A fraud identification method, device and electronic equipment based on voiceprint technology

Abstract

This invention provides a fraud identification method, device, and electronic device based on voiceprint technology. The method includes: acquiring voice information to be matched, the voice information including a user identifier and voice data; searching a voiceprint database for voiceprint records that match the voice information based on the user identifier and / or the voice data; if a fraud tag exists in the voiceprint record matching the voice information, then the voice information to be matched is determined to be fraudulent. This invention achieves simple and rapid identification of fraudulent information, improving the timeliness of fraud identification.

Description

Technical Field

[0001] This invention relates to the field of mobile communication technology, and in particular to a fraud identification method, device, and electronic device based on voiceprint technology. Background Technology

[0002] With the rapid development and widespread adoption of wireless communication technology, while bringing great convenience to users, the use of telecommunications networks for fraud has also seen a high incidence and rapid growth. Telecommunications fraud refers to criminal acts in which criminals fabricate false information and set up scams through telephone, internet, or text messages to remotely and non-contactly defraud victims, inducing them to make payments or transfers to the criminals.

[0003] Traditional methods for identifying telecom fraud often rely on call signaling and call detail record analysis to analyze user communication behavior characteristics and manually summarize rules. For example, anti-telecom fraud models are developed based on features such as caller ID behavior, International Mobile Equipment Identity (IMEI) communication behavior, and communication patterns of user caller ID numbers.

[0004] Traditional anti-fraud models for telecommunications require a large amount of user communication behavior data for analysis and judgment, making it difficult to easily and quickly identify fraudulent users. Summary of the Invention

[0005] The purpose of this invention is to provide a fraud identification method, device, and electronic device based on voiceprint technology to solve the problem of not being able to easily and quickly identify fraudulent users.

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

[0007] In a first aspect, embodiments of the present invention provide a fraud identification method based on voiceprint technology, comprising:

[0008] Obtain the voice information to be matched, the voice information including: user identifier and voice data;

[0009] Based on user identifier and / or voice data, search the voiceprint database for voiceprint records that match the voice information;

[0010] If a fraud tag is found in the voiceprint record that matches the voice information, then the voice information to be matched is determined to be fraudulent.

[0011] Secondly, embodiments of the present invention provide a fraud identification device based on voiceprint technology, comprising:

[0012] The acquisition unit is used to acquire the voice information to be matched, the voice information including: user identifier and voice data;

[0013] The matching unit is used to search for voiceprint records that match the voice information from the voiceprint database based on the user identifier and / or voice data, respectively.

[0014] The identification unit is used to determine that the voice information to be matched is fraudulent if a fraud tag exists in the voiceprint record that matches the voice information.

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

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

[0017] As can be seen from the technical solutions provided by the above embodiments of the present invention, the embodiments of the present invention obtain voice information to be matched, the voice information including: user identifier and voice data; based on the user identifier and / or voice data respectively, search for voiceprint records that match the voice information from a voiceprint database; if a fraud tag exists in the voiceprint record that matches the voice information, then the voice information to be matched is determined to be fraudulent information. Through the embodiments of the present invention, simple and rapid identification of fraudulent information is achieved, improving the timeliness of fraud identification. Attached Figure Description

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

[0019] Figure 1 A schematic flowchart of a fraud identification method based on voiceprint technology provided in an embodiment of the present invention;

[0020] Figure 2 This is another flowchart illustrating the fraud identification method based on voiceprint technology provided in an embodiment of the present invention.

[0021] Figure 3 A schematic diagram of a fraud identification device based on voiceprint technology provided in an embodiment of the present invention;

[0022] Figure 4 Another structural schematic diagram of a fraud identification device based on voiceprint technology provided in an embodiment of the present invention;

[0023] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0024] This invention provides a fraud identification method, device, and electronic device based on voiceprint technology.

[0025] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0026] like Figure 1 As shown, this embodiment of the invention provides a fraud identification method based on voiceprint technology. The execution subject of this method can be a server, which can be a standalone server or a server cluster composed of multiple servers. Furthermore, the server can be a server capable of network operation processing, such as a server configuring network resources. The method specifically includes the following steps:

[0027] Step S110: Obtain the voice information to be matched, which includes: user identifier and voice data. The user identifier can be a user number, and the data format of the voice data can be determined according to actual needs or acquisition methods, such as WAV, AVI, MP3, etc.

[0028] There are various ways for a server to obtain voice information. This application only provides examples of a few of these methods:

[0029] In one implementation, when a new user opens an account, the voiceprint acquisition module collects the user's voice during the identity verification process to obtain voice data. The user's associated phone number is used as the user identifier to obtain the voice information to be matched. Alternatively, when an existing user conducts business, the user's voice is collected as voice data, and the existing user's associated phone number is used as the user identifier to obtain the voice information to be matched.

[0030] In another implementation, during a voice call, signaling messages and voice packets are collected by a voice acquisition module, and the voice information to be matched is obtained by parsing the signaling messages and voice packets.

[0031] In one implementation, the process of acquiring voice information through the voice acquisition module can be divided into a signaling acquisition part, a voice acquisition part, a signaling parsing part, a signaling synthesis part, a voice parsing part, a voice synthesis part, and a voice encoding and decoding part, with various devices using the Network Time Protocol (NTP) service to unify the time.

[0032] In the signaling acquisition section and the voice acquisition section, acquisition devices for signaling data and user plane voice data can be deployed in the voice core network to collect raw signaling messages and voice packets:

[0033] For calls on 2G and 3G networks, BICC signaling messages from the MSC Server network element and voice packets from the MGW network element are collected; for calls on 4G networks and CM_IMS fixed networks, SIP signaling messages from the SBC network element and voice packets from the user plane are collected.

[0034] In the signaling parsing section, the original signaling message is parsed into a series of signaling data records, which may include: message type, header fields and values, timestamps, etc.

[0035] In the signaling synthesis section, the signaling data records of the same call are associated into a single synthesized signaling data record by using the calling and called numbers and call-id in the header field and value. The synthesized signaling data record may include: the calling number, the called number, the calling side IP address and port of the user plane used in the call, the called side IP address and port, the voice code used in the call, the call start time, and the call end time.

[0036] In the voice parsing section, based on the source IP address and port, destination IP address and port of the voice packets, the voice packets are synthesized into voice segments at a preset period, such as 30 seconds, and the source IP address and port, destination IP address and port, time of the first voice packet and time of the last voice packet of the voice segment are recorded.

[0037] In the speech synthesis section, the caller's IP address and port, the called party's IP address and port, the call start time and the call end time are obtained from the synthesized signaling data record. The speech segments obtained from the speech parsing section are then synthesized into a complete call speech on the caller's side and the called party's side, and associated with the voice-generated number, the caller's number, the called party's number, the call start time, the call end time, the speech encoding and other information of the call.

[0038] The voice messages from the calling side are sourced from the calling side's IP address and port, and the destination IP address and port are from the called side's IP address and port. The voice segments within this time interval are merged based on the call start time and call end time. The voice messages from the called side are sourced from the called side's IP address and port, and the destination IP address and port are from the calling side's IP address and port. The voice segments within this time interval are merged based on the call start time and call end time.

[0039] In the voice encoding and decoding section, the synthesized call voice is converted into a preset data format, such as WAV format, based on the voice encoding. The final voice information includes: the user's number in the call voice, the voice data in the preset data format, and basic information about the call, such as the caller's number, the called number, the call start time, and the call end time.

[0040] In the voice encoding and decoding section, commonly used voice codes in communication networks can be adopted, such as G711 or AMR.

[0041] The following describes the signaling message processing procedure during a VoLTE call:

[0042] VoLTE calls use SIP signaling. The call signaling flow is shown in the figure below. The signaling plane messages for voice calls and video calls are basically the same, except for the encoding in the Session Description Protocol (SDP).

[0043] In the signaling acquisition section, the INVITE message, UPDATE message, and the 200 OK message corresponding to UPDATE in the signaling plane messages of the VoLTE call are collected as the calling number, the called number, the IP address and port of the user plane used in the call, and the source of the voice encoding used in the call. The BYE message is collected to indicate the end time of the call.

[0044] A VoLTE call is initiated with an INVITE message. The INVITE message contains a header field: From, which identifies the calling number; To, which identifies the called number; and a call-ID to identify the call. The SDP media plane information in the INVITE message is still pending confirmation and cannot be used later.

[0045] The NVITE message contains the same header fields From, To, and call-ID. In the SDP media plane information of the UPDATE message, the Connection Information header field indicates the calling party's media plane IP address, the Media Description contains the calling party's media plane port number, and the Media Attribute contains the voice encoding method.

[0046] The 200 OK message corresponding to UPDATE contains the same From, To, and call-ID header fields as the INVITE message. In the SDP media plane information of the 200 OK message, the Connection Information header field indicates the called party's media plane IP address, the Media Description contains the called party's media plane port number, and the Media Attribute contains the voice encoding method (the same as in UPDATE).

[0047] The 200 OK message corresponding to UPDATE contains the same From, To, and call-ID header fields as the INVITE message. In the SDP media plane information of the 200 OK message, the Connection Information header field indicates the called party's media plane IP address, the Media Description contains the called party's media plane port number, and the Media Attribute contains the voice encoding method (the same as in UPDATE).

[0048] Similar to VoLTE calls, signaling from the 2G / 3G voice core network is collected. The call_id, calling number, called number, and call start time are extracted from the IAM message; the calling side media plane IP address and port, the called side media plane IP address and port, and the media encoding method are extracted from the APM message; and the call end time is extracted from the REL message.

[0049] Step S120: Based on the user identifier and / or voice data, search the voiceprint database for voiceprint records that match the voice information.

[0050] The server pre-establishes a voiceprint database containing a large number of voiceprint records. Each voiceprint record may include a user identifier, voice data, and voiceprint features corresponding to the voice data.

[0051] The voiceprint features can be obtained by processing the speech data using a preset voiceprint recognition model and then extracting the speech features. Specifically, they can be represented in the form of a voiceprint representation vector.

[0052] The voiceprint record may also include the source of the voice data, which can be of various types, including new user registration, business processing, and voice calls.

[0053] The voiceprint record may also include other voice-related information, such as storage path, whether it is a high-risk voice, user number-related information, such as whether it is a high-risk user, and the correlation between voiceprints of different users.

[0054] Based on the user identifier and / or voice data of the voice information, the voice information is matched with each voiceprint record in the voiceprint database to find whether there is a voiceprint record that matches the voice information.

[0055] In one implementation, the matching voiceprint record can specifically be a voiceprint record that satisfies at least one of the following matching conditions:

[0056] The user identifier recorded in the voiceprint is the same as the user identifier in the voice information.

[0057] The voiceprint data recorded by voiceprint has the same voiceprint features as the voice data of voice information;

[0058] The voiceprint features recorded are the same as those of the voiceprint data in the speech information.

[0059] Step S130: If a fraud tag exists in the voiceprint record that matches the voice information, then the voice information to be matched is determined to be fraud information.

[0060] The voiceprint database can include voiceprint records marked as fraudulent. These records can be marked by adding a fraud tag. Voiceprint records containing the fraud tag are identified as fraudulent, while those without the fraud tag are identified as ordinary voiceprint records.

[0061] In one implementation, the voiceprint database can be divided into a general voiceprint database and a fraud-related voiceprint database. All general voiceprint records are saved to the general voiceprint database, and fraud-related voiceprint records are saved to the fraud-related voiceprint database.

[0062] The acquisition of ordinary voiceprint records can be achieved in various ways, including through the voiceprint acquisition module or voice acquisition module during new user registration, business processing, and voice calls. The ordinary voiceprint record includes: user identifier and voice data. It may also include voiceprint features corresponding to the voice data.

[0063] There are various ways to obtain the fraudulent voiceprint records. This application embodiment only provides a few of these methods:

[0064] 1. Voice data related to fraud provided by a third party, or voice data related to fraud confirmed by other means, such as voice data of individuals involved in fraud provided by the police;

[0065] 2. User identifiers suspected of fraud provided by a third party, or user identifiers suspected of fraud confirmed through other means;

[0066] 3. Based on the user identifier or voice data involved in fraud, match it with the user identifier or voice data in ordinary voiceprint records. Then, mark the matched ordinary voiceprint records with a fraud tag and import them as new fraud-related voiceprint records into the fraud-related voiceprint database. It can be seen that the fraud-related voiceprint records may include user identifiers and / or voice data. If voice data exists, voiceprint features corresponding to the voice data can be obtained through processing by a voiceprint recognition model.

[0067] When the server matches the voice information to be matched with each voiceprint record based on the user identifier and / or voice data, if a fraud tag is found in the voiceprint record that matches the voice information, that is, the matched voiceprint record is a fraudulent voiceprint record, then the voice information can be determined to be fraudulent information, and the user identifier and voice data in the voice information will be reported as fraudulent user identifier and voice data; if no fraud tag is found in the voiceprint record that matches the voice information, that is, the matched voiceprint record is a normal voiceprint record, then the voice information can be determined to be normal information.

[0068] As can be seen from the technical solutions provided by the above embodiments of the present invention, the embodiments of the present invention obtain voice information to be matched, the voice information including: user identifier and voice data; based on the user identifier and / or voice data respectively, search for voiceprint records that match the voice information from a voiceprint database; if a fraud tag exists in the voiceprint record that matches the voice information, then the voice information to be matched is determined to be fraudulent information. Through the embodiments of the present invention, simple and rapid identification of fraudulent information is achieved, improving the timeliness of fraud identification.

[0069] Based on the above embodiments, further, such as Figure 2 As shown, there are many ways to process step S120 above. Here is one optional processing method, which can be found in the following steps.

[0070] Step S121: Select voiceprint records with the same user identifier a1 as the voice information S1 from the voiceprint database as voiceprint records R1 to be matched.

[0071] The voice information S1 to be matched includes: user identifier a1 and voice data b1.

[0072] Based on user identifier a1, select all voiceprint records under user identifier a1 from the voiceprint database as voiceprint record R1 to be matched.

[0073] If no voiceprint record with user identifier a1 is found in the voiceprint database, a new voiceprint record R2 can be obtained based on the voice information S1 and saved to the voiceprint database.

[0074] If the selected voiceprint record R1 to be matched is a fraudulent voiceprint record, then the voice information S1 is directly identified as fraudulent information. Alternatively, it can be determined directly based on user identifier a1 whether user identifier a1 is a fraudulent user identifier. If user identifier a1 is determined to be a fraudulent user identifier, then the voice information S1 is directly identified as fraudulent information.

[0075] Step S122: Using a preset voice matching method, compare the voiceprint feature c1 of the voice data b1 of the voice information S1 with the voiceprint feature r1 of the voice data of each voiceprint record R1 to be matched.

[0076] In one embodiment, a voiceprint recognition and comparison model is established. The voice data b1 of the voice information is processed by the voiceprint recognition and comparison model to extract the voiceprint feature c1 of the voice data b1. The voiceprint feature c1 is compared with the voiceprint feature r1 of each voiceprint record R1 to be matched to obtain the similarity value between the voiceprint feature c1 and the voiceprint feature r1. If the similarity value meets the preset similarity threshold, the voiceprint feature c1 and the voiceprint feature r1 are determined to be the same voiceprint feature.

[0077] Step S123: If no matching voiceprint record R1 with the same voiceprint feature c1 as the voice data b1 of the voice information S1 is found, then a new voiceprint record R2 is added to the voiceprint database according to the voice information S1, and the voiceprint feature c1 of the voice data b1 of the new voiceprint record R2 is compared with the voiceprint feature r3 of the voice data of other voiceprint records R3 in the voiceprint database.

[0078] Further, the voiceprint record includes a matching count, which indicates the number of times the voiceprint record successfully matches the voice information to be matched. After step S122, the method further includes:

[0079] Step S124: If it is determined that the voiceprint record R1 to be matched has the same voiceprint feature c1 as the voice data b1 of the voice information S1, the matching count of the voiceprint record R1 to be matched is incremented by 1.

[0080] Step S125: Take other voiceprint records R3 that have the same voiceprint feature c1 as the new voiceprint record R2 as voiceprint records that match the speech information S1, and associate them with the new voiceprint record R2. The association can be performed by adding corresponding association information to the new voiceprint record R2 and the other voiceprint records R3.

[0081] In one implementation, if the voiceprint database is divided into a general voiceprint database and a fraud-related voiceprint database as described above, then the matching process described above may include:

[0082] 1. Based on the user identifier a1 of the voice information S1, extract the ordinary voiceprint records under the user identifier a1 from the ordinary voiceprint database as voiceprint records R1 to be matched, and compare the voiceprint features c1 of the voice data b1 of the voice information S1 with the voiceprint features r1 of each voiceprint record R1 to be matched one by one through the voiceprint recognition and comparison model.

[0083] 2. If a match is successful, increment the match count in the voiceprint record R1 to be matched by 1;

[0084] 3. If the matching fails, a new voiceprint record R2 is added to the ordinary voiceprint database according to the voice information S1.

[0085] 4. Using the voiceprint recognition and comparison model, the voiceprint feature c1 of the new voiceprint record R2 is compared one by one with the voiceprint features of other ordinary voiceprint records R3 in the ordinary voiceprint database.

[0086] 5. If the match is successful, the new voiceprint record R2 is associated with the voiceprint record R3;

[0087] 6. If the matching fails, the voice information S1 is matched with the fraudulent voiceprint records in the fraud database based on the user identifier a1 and the voice data b1.

[0088] 7. If a match is successful, a fraud-related tag is added to the new voiceprint record R2 and imported into the fraud-related database.

[0089] As can be seen from the technical solutions provided by the above embodiments of the present invention, the embodiments of the present invention match voice information with voiceprint records based on user identifiers and voice data respectively, and perform corresponding processing according to the matching results. Through the embodiments of the present invention, simple and rapid identification of fraudulent information is achieved, the timeliness of fraud identification is improved, and the matching results are accurately recorded.

[0090] Based on the above embodiments, in order to further assist the early warning system in analyzing fraudulent activities, the method further includes:

[0091] If the voiceprint record without fraud label meets the preset conditions, the voiceprint record without fraud label is labeled with a person-card separation scenario label.

[0092] The preset conditions include at least one of the following:

[0093] The number of voiceprints with the same user identifier that match more than the first threshold exceeds the second threshold.

[0094] The number of voiceprints with a matching count greater than the third threshold among the associated voiceprint information exceeds the fourth threshold.

[0095] In one implementation, a pre-set fraud analysis module is used to analyze ordinary voiceprint records in the ordinary voiceprint database based on the user identifier, matching frequency, and association relationships of each ordinary voiceprint record. The ordinary voiceprint database can include scenarios such as one card for multiple users and one person having multiple cards.

[0096] Specifically, the "one card, multiple users" situation refers to a situation where the number of voiceprint features that match the same user identifier more than the first threshold exceeds the second threshold. In this case, the user identifier can be identified as belonging to the "one card, multiple users" situation, and a user-card separation scenario label can be marked in the ordinary voiceprint record with the user identifier.

[0097] The "one person, multiple SIM cards" scenario specifically refers to a situation where the number of user identifiers matching the same voiceprint feature more than the third threshold exceeds the fourth threshold. In this case, the user identifier corresponding to the voiceprint feature can be identified as belonging to the "one person, multiple SIM cards" scenario, and the user-SIM card separation scenario can be marked in the ordinary voiceprint records with that voiceprint feature. Related ordinary voiceprint records can be considered to have the same voiceprint feature.

[0098] The fraud analysis module can also be used to report voiceprint records marked with fraud tags and / or card-person separation scenario tags to a preset early warning system or an external anti-fraud system, thereby increasing the data sources of the external anti-fraud system and the accuracy of fraud analysis.

[0099] In one implementation, the fraud analysis module can also increase the accuracy of the analysis by importing information such as high-risk user numbers and external user tags provided by an external anti-fraud system.

[0100] Based on the above embodiments, the method further includes:

[0101] Obtain user identifiers or voice data suspected of being involved in fraud, and import them into the fraud-related voiceprint database as new fraud-related voiceprint records.

[0102] Voiceprint records matching the fraudulent user identifier or voice data are selected from the voiceprint database and tagged with fraud. Ordinary voiceprint records with the fraudulent user identifier or voice data identical to the fraudulent voice data are selected from the ordinary voiceprint database and tagged with fraud. These ordinary voiceprint records are then imported into the fraudulent voiceprint database.

[0103] like Figure 3 As shown, the server used to execute the voiceprint-based fraud as described above may include: a voiceprint acquisition module 301, a voice acquisition module 302, a voiceprint recognition and comparison module 303, a general voiceprint library 304, a fraud-related voiceprint library 305, and a fraud-related analysis module 306. The fraud-related analysis module 306 is also connected to an early warning system 307, an external anti-fraud system 308, and an external data module 309.

[0104] As can be seen from the technical solutions provided by the above embodiments of the present invention, the embodiments of the present invention, when the voiceprint records without fraud tags meet preset conditions, mark the person-card separation scenario tag in the voiceprint records without fraud tags. Through the embodiments of the present invention, rapid identification of fraud information is achieved, and the accuracy of fraud identification is improved.

[0105] Corresponding to the fraud identification method based on voiceprint technology provided in the above embodiments, based on the same technical concept, this embodiment of the invention also provides a fraud identification device based on voiceprint technology. Figure 4 This is a schematic diagram of the module composition of a fraud detection device based on voiceprint technology provided in an embodiment of the present invention. This fraud detection device based on voiceprint technology is used to perform... Figures 1 to 2 The described fraud detection method based on voiceprint technology, such as Figure 4 As shown, the fraud identification device based on voiceprint technology includes: a collection unit 401, a matching unit 402, and an identification unit 403.

[0106] The acquisition unit 401 is used to acquire voice information to be matched, the voice information including: user identifier and voice data; the matching unit 402 is used to search for voiceprint records that match the voice information from the voiceprint database based on the user identifier and / or voice data respectively; the identification unit 403 is used to determine that the voice information to be matched is fraudulent information if there is a fraud tag in the voiceprint record that matches the voice information.

[0107] Furthermore, the acquisition unit is used to acquire signaling messages and voice packets during a voice call, and obtain the voice information to be matched by parsing the signaling messages and voice packets.

[0108] As can be seen from the technical solutions provided by the above embodiments of the present invention, the embodiments of the present invention obtain voice information to be matched, the voice information including: user identifier and voice data; based on the user identifier and / or voice data respectively, search for voiceprint records that match the voice information from a voiceprint database; if a fraud tag exists in the voiceprint record that matches the voice information, then the voice information to be matched is determined to be fraudulent information. Through the embodiments of the present invention, simple and rapid identification of fraudulent information is achieved, improving the timeliness of fraud identification.

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

[0110] Voiceprint records with the same user identifier as the voice information are selected from the voiceprint database as voiceprint records to be matched.

[0111] Using a preset voice matching method, the voiceprint features of the voice data of the voice information are compared with the voiceprint features of the voice data of each voiceprint record to be matched.

[0112] If no matching voiceprint record with the same voiceprint features as the voice information is found, a new voiceprint record is added to the voiceprint database according to the voice information, and the voiceprint features of the voice data of the new voiceprint record are compared with the voiceprint features of the voice data of other voiceprint records in the voiceprint database.

[0113] Other voiceprint records with the same voiceprint features as the new voiceprint record are identified as voiceprint records that match the speech information and are associated with the new voiceprint record.

[0114] Furthermore, the voiceprint record includes a matching count, and the matching unit is further configured to increment the matching count of the voiceprint record to be matched by 1 when it is determined that the voiceprint record to be matched has the same voiceprint features as the voice data of the voice information.

[0115] As can be seen from the technical solutions provided by the above embodiments of the present invention, the embodiments of the present invention match voice information with voiceprint records based on user identifiers and voice data respectively, and perform corresponding processing according to the matching results. Through the embodiments of the present invention, rapid identification of fraudulent information is achieved, the timeliness of fraud identification is improved, and the matching results are accurately recorded.

[0116] Furthermore, the identification unit is also used for:

[0117] If the voiceprint record without fraud label meets the preset conditions, mark the person-card separation scenario label in the voiceprint record without fraud label;

[0118] The preset conditions include at least one of the following:

[0119] The number of voiceprint records with the same user identifier and a matching count greater than the first threshold exceeds the second threshold;

[0120] The number of voiceprint records with a matching count greater than the third threshold among the associated voiceprint records exceeds the fourth threshold.

[0121] Furthermore, the acquisition unit is also used to acquire user identifiers or voice data involved in fraud;

[0122] The identification unit is also used to filter out voiceprint records from the voiceprint database that match the fraudulent user identifier or voice data and mark them with fraud tags.

[0123] As can be seen from the technical solutions provided by the above embodiments of the present invention, the embodiments of the present invention, when the voiceprint records without fraud tags meet preset conditions, mark the person-card separation scenario tag in the voiceprint records without fraud tags. Through the embodiments of the present invention, rapid identification of fraud information is achieved, and the accuracy of fraud identification is improved.

[0124] The fraud identification device based on voiceprint technology provided in this embodiment of the invention can realize the various processes in the embodiments of the fraud identification method based on voiceprint technology described above. To avoid repetition, these processes will not be described again here.

[0125] It should be noted that the fraud identification device based on voiceprint technology provided in this embodiment of the invention and the fraud identification method based on voiceprint technology provided in this embodiment of the invention are based on the same inventive concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned fraud identification method based on voiceprint technology, and the repeated parts will not be described again.

[0126] Corresponding to the fraud identification method based on voiceprint technology provided in the above embodiments, based on the same technical concept, this embodiment of the invention also provides an electronic device for executing the above-described fraud identification method based on voiceprint technology. Figure 5 To illustrate the structure of an electronic device according to various embodiments of the present invention, as shown in the schematic diagram... Figure 5As shown. Electronic devices can vary considerably due to differences in configuration or performance, and may include one or more processors 501 and memory 502. Memory 502 may store one or more application programs or data. Memory 502 may be temporary or persistent storage. The application programs stored in memory 502 may include one or more modules (not shown), each module may include a series of computer-executable instructions for the electronic device. Furthermore, processor 501 may be configured to communicate with memory 502 and execute the series of computer-executable instructions in memory 502 on the electronic device. The electronic device may also include one or more power supplies 503, one or more wired or wireless network interfaces 504, one or more input / output interfaces 505, and one or more keyboards 506.

[0127] Specifically, in this embodiment, the electronic device includes a processor, a communication interface, a memory, and a communication bus; wherein, the processor, the communication interface, and the memory communicate with each other via the bus; the memory is used to store computer programs; the processor is used to execute the programs stored in the memory to implement the following method steps:

[0128] Obtain the voice information to be matched, the voice information including: user identifier and voice data;

[0129] Based on user identifier and / or voice data, search the voiceprint database for voiceprint records that match the voice information;

[0130] If a fraud tag is found in the voiceprint record that matches the voice information, then the voice information to be matched is determined to be fraudulent.

[0131] This application embodiment also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the following method steps:

[0132] Obtain the voice information to be matched, the voice information including: user identifier and voice data;

[0133] Based on user identifier and / or voice data, search the voiceprint database for voiceprint records that match the voice information;

[0134] If a fraud tag is found in the voiceprint record that matches the voice information, then the voice information to be matched is determined to be fraudulent.

[0135] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0136] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0137] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0138] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0139] In a typical configuration, an electronic device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0140] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0141] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0142] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0143] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0144] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A fraud identification method based on voiceprint technology, characterized in that, The method includes: Obtain the voice information to be matched, the voice information including: user identifier and voice data; Based on user identifier and / or voice data, search the voiceprint database for voiceprint records that match the voice information; If a fraud tag is found in the voiceprint record that matches the voice information, then the voice information to be matched is determined to be fraud information. The method further includes: If the voiceprint record without fraud label meets the preset conditions, mark the person-card separation scenario label in the voiceprint record without fraud label; The preset conditions include at least one of the following: The number of voiceprint records with the same user identifier and a matching count greater than the first threshold exceeds the second threshold; The number of voiceprint records with a matching count greater than the third threshold among the associated voiceprint records exceeds the fourth threshold.

2. The method according to claim 1, characterized in that, After searching a voiceprint database for a voiceprint record matching the voice information based on user identifier and / or voice data respectively, the method includes: Voiceprint records with the same user identifier as the voice information are selected from the voiceprint database as voiceprint records to be matched. Using a preset voice matching method, the voiceprint features of the voice data of the voice information are compared with the voiceprint features of the voice data of each voiceprint record to be matched. If no matching voiceprint record with the same voiceprint features as the voice information is found, a new voiceprint record is added to the voiceprint database according to the voice information, and the voiceprint features of the voice data of the new voiceprint record are compared with the voiceprint features of the voice data of other voiceprint records in the voiceprint database. Other voiceprint records with the same voiceprint features as the new voiceprint record are identified as voiceprint records that match the speech information and are associated with the new voiceprint record.

3. The method according to claim 2, characterized in that, The voiceprint record includes a number of matches. After comparing the voiceprint features of the voice data of the voice information with the voiceprint features of the voice data of each voiceprint record to be matched using a preset voice matching method, the method further includes: If it is determined that the voiceprint record to be matched has the same voiceprint features as the voice data of the voice information, the matching count of the voiceprint record to be matched is incremented by 1.

4. The method according to claim 1, characterized in that, The process of obtaining the voice information to be matched includes: The system collects signaling messages and voice packets during voice calls, and obtains the voice information to be matched by parsing the signaling messages and voice packets.

5. The method according to any one of claims 1-4, characterized in that, The method further includes: Obtaining user identifiers or voice data suspected of being fraudulent; Voiceprint records that match the fraudulent user identifier or voice data are filtered from the voiceprint database and labeled with fraud tags.

6. A fraud detection device based on voiceprint technology, characterized in that, The device includes: The acquisition unit is used to acquire the voice information to be matched, the voice information including: user identifier and voice data; The matching unit is used to search for voiceprint records that match the voice information from the voiceprint database based on the user identifier and / or voice data, respectively. The identification unit is used to determine that the voice information to be matched is fraudulent if a fraud tag exists in the voiceprint record that matches the voice information. The identification unit is also used to mark a person-card separation scenario label in the voiceprint record without a fraud label when the voiceprint record without a fraud label meets the preset conditions. The preset conditions include at least one of the following: The number of voiceprint records with the same user identifier and a matching count greater than the first threshold exceeds the second threshold; The number of voiceprint records with a matching count greater than the third threshold among the associated voiceprint records exceeds the fourth threshold.

7. The apparatus according to claim 6, characterized in that, The matching unit is used for: Voiceprint records with the same user identifier as the voice information are selected from the voiceprint database as voiceprint records to be matched. Using a preset voice matching method, the voiceprint features of the voice data of the voice information are compared with the voiceprint features of the voice data of each voiceprint record to be matched. If no matching voiceprint record with the same voiceprint features as the voice information is found, a new voiceprint record is added to the voiceprint database according to the voice information, and the voiceprint features of the voice data of the new voiceprint record are compared with the voiceprint features of the voice data of other voiceprint records in the voiceprint database. Other voiceprint records with the same voiceprint features as the new voiceprint record are identified as voiceprint records that match the speech information and are associated with the new voiceprint record.

8. An electronic device, characterized in that, The system includes a processor, a communication interface, a memory, and a communication bus; wherein the processor, the communication interface, and the memory communicate with each other via the bus; the memory is used to store computer programs; and the processor is used to execute the programs stored in the memory to implement the steps of the fraud identification method based on voiceprint technology as described in any one of claims 1-5.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the fraud identification method based on voiceprint technology as described in any one of claims 1-5.

Images