main card will be identified. When identifying the main card, a hierarchical analysis method is used to establish multi-dimensional evaluation indicators from grade information, network age contribution, consumption amount, call minutes and data traffic value. In this way, the importance score of each card is calculated for main card identification to improve the accuracy of main card identification.

[0044] In a second aspect, the present invention provides a device for identifying a user with multiple cards, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for identifying a user with multiple cards as described in the first aspect is implemented.

[0045] Among them, the technical effect corresponding to the device for identifying one person with multiple cards provided in the second aspect refers to the relevant description of the method for identifying one person with multiple cards provided in the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 A flow chart of a method for identifying a user with multiple cards provided in this embodiment;

[0047] Figure 2 A schematic diagram of the overall flow of a method for identifying a user with multiple cards provided in this embodiment;

[0048] Figure 3 A schematic diagram of the structure of a device for identifying a multi-card user provided in this embodiment.

[0049] [Description of Reference Numerals]

[0050] 1. A device for identifying a user with multiple cards;

[0051] 2. Processor;

[0052] 3. Memory. DETAILED DESCRIPTION

[0053] In order to better understand the above technical solution, 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 described herein. On the contrary, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0054] Embodiment 1

[0055] Please refer to Figure 1 to Figure 2 The present invention provides a method for identifying a user with multiple cards, comprising the steps of:

[0056] S1. Obtaining fingerprint data of a user number according to a preset fingerprint variable, wherein the fingerprint data includes operator attributes, social circle data and location data;

[0057] In this embodiment, if Figure 2 As shown, the fingerprint data of the user number is obtained according to the preset fingerprint variables, wherein the fingerprint data generally includes: operator attributes, social circle data and location data, and the operator attributes are divided into the operator itself and competitors from the perspective of the operator. For example, there are currently three major operators, namely China Telecom, China Mobile and China Unicom. When the operator is China Telecom, its operator attribute is itself, while the operator attributes of China Mobile and China Unicom are competitors. When the operator attribute is itself, its fingerprint data also includes the user's account opening certificate number, terminal data and content preference data. Of course, if the operator data is a competitor, the user's account opening certificate number, terminal data and content preference data will be obtained together if they can be obtained through legal channels.

[0058] S2. Obtain an analysis time window of the fingerprint variable according to the operator attribute, combine the fingerprint data under the analysis time window according to a preset construction rule, obtain unique fingerprint information of user numbers corresponding to different operator attributes, and aggregate the unique fingerprint information of all user numbers to obtain a fingerprint database;

[0059] In this embodiment, if Figure 2 As shown, the analysis time window of each fingerprint variable is set according to the operator attributes, so that the fingerprint data under the analysis time window is combined according to the preset construction rules, that is, when constructing unique fingerprint information, the fingerprint data is screened in terms of time, so as to construct unique fingerprint information of user numbers corresponding to different operator attributes, so as to aggregate the unique fingerprint information of all user numbers to obtain a fingerprint database, wherein the unique fingerprint information is dynamically updated as the fingerprint data changes, and the dynamic update rules can be based on a certain period, such as dynamic update every month, three months, half a year, etc., or can be dynamic update every day, and the specific rules can be set according to actual conditions.

[0060] At this time, the analysis time window of the fingerprint variable is obtained according to the operator attribute in step S2, and the fingerprint data under the analysis time window is combined according to the preset construction rule to obtain the unique fingerprint information of the user number corresponding to different operator attributes, including:

[0061] S21. When the operator attribute is itself, the fingerprint data includes the user account opening certificate number, terminal data and content preference data; the first contact number, the non-local number within the province, the number outside the province, the first family number and the first colleague number within the first preset time period are obtained from the contact circle data; the historical terminal data within the second preset time period and the frequently used terminal data within the third preset time period are obtained from the terminal data; the first working base station, the first residential base station, the first contact circle location, the roaming location within the fourth preset time period and the signaling trajectory within the fifth preset time period are obtained from the location data; and the APP usage preference data and the traffic content preference data within the sixth preset time period are obtained from the content preference data;

[0062] In this embodiment, when the operator attribute is itself, the first contact number, the non-local number in the province, the number outside the province, the first family number and the first colleague number within the first preset time period are obtained from the contact circle data, wherein the first preset time period is nearly 6 months, and the historical terminal data within the second preset time period and the commonly used terminal data within the third preset time period are obtained from the terminal data, wherein the second preset time period is nearly one year, and the third preset time period is nearly 31 days, and the obtained historical terminal data and commonly used terminal data exclude the copycat phone imei (International Mobile EquipmentIdentity, International Mobile Equipment Identity), obtain the first working base station, the first residential base station, the first social circle base station, the roaming position within the fourth preset time period and the signaling trajectory within the fifth preset time period from the location data, wherein the first social circle base station refers to the city of the user number, the fourth preset time period is the past 31 days, and the fifth preset time period is the past 31 days, and for the obtained signaling trajectory within the fifth preset time period, that is, the signaling trajectory of the past 31 days, the base station with the longest stay in each time period will be screened, and the screened longest base station will be used as the final signaling trajectory, and the APP usage preference data and traffic content preference data within the sixth preset time period are obtained from the content preference data, wherein the sixth preset time period is the past 31 days, the APP usage preference data refers to the top 10 APPs in APP usage in the past 31 days, and the traffic content preference data refers to the top 10 data in traffic usage in the past 31 days, wherein the first preset time period, the second preset time period, the third preset time period, the fourth preset time period, the fifth preset time period and the sixth preset time period can be adjusted and set according to actual conditions.

[0063] At this time, the step S21 of acquiring the first contact number within the first preset time period from the contact circle data includes:

[0064] S211, obtaining all the first contact numbers ranked in the top N places and all the first contact numbers ranked in the top M places for nighttime calls within a first preset time period from the social circle data, and filtering out all the first contact numbers of stable contacts from the social circle data according to a filtering rule;

[0065] In this embodiment, if Figure 2As shown, the first contact numbers within the first preset time period obtained from the social circle data are further divided to obtain all the first contact numbers ranked in the top N and all the first contact numbers ranked in the top M for night calls, where N is 10 and M is 5. Night calls refer to calls from 18:00 to 6:00 in the morning, that is, all the first contact numbers ranked in the top 10 and all the first contact numbers ranked in the top 5 in the past 6 months are obtained from the social circle data, and all the first contact numbers with stable contacts are filtered out from the social circle data according to the screening rules, that is, according to the screening rules, the contact time within the eighth preset time period is not less than the first threshold, the eighth preset time period is nearly 6 months, and the first threshold is 4 months, that is, the first contact numbers with stable contacts have a contact time of not less than 4 months in the past 6 months, where N, M and the screening rules can be adjusted and set according to actual conditions.

[0066] S22, combining the user's account opening certificate number, first contact number, in-province number, out-of-province number, first family number, first colleague number, historical terminal data, frequently used terminal data, first work base station, first residence base station, first contact circle location, roaming location, signaling trace, APP usage preference data and traffic content preference data to obtain unique fingerprint information of the user number when the operator attribute is itself;

[0067] In this embodiment, the acquired user account opening certificate number, first contact number, out-of-province number within the province, out-of-province number, first family number, first colleague number, historical terminal data, frequently used terminal data, first work base station, first residence base station, first social circle location, roaming location, signaling trajectory, APP usage preference data and traffic content preference data are combined as shown in Table 1, thereby obtaining the unique fingerprint information of the user number when the operator attribute is itself.

[0068] Table 1. Combination table of unique fingerprint information when the operator attribute is itself

[0069]

[0070]

[0071] S23. When the operator attribute is a competitor, obtain the second contact number, the second family number and the second colleague number within the seventh preset time period from the social circle data, obtain the second work base station, the second residential base station and the second social circle location from the location data, combine the second contact number, the second family number, the second colleague number, the second work base station, the second residential base station and the second social circle location to obtain unique fingerprint information of the user number when the operator attribute is a competitor.

[0072] At this time, the step S23 of acquiring the second contact number within the seventh preset time period from the contact circle data includes:

[0073] S231, obtaining the first second contact number ranked in the top 1, all second contact numbers ranked in the top Q, and all second contact numbers ranked in the top P for night calls within the seventh preset time period from the social circle data, and filtering out all second contact numbers of stable contacts from the social circle data according to the filtering rule;

[0074] In this embodiment, when the operator attribute is a competitor, the second contact number, the second family number, and the second colleague number within the seventh preset time period are obtained from the circle of contacts data, wherein the seventh preset time period is nearly 6 months, and the obtained second contact numbers are further processed to obtain the second contact number ranked first, i.e., the first ranked first, all the second contact numbers ranked in the top Q positions, and all the second contact numbers ranked in the top P position for night calls, wherein Q is 10 and P is 5, and all the second contact numbers with stable contacts are filtered out from the circle of contacts data according to the screening rules, i.e., the contact time within the eighth preset time period according to the screening rules is not less than the first Threshold, the eighth preset time period is nearly 6 months, and the first threshold is 4 months, that is, the second contact number with a contact time of not less than 4 months in the past 6 months is a stable contact number, wherein Q, P and screening rules can be adjusted and set according to actual conditions, and the second contact number, second family number and second colleague number obtained are obtained from the location data. The second work base station, second residence base station and second circle of contacts are obtained, and the second contact number, second family number, second colleague number, second work base station, second residence base station and second circle of contacts are combined to obtain the unique fingerprint information of the user number when the operator attribute is a competitor.

[0075] S3. Compare the unique fingerprint information in the fingerprint database in pairs in sequence, calculate the overlap degree, and obtain the multi-card user based on the overlap degree.

[0076] At this time, step S3 includes:

[0077] S31, classifying the unique fingerprint information in the fingerprint database according to the type of the operator attribute to obtain classified unique fingerprint information;

[0078] S32, cross-combining the classified unique fingerprint information to form D different types of comparison scenarios, and comparing the unique fingerprint information in the fingerprint database in pairs in each different type of comparison scenario and each same type of comparison scenario, calculating the overlap degree, and obtaining the multi-card user according to the overlap degree.

[0079] In this embodiment, if Figure 2 As shown, the unique fingerprint information in the fingerprint database is classified according to the type of operator attribute. The types of operator attributes are divided into three categories, namely: telecommunications, mobile and Unicom, and the classified unique fingerprint information is obtained. The classified unique fingerprint information is cross-combined to form D different types of comparison scenarios, such as: comparison scenarios between telecommunications and mobile, comparison scenarios between telecommunications and Unicom, and comparison scenarios between mobile and Unicom. In each different type of comparison scenario and each same type of comparison scenario, the unique fingerprint information in the fingerprint database is respectively compared two by two in turn, wherein the same type of comparison scenarios are: comparison scenarios between telecommunications and telecommunications, comparison scenarios between Unicom and Unicom, and comparison scenarios between mobile and mobile, so as to calculate the overlap degree, and obtain the one-person multi-card user according to the overlap degree.

[0080] At this time, in step S32, in each different type comparison scenario and each same type comparison scenario, the unique fingerprint information in the fingerprint database is respectively compared in pairs, and the overlap degree is calculated. The one-person multi-card user obtained according to the overlap degree includes:

[0081] S321, when calculating the coincidence degree, the corresponding method is used for calculation according to the type of the fingerprint variable. When the type of the fingerprint variable is attribute type, the coincidence degree is calculated by the precise comparison method. When the type of the fingerprint variable is sequence type, the coincidence degree is calculated by the overlap comparison method. When the type of the fingerprint variable is position type, the coincidence degree is calculated by the distance comparison method.

[0082] In this embodiment, if Figure 2 As shown, when calculating the coincidence, a diversified calculation method is used according to the type of fingerprint variable. For example, when the type of fingerprint variable is attribute class, such as user account opening certificate number, the coincidence is calculated by precise comparison method, that is, the fingerprint variable of attribute class needs to be completely consistent. If there is inconsistency, the coincidence is 0. When the type of fingerprint variable is sequence class, the overlap comparison method is used to calculate the coincidence, such as social circle data and terminal data. When the type of fingerprint variable is location class, the distance comparison method is used to calculate the coincidence, such as location data.

[0083] At this time, when the type of the fingerprint variable in step S321 is a sequence type, the overlap comparison method is used to calculate the coincidence degree, including:

[0084] S3211. When the sequence class is social circle data, in the same type comparison scenario, the social circle data is directly overlapped and the calculated first overlap is used as the coincidence. In different types of comparison scenarios, if the operator attributes are the operator itself and the competitor respectively, candidate numbers are screened out from the user numbers of the operator being the operator itself and the operator being the competitor according to the first coincidence rule, and then the overlap comparison method is used to calculate the second overlap between the unique fingerprint information corresponding to the candidate numbers, and the second overlap is used as the coincidence. If the operator attributes are all competitors, suspected numbers are screened out from the user numbers of the operator being the competitor according to the second coincidence rule, and then the overlap comparison method is used to calculate the third overlap between the unique fingerprint information corresponding to the suspected numbers, and the third overlap is used as the coincidence.

[0085] In this embodiment, when calculating the overlap of the sequence-type social circle data, further division is performed. In the same type of comparison scenario, such as the comparison scenario between telecom and telecom, the comparison scenario between Unicom and Unicom, and the comparison scenario between Mobile and Mobile, the overlap of the social circle data is directly calculated. For the comparison scenarios of different types, if the operator attributes are the operator itself and the competitor, the candidate numbers are screened out from the user numbers whose operator attributes are the operator itself and the operator attributes are the competitor according to the first overlap rule, wherein the first overlap rule has three judgment conditions, namely:

[0086] 1. Cross-check the contact numbers of the current month, the previous month and the month before that of the user numbers whose operator attributes are their own and their competitors. If more than 3 of the top 10 contact numbers overlap at least once and the contact numbers with the most calls between the two are the same, then select them as candidate numbers;

[0087] 2. Cross-check the contact numbers of the current month, the previous month and the month before that of the user numbers whose operator attributes are their own and their competitors. If there is at least one overlap of more than 3 contact numbers in the top 10 and the overlapping contact numbers include the number with the most calls, select it as a candidate number;

[0088] 3. Cross-check the numbers of the users with the same operator and the competitors in the current month, the previous month and the month before that. If the number of calls between the two numbers in the past three months is less than three times and the call duration is less than five minutes, select them as candidate numbers.

[0089] The above three judgment conditions are in an or relationship, and the overlap comparison method is used to calculate the second overlap between the unique fingerprint information corresponding to the selected candidate numbers. When the operator attributes are all competitors, the suspected numbers are screened out from the user numbers whose operator attributes are competitors according to the second overlap rule. The second overlap rule has two judgment conditions, namely:

[0090] 1. Cross-check the contact numbers of the current month, the previous month, and the month before that for user numbers whose operator attributes are all competitors. If there is at least one overlap of more than 3 of the top 10 contact numbers and the overlapping contact numbers include the number with the most calls, then screen it out as a suspected number;

[0091] 2. Cross-check the contact numbers of the current month, the previous month, and the month before that for user numbers whose operator attributes are all competitors. If there is at least one overlap of more than 3 contact numbers among the top 10 contact numbers and the overlapping contact numbers include the number with the most calls, then screen it out as a suspected number;

[0092] The above two judgment conditions are in an OR relationship, and then the overlap comparison method is used to calculate the third overlap between the unique fingerprint information corresponding to the screened suspected numbers.

[0093] S322. In the same type of comparison scenario, the first conversion rule is used to convert the overlap degree into an overlap score value. In different types of comparison scenarios, the second conversion rule is used to convert the overlap degree into an overlap score value, and a one-person-multiple-card user is obtained according to the overlap score value.

[0094] In this embodiment, in the same type comparison scenario, the first conversion rule is used to convert the overlap degree into an overlap score value, wherein the first conversion rule is shown in Table 2:

[0095] Table 2. First transformation rules

[0096]

[0097]

[0098]

[0099] The second transformation rule is shown in Table 3:

[0100] Table 3. Second transformation rules

[0101]

[0102]

[0103] At this time, the step S322 of obtaining a user with multiple cards according to the overlap score includes:

[0104] S3221. In the same type comparison scenario, if the overlap score value exceeds the first score threshold, and the user numbers of the two unique fingerprint information corresponding to the first score threshold are both integrated cards, then the user numbers of the two unique fingerprint information are the same user, and a multi-card user is obtained, and the integrated cards are a voice card and a traffic card;

[0105] Or, in the same type comparison scenario, if the overlap score value exceeds the second score threshold, and the user numbers of the two unique fingerprint information corresponding to the second score threshold are a voice card and a traffic card respectively, then the user numbers of the two unique fingerprint information are the same user, and a multi-card user is obtained;

[0106] In this embodiment, if Figure 2 As shown, for the same type of comparison scenario, when identifying a user with multiple cards, the judgment is not only made based on the overlap score value, but also based on whether the user number is a comprehensive card. When the overlap score value exceeds the first score threshold, and the user numbers of the two unique fingerprint information corresponding to the first score threshold are both comprehensive cards, that is, both are dual-card comprehensive cards of a voice card and a traffic card, then the user numbers of the two unique fingerprint information are the same user, and a user with multiple cards is obtained, wherein the first score threshold is 70 points, or, the overlap score value exceeds the second score threshold, and the user numbers of the two unique fingerprint information corresponding to the first score threshold are a language card and a traffic card, respectively, then the user numbers of the two unique fingerprint information are the same user, and a user with multiple cards is obtained, wherein the second score threshold is 60 points. The specific first score threshold and the second score threshold can be adjusted according to actual conditions.

[0107] S3222. In different types of comparison scenarios, if the overlap score value exceeds a third score threshold, the user numbers of the two unique fingerprint information are the same user, and a one-person-multiple-card user is obtained.

[0108] In this embodiment, if Figure 2 As shown, in different types of comparison scenarios, if the overlap score value exceeds the third score threshold, the user numbers of the two unique fingerprint information are the same user, and a one-person multi-card user is obtained, where the third score threshold is 35 points. The specific third score threshold can be adjusted according to actual conditions.

[0109] In this embodiment, after identifying a card user, the primary card identification is also performed. The specific steps of the primary card identification are as follows:

[0110] Obtaining the usage of all users' cards, classifying the cards according to the usage, and obtaining corresponding grade information;

[0111] In this embodiment, if Figure 2 As shown, by obtaining the card usage of all users, the cards are classified into grades according to the usage, and the corresponding grade information is obtained, wherein the grade classification standard is: the cards of the entire network users are ranked in descending order according to their usage, and each 10% of users is divided into 1 grade, and there are a total of 10 grades.

[0112] Calculate the usage ratio of each card of the same user in the usage ratio of all cards, use the usage ratio as the user's network age contribution, and calculate the consumption amount, call minutes and data traffic value of each card of the same user within a preset period;

[0113] A multi-dimensional evaluation index is established from the grade information, the network age contribution, the consumption amount, the number of call minutes and the data traffic value through the hierarchical analysis method, and the importance score of each card corresponding to each user is calculated. The card with the highest importance score is screened out according to the importance score. If there are multiple cards with the highest importance score, they are sorted in descending order according to the consumption amount or the number of call minutes or the data traffic value to obtain the cards in descending order. The card with the highest importance score that is ranked first is used as the main card. If there are not multiple cards with the highest importance scores, the card with the highest importance score is directly used as the main card.

[0114] In this embodiment, if Figure 2 As shown, the usage ratio of each card of the same user in the usage ratio of all the cards owned by the same user is calculated, and the usage ratio is used as the user's network age contribution, and the consumption amount, call minutes and data flow value of each card of the same user within a preset period are calculated, wherein the preset period can be annual, quarterly, time period from card opening time to present, etc., and can be adjusted according to actual conditions, so as to establish a multi-dimensional evaluation index from grade information, network age contribution, consumption amount, call minutes and data flow value through hierarchical analysis method, and calculate the importance score of each card. When calculating the importance score, corresponding weights are set for different dimensions according to operator attributes, so as to calculate the importance score in combination with the weights, and then the card with the highest importance score is screened out according to the importance score. If multiple cards with the highest importance score are screened out, they are sorted in descending order according to the consumption amount or call minutes or data flow value to obtain the cards after descending order, and the card with the highest importance score at the first place is used as the main card, and the rest are used as secondary cards. If one card with the highest importance score is screened out, that is, there are no multiple cards, then the card with the highest importance score is directly used as the main card, and the rest are used as secondary cards.

[0115] Embodiment 2

[0116] Please refer to Figure 3A device 1 for identifying a user with multiple cards includes a memory 3, a processor 2, and a computer program stored in the memory 3 and executable on the processor 2. When the processor 2 executes the computer program, the steps in the first embodiment are implemented.

[0117] Since the system / device described in the above embodiments of the present invention is a system / device used to implement the method of the above embodiments of the present invention, a person skilled in the art can understand the specific structure and deformation of the system / device based on the method described in the above embodiments of the present invention, and thus will not be described in detail here. All systems / devices used in the method of the above embodiments of the present invention belong to the scope of protection of the present invention.

[0118] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented 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.

[0119] The present invention is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present invention. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions.

[0120] It should be noted that in the claims, any reference numerals placed between brackets shall not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claims. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention may be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In the claims enumerating several means, several of these means may be embodied by the same hardware. The use of the words first, second, third, etc., is for convenience of expression only and does not indicate any order. These words may be understood as part of the component name.

[0121] In addition, it should be noted that, in the description of this specification, the description of the terms "one embodiment", "some embodiments", "embodiment", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0122] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments after knowing the basic creative concept. Therefore, the claims should be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present invention.

[0123] Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention should also include these modifications and variations.

Claims

1. A method for identifying a user with multiple cards, characterized in that: include: Acquire fingerprint data of the user number according to preset fingerprint variables, wherein the fingerprint data includes operator attributes, social circle data and location data; Acquire an analysis time window of the fingerprint variable according to the operator attribute, combine the fingerprint data under the analysis time window according to a preset construction rule, obtain unique fingerprint information of user numbers corresponding to different operator attributes, and aggregate the unique fingerprint information of all user numbers to obtain a fingerprint database; The unique fingerprint information in the fingerprint database is compared in pairs in sequence, the overlap degree is calculated, and the multi-card user is obtained according to the overlap degree.

2. A method for identifying a user with multiple cards as claimed in claim 1, characterized in that: The step of obtaining the analysis time window of the fingerprint variable according to the operator attribute, and combining the fingerprint data under the analysis time window according to a preset construction rule to obtain unique fingerprint information of the user number corresponding to different operator attributes includes: When the operator attribute is itself, the fingerprint data includes the user account opening certificate number, terminal data and content preference data, and the first contact number, the out-of-province number, the out-of-province number, the first family number and the first colleague number within the first preset time period are obtained from the social circle data; the historical terminal data within the second preset time period and the commonly used terminal data within the third preset time period are obtained from the terminal data; the first working base station, the first residential base station, the first social circle location, the roaming location within the fourth preset time period and the signaling track within the fifth preset time period are obtained from the location data; the APP usage preference data and the traffic content preference data within the sixth preset time period are obtained from the content preference data; Combine the user's account opening certificate number, first contact number, in-province number, out-of-province number, first family number, first colleague number, historical terminal data, frequently used terminal data, first work base station, first residential base station, first contact circle location, roaming location, signaling trace, APP usage preference data and traffic content preference data to obtain the unique fingerprint information of the user number when the operator attribute is itself; When the operator attribute is a competitor, the second contact number, the second family number and the second colleague number within the seventh preset time period are obtained from the social circle data, the second work base station, the second residence base station and the second social circle location are obtained from the location data, and the second contact number, the second family number, the second colleague number, the second work base station, the second residence base station and the second social circle location are combined to obtain unique fingerprint information of the user number when the operator attribute is a competitor.

3. A method for identifying a user with multiple cards as claimed in claim 2, characterized in that: The step of acquiring the first contact number within the first preset time period from the circle of contacts data comprises: Obtaining all first contact numbers ranked in the top N places and all first contact numbers ranked in the top M places for nighttime calls within a first preset time period from the contact circle data, and filtering out all first contact numbers of stable contacts from the contact circle data according to the filtering rule; The step of acquiring the second contact number within the seventh preset time period from the circle of contacts data includes: Obtaining the first second contact number ranked in the top 1, all second contact numbers ranked in the top Q, and all second contact numbers ranked in the top P for night calls from the social circle data within the seventh preset time period, and filtering out all second contact numbers of stable contacts from the social circle data according to the screening rule; The screening rule is that the contact time within the eighth preset time period is not less than the first threshold.

4. A method for identifying a user with multiple cards as claimed in claim 1, characterized in that: The unique fingerprint information is dynamically updated as the fingerprint data changes.

5. A method for identifying a user with multiple cards as claimed in claim 1, characterized in that: The unique fingerprint information in the fingerprint database is sequentially compared with each other, and the overlap degree is calculated. According to the overlap degree, the user with multiple cards is obtained, which includes: Classifying the unique fingerprint information in the fingerprint database according to the type of the operator attribute to obtain classified unique fingerprint information; The classified unique fingerprint information is cross-combined to form D different types of comparison scenarios. In each different type of comparison scenario and each same type of comparison scenario, the unique fingerprint information in the fingerprint database is respectively compared two by two in sequence, and the overlap degree is calculated. According to the overlap degree, the multi-card user is obtained.

6. A method for identifying a user with multiple cards as claimed in claim 5, characterized in that: In each of the different types of comparison scenarios and each of the same type of comparison scenarios, the unique fingerprint information in the fingerprint database is respectively compared in pairs, and the overlap degree is calculated. The one-person multi-card user is obtained according to the overlap degree, including: When calculating the coincidence degree, the corresponding method is used for calculation according to the type of the fingerprint variable. When the type of the fingerprint variable is attribute type, the coincidence degree is calculated by the precise comparison method. When the type of the fingerprint variable is sequence type, the coincidence degree is calculated by the overlap comparison method. When the type of the fingerprint variable is position type, the coincidence degree is calculated by the distance comparison method. In the same type of comparison scenario, the first conversion rule is used to convert the overlap degree into an overlap score value. In different types of comparison scenarios, the second conversion rule is used to convert the overlap degree into an overlap score value, and a one-person multi-card user is obtained according to the overlap score value.

7. A method for identifying a user with multiple cards as claimed in claim 6, characterized in that: When the type of the fingerprint variable is a sequence type, the overlap comparison method is used to calculate the coincidence degree, including: When the sequence class is social circle data, in the same type of comparison scenario, the social circle data is directly overlapped and the calculated first overlap is used as the coincidence degree. In different types of comparison scenarios, if the operator attributes are respectively itself and competitors, candidate numbers are screened out from the user numbers of the operator being itself and the operator being competitors according to the first overlap rule, and then the overlap comparison method is used to calculate the second overlap between the unique fingerprint information corresponding to the candidate numbers, and the second overlap is used as the coincidence degree. If the operator attributes are all competitors, suspected numbers are screened out from the user numbers of the operator being competitors according to the second overlap rule, and then the overlap comparison method is used to calculate the third overlap between the unique fingerprint information corresponding to the suspected numbers, and the third overlap is used as the coincidence degree.

8. A method for identifying a user with multiple cards as claimed in claim 6, characterized in that: The method of obtaining a user with multiple cards according to the overlap score includes: In the same type comparison scenario, if the overlap score value exceeds the first score threshold, and the user numbers of the two unique fingerprint information corresponding to the first score threshold are both comprehensive cards, then the user numbers of the two unique fingerprint information are the same user, and a multi-card user is obtained, and the comprehensive card is a voice card and a traffic card; Or, in the same type comparison scenario, if the overlap score value exceeds the second score threshold, and the user numbers of the two unique fingerprint information corresponding to the second score threshold are a voice card and a traffic card respectively, then the user numbers of the two unique fingerprint information are the same user, and a multi-card user is obtained; In different types of comparison scenarios, if the overlap score value exceeds the third score threshold, the user numbers of the two unique fingerprint information are the same user, and a one-person-multiple-card user is obtained.

9. A method for identifying a user with multiple cards as claimed in claim 1, characterized in that: The obtaining of a user with multiple cards according to the overlap degree further comprises: Obtaining the usage of all users' cards, classifying the cards according to the usage, and obtaining corresponding grade information; Calculate the usage ratio of each card of the same user in the usage ratio of all cards, use the usage ratio as the user's network age contribution, and calculate the consumption amount, call minutes and data traffic value of each card of the same user within a preset period; A multi-dimensional evaluation index is established from the grade information, the network age contribution, the consumption amount, the number of call minutes and the data traffic value through the hierarchical analysis method, and the importance score of each card corresponding to each user is calculated. The card with the highest importance score is screened out according to the importance score. If there are multiple cards with the highest importance score, they are sorted in descending order according to the consumption amount or the number of call minutes or the data traffic value to obtain the cards in descending order. The card with the highest importance score that is ranked first is used as the main card. If there are not multiple cards with the highest importance scores, the card with the highest importance score is directly used as the main card.

10. A device for identifying a user with multiple cards, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 9 is implemented.

Images