Anti-fraud system and method

Abstract

The present invention provides an anti-fraud method, comprising the following steps: obtaining voiceprint features according to the voice of the other party; matching and identifying the voiceprint features with the pre-stored voiceprint of the fraudster, and if the voiceprint features match the pre-stored voiceprint of the fraudster, generating a first-level alarm instruction; issuing a first-level alarm prompt according to the alarm instruction. The present invention can improve the efficiency of identifying fraudsters, and can solve the problem that the existing anti-fraud method is completed by marking the fraudster's phone number and has low anti-fraud efficiency. Accordingly, the present invention also provides an anti-fraud system.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an anti-fraud system and method. Background Art

[0002] Telephone fraud, which is the use of telephone to commit fraud, has more and more methods, making it difficult for people to guard against and causing great impact on people's lives.

[0003] Currently, fraud prevention methods are generally completed by marking the fraudsters' phone numbers. However, after the fraudsters realize that their phone numbers have been marked, they usually change their phone numbers to continue their fraud. Therefore, the efficiency of this method in preventing fraud is low and needs to be improved. Summary of the invention

[0004] Based on this, in order to solve the problem that the existing anti-fraud method is completed by marking the fraudsters' phone numbers and has low anti-fraud efficiency, the present invention provides an anti-fraud system and method, and its specific technical solution is as follows:

[0005] An anti-fraud system includes a mobile terminal, wherein the mobile terminal includes an acquisition module, an analysis module and an alarm module.

[0006] The acquisition module is used to acquire voiceprint features according to the conversation voice of the other party.

[0007] The analysis module is used to match and identify the voiceprint feature with the pre-stored voiceprint of the fraudster, and if the voiceprint feature matches the pre-stored voiceprint of the fraudster, a first-level alarm instruction is generated.

[0008] The alarm module is used to issue a first-level alarm prompt according to the first-level alarm instruction.

[0009] By obtaining the voiceprint features of the other party and comparing the voiceprint features with the pre-stored fraudster voiceprints, it is possible to accurately and efficiently determine whether the other party is a fraudster. In this way, when the voiceprint features of the other party match the pre-stored fraudster voiceprint, even if the other party changes their mobile phone number, it can be determined in real time that the other party is a fraudster, and the user can be reminded of the fraudulent behavior of the other party in a timely manner, reducing the probability of the user being deceived.

[0010] That is to say, the anti-fraud system solves the problem of low anti-fraud efficiency in existing anti-fraud methods that are completed by marking the fraudsters' phone numbers, and can improve the anti-fraud efficiency and reduce the probability of being deceived.

[0011] The pre-stored fraudster's voiceprint is stored in a cloud server, and the analysis module obtains the pre-stored fraudster's voiceprint from the cloud server.

[0012] Furthermore, the acquisition module is further used to acquire specific voice information according to the call voice of the other party, and the analysis module is further used to match and identify the specific voice information with the pre-stored voice information of the fraudster, and if the specific voice information matches the pre-stored voice information of the fraudster, a first-level alarm instruction is generated;

[0013] The pre-stored voice information of the fraudster is stored in the cloud server, and the analysis module obtains the pre-stored voice information of the fraudster from the cloud server.

[0014] Furthermore, the analysis module includes a keyword extraction unit and a first matching identification unit.

[0015] The keyword extraction unit is used to extract multiple first keywords from the specific voice information and multiple second keywords from the pre-stored fraudster voice information.

[0016] The first matching and identification module is used to match and identify multiple first keywords and multiple second keywords. If the matching ratio of multiple first keywords to multiple second keywords is greater than a preset matching threshold, it is determined that the specific voice information matches the pre-stored fraudster voice information and a first-level alarm instruction is generated.

[0017] Furthermore, the analysis module also includes a text extraction unit and a second matching recognition unit.

[0018] The text extraction unit is used to extract the first text information of the specific voice information and the second text information of the pre-stored fraudster voice information.

[0019] The second matching identification unit is used to calculate the similarity between the first text information and the second text information. If the similarity is greater than a first preset similarity threshold, it is determined that the specific voice information matches the pre-stored fraudster voice information and a first-level alarm instruction is generated.

[0020] Furthermore, a fraud prevention method comprises the following steps:

[0021] Acquire voiceprint features based on the other party's voice;

[0022] Matching and identifying the voiceprint feature with the pre-stored voiceprint of the fraudster, and generating a first-level alarm instruction if the voiceprint feature matches the pre-stored voiceprint of the fraudster;

[0023] A first-level alarm prompt is issued according to the first-level alarm instruction.

[0024] Furthermore, the anti-fraud method further comprises the following steps:

[0025] Obtaining specific voice information based on the other party's voice;

[0026] The specific voice information is matched and identified with the pre-stored voice information of the fraudster. If the specific voice information matches the pre-stored voice information of the fraudster, a first-level alarm instruction is generated.

[0027] Furthermore, the specific voice information is matched and identified with the pre-stored voice information of the fraudster, and if the specific voice information matches the pre-stored voice information of the fraudster, the specific method of generating a first-level alarm instruction includes the following steps:

[0028] Extracting a plurality of first keywords from the specific voice information and a plurality of second keywords from the pre-stored fraudster voice information;

[0029] Match and identify multiple first keywords and multiple second keywords. If the matching ratio of multiple first keywords to multiple second keywords is greater than a preset matching threshold, it is determined that the specific voice information matches the pre-stored fraudster voice information and a first-level alarm instruction is generated.

[0030] Furthermore, the specific voice information is matched and identified with the pre-stored voice information of the fraudster, and if the specific voice information matches the pre-stored voice information of the fraudster, the specific method of generating a first-level alarm instruction also includes the following steps:

[0031] Extracting the first text information of the specific voice information and the second text information of the pre-stored fraudster voice information;

[0032] The similarity between the first text information and the second text information is calculated. If the similarity is greater than a first preset similarity threshold, it is determined that the specific voice information matches the pre-stored fraudster voice information and a first-level alarm instruction is generated.

[0033] Furthermore, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the anti-fraud method is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The present invention can be further understood from the following description in conjunction with the accompanying drawings. The components in the figures are not necessarily drawn to scale, but the emphasis is placed on illustrating the principles of the embodiments. In different views, the same reference numerals designate corresponding parts.

[0035] Figure 1 This is a schematic diagram of the overall process of an anti-fraud method in one embodiment of the present invention. Figure 1 ;

[0036] Figure 2 This is a schematic diagram of the overall process of an anti-fraud method in one embodiment of the present invention. Figure 2 . DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with its embodiments. It should be understood that the specific implementation methods described herein are only used to explain the present invention and do not limit the protection scope of the present invention.

[0038] It should be noted that when an element is referred to as being "fixed to" another element, it may be directly on the other element or there may be a central element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may be a central element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only and are not intended to be the only implementation method.

[0039] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0040] The “first” and “second” mentioned in the present invention do not represent specific quantities and orders, but are merely used to distinguish names.

[0041] An anti-fraud system in one embodiment of the present invention includes a mobile terminal, wherein the mobile terminal includes an acquisition module, an analysis module, and an alarm module.

[0042] The acquisition module is used to acquire voiceprint features according to the conversation voice of the other party.

[0043] The analysis module is used to match and identify the voiceprint feature with the pre-stored voiceprint of the fraudster, and if the voiceprint feature matches the pre-stored voiceprint of the fraudster, a first-level alarm instruction is generated.

[0044] The alarm module is used to issue a first-level alarm prompt according to the first-level alarm instruction. Specifically, the alarm prompt includes but is not limited to an alarm text message and an alarm voice.

[0045] The pre-stored fraudster's voiceprint is stored in a cloud server, the mobile terminal is in communication connection with the cloud server, and the analysis module obtains the pre-stored fraudster's voiceprint from the cloud server.

[0046] By obtaining the voiceprint features of the other party and comparing the voiceprint features with the pre-stored fraudster voiceprints, it is possible to accurately and efficiently determine whether the other party is a fraudster. In this way, when the voiceprint features of the other party match the pre-stored fraudster voiceprint, even if the other party changes their mobile phone number, it can be determined in real time that the other party is a fraudster, and the user can be reminded of the fraudulent behavior of the other party in a timely manner, reducing the probability of the user being deceived.

[0047] That is to say, the anti-fraud system solves the problem of low anti-fraud efficiency in existing anti-fraud methods that are completed by marking the fraudsters' phone numbers, and can improve the anti-fraud efficiency and reduce the probability of being deceived.

[0048] In one of the embodiments, the acquisition module is also used to obtain specific voice information based on the call voice of the other party, and the analysis module is also used to match and identify the specific voice information with the pre-stored voice information of the scammer. If the specific voice information matches the pre-stored voice information of the scammer, a first-level alarm instruction is generated.

[0049] The pre-stored voice information of the fraudster is stored in the cloud server, and the analysis module obtains the pre-stored voice information of the fraudster from the cloud server.

[0050] Specifically, the analysis module includes a keyword extraction unit and a first matching identification unit.

[0051] The keyword extraction unit is used to extract multiple first keywords from the specific voice information and multiple second keywords from the pre-stored fraudster voice information.

[0052] Since extracting keywords based on speech recognition is a conventional technical means in this field, it will not be described in detail here.

[0053] The first matching and identification module is used to match and identify multiple first keywords and multiple second keywords. If the matching ratio of multiple first keywords to multiple second keywords is greater than a preset matching threshold, it is determined that the specific voice information matches the pre-stored fraudster voice information and a first-level alarm instruction is generated.

[0054] For example, the number of second keywords in the pre-stored fraudster voice information is N, and the number of first keywords in the specific voice information is M. The specific method for matching and identifying the M first keywords and the N second keywords may be: matching and identifying each of the first keywords with the N second keywords, if L first keywords match one of the second keywords, then the matching ratio is L / M.

[0055] When the matching ratio is greater than a preset matching threshold, it can be considered that the specific voice information matches the pre-stored fraudster voice information.

[0056] In one of the embodiments, the analysis module further includes a text extraction unit and a second matching recognition unit.

[0057] The text extraction unit is used to extract the first text information of the specific voice information and the second text information of the pre-stored fraudster voice information.

[0058] The second matching identification unit is used to calculate the similarity between the first text information and the second text information. If the similarity is greater than a first preset similarity threshold, it is determined that the specific voice information matches the pre-stored fraudster voice information and a first-level alarm instruction is generated.

[0059] Here, the first text information and the second text information are both within a preset character string length range, for example, the character string lengths of the first text information and the second text information are limited to within 10. In this way, by limiting the character string lengths of the first text information of the specific voice information and the second text information of the pre-stored fraudster voice information, the similarity between the first text information and the second text information can be more accurately calculated, thereby improving the recognition efficiency and accuracy of the specific voice information.

[0060] In one embodiment, the mobile terminal includes a plurality of mobile terminals, and the anti-fraud system further includes a trusted server that is communicatively connected to the plurality of mobile terminals. Specifically, the trusted server may be a backend server of a public security agency or a backend server of a communication operator.

[0061] The multiple mobile terminals send their own address book lists to a trusted server in an encrypted manner based on a commonly recognized protocol or an agreed protocol. The trusted server decrypts the encrypted address book lists of the multiple mobile terminals. Here, the encryption and decryption method is preferably an asymmetric encryption and decryption method such as RSA. The multiple mobile terminals encrypt their own address book lists with public keys, and the trusted server decrypts the encrypted address book lists with private keys. In this way, the mobile terminal address book lists can be better kept confidential to avoid personal privacy leaks.

[0062] When the mobile terminal determines that the other party is making the first call, it obtains the other party's phone number and sends the other party's phone number and a verification request to the trusted server. The trusted server traverses the address book lists of multiple mobile terminals based on the verification request and the other party's phone number. If the other party's phone number is not in the address book lists of multiple mobile terminals, a secondary alarm indication is generated and fed back to the corresponding mobile terminal.

[0063] If the mobile terminal receives the secondary alarm instruction and the similarity is greater than the second preset similarity threshold and less than the first preset similarity threshold, a primary alarm instruction is generated; if the mobile terminal receives the secondary alarm instruction and the similarity is less than the second preset similarity threshold, a secondary alarm instruction is generated and a secondary alarm prompt is issued according to the secondary alarm instruction.

[0064] The second preset similarity threshold is less than the first preset similarity threshold. The second level alarm prompt is a moderate fraud suspicion alarm prompt, and the first level alarm prompt is a severe fraud suspicion alarm prompt.

[0065] Since fraudsters may change their words to avoid the fraud system's matching and identification of the specific voice information of the call voice with the pre-stored fraudster's voice information, when the similarity is greater than the second preset similarity threshold and less than the first preset similarity threshold, the other party in the call may also be a fraudster.

[0066] At this time, the trusted server traverses the address book lists of multiple mobile terminals to match and identify the phone number of the other party of the call, and if the phone number of the other party of the call is not in the address book lists of multiple mobile terminals, it generates a secondary alarm indication and feeds back the secondary alarm indication to the corresponding mobile terminal. The mobile terminal then generates a primary alarm instruction and a secondary alarm instruction according to the secondary alarm indication and the relationship between the similarity and the second preset similarity threshold and the first preset similarity threshold. The fraudsters can be further identified and judged, and the user can be promptly reminded of the fraudulent behavior of the other party of the call, thereby reducing the probability of the user being deceived.

[0067] Since the phone number of a fraudster may be stored in the address book list of one of the mobile terminals due to mistakenly storing an unfamiliar phone number, etc., at this time, only using the address book lists of multiple mobile terminals to match and identify the phone number of the other party, the problem that the other party is a fraudster and the trusted server cannot generate a secondary alarm indication may occur.

[0068] In order to solve the above problem, in one embodiment, the multiple mobile terminals also send their own WeChat address book friend lists to a trusted server in an encrypted manner based on a commonly recognized protocol or an agreed protocol. The trusted server decrypts the encrypted WeChat address book friend lists of the multiple mobile terminals.

[0069] When the mobile terminal determines that the call is the first call, it obtains the phone number of the other party and sends the phone number of the other party and a verification request to the trusted server. The trusted server traverses the address book lists of multiple mobile terminals and the WeChat address book friend lists of multiple mobile terminals according to the verification request and the phone number of the other party. If the phone number of the other party is not in the address book lists of multiple mobile terminals and the WeChat address book friend lists of the multiple mobile terminals, a secondary alarm indication is directly generated and the secondary alarm indication is fed back to the corresponding mobile terminal.

[0070] In this way, even if the address book list of one of the mobile terminals may store the phone number of a scammer due to the mistaken storage of an unfamiliar phone number, etc., since the WeChat account corresponding to the phone number of the other party of the call is not in the WeChat address book friend list of multiple mobile terminals, the trusted server will still generate a secondary alarm indication, avoiding the problem that the other party of the call is a scammer and the trusted server cannot generate a secondary alarm indication.

[0071] That is to say, by combining the address book lists of multiple mobile terminals and the WeChat address book friend lists to identify the first incoming call, it is possible to more accurately identify whether the other party of the call is a scammer.

[0072] In one embodiment, if Figure 1 As shown, a fraud prevention method comprises the following steps:

[0073] S1, obtaining voiceprint features according to the voice of the other party of the call.

[0074] S2, matching and identifying the voiceprint feature with the pre-stored voiceprint of the fraudster, and generating a first-level alarm instruction if the voiceprint feature matches the pre-stored voiceprint of the fraudster.

[0075] S3, issuing a first-level alarm prompt according to the first-level alarm instruction. The alarm prompt includes but is not limited to a text message or an alarm voice.

[0076] Specifically, the pre-stored fraudster's voiceprint is stored in a cloud server, and the analysis module of the mobile terminal obtains the pre-stored fraudster's voiceprint from the cloud server.

[0077] By obtaining the voiceprint features of the other party and comparing the voiceprint features with the pre-stored voiceprints of the fraudsters, the anti-fraud method can accurately and efficiently determine whether the other party is a fraudster. In this way, when the voiceprint features of the other party match the pre-stored voiceprints of the fraudsters, even if the other party changes their mobile phone number, it can be accurately determined in real time that the other party is a fraudster, and the user can be reminded of the fraudulent behavior of the other party in time, reducing the probability of the user being deceived.

[0078] That is to say, the anti-fraud method improves the efficiency of identifying fraudsters, and can solve the problem of low anti-fraud efficiency in existing anti-fraud methods that are completed by marking the fraudsters' phone numbers.

[0079] In one embodiment, if Figure 2 As shown, the anti-fraud method also includes the following steps:

[0080] S4, obtaining specific voice information according to the call voice of the other party.

[0081] S5, matching and identifying the specific voice information with the pre-stored voice information of the fraudster, if the specific voice information matches the pre-stored voice information of the fraudster, generating a first-level alarm instruction.

[0082] Specifically, in step S5, the specific voice information is matched and identified with the pre-stored voice information of the fraudster. If the specific voice information matches the pre-stored voice information of the fraudster, a first-level alarm instruction is generated. The specific method also includes the following steps:

[0083] S50, extracting the first text information of the specific voice information and the second text information of the pre-stored fraudster's voice information.

[0084] S51, calculating the similarity between the first text information and the second text information, if the similarity is greater than a first preset similarity threshold, determining that the specific voice information matches the pre-stored fraudster voice information and generating a first-level alarm instruction.

[0085] Alternatively, in step S5, the specific voice information is matched and identified with the pre-stored voice information of the fraudster, and if the specific voice information matches the pre-stored voice information of the fraudster, a first-level alarm instruction is generated. The specific method also includes the following steps:

[0086] S52, extracting a plurality of first keywords from the specific voice information and a plurality of second keywords from the pre-stored fraudster voice information.

[0087] S53, matching and identifying the plurality of the first keywords and the plurality of the second keywords, if the matching ratio of the plurality of the first keywords to the plurality of the second keywords is greater than a preset matching threshold, it is determined that the specific voice information matches the pre-stored fraudster voice information and a first-level alarm instruction is generated.

[0088] Specifically, each of the first keywords is compared and matched with a plurality of the second keywords one by one to identify whether each of the first keywords matches one of the second keywords, and the number of the first keywords matching one of the second keywords is calculated L. Assuming that the number of the first keywords is M, the matching ratio is L / M.

[0089] By matching and identifying the voiceprint feature with the pre-stored voiceprint of the fraudster and the specific voice information with the pre-stored voice information of the fraudster, the identity of the other party can be further identified to determine whether the other party is a fraudster. In this way, the anti-fraud method can further improve the efficiency of identifying fraudsters, and solve the problem of low anti-fraud efficiency in the existing anti-fraud method that is completed by marking the fraudster's phone number.

[0090] In one embodiment, a computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the anti-fraud method is implemented.

[0091] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0092] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. An anti-fraud system, comprising a mobile terminal, characterized in that: The mobile terminal comprises: An acquisition module is used to acquire voiceprint features based on the voice of the other party of the call; An analysis module is used to match and identify the voiceprint feature with the pre-stored voiceprint of the fraudster, and if the voiceprint feature matches the pre-stored voiceprint of the fraudster, a first-level alarm instruction is generated; When the mobile terminal determines that the call is the first call, the phone number of the other party is obtained and the phone number of the other party and a verification request are sent to the trusted server. The trusted server traverses the address book lists of multiple mobile terminals according to the verification request and the phone number of the other party. If the phone number of the other party is not in the address book lists of multiple mobile terminals, a secondary alarm indication is generated and the secondary alarm indication is fed back to the corresponding mobile terminal; if the mobile terminal receives the secondary alarm instruction and the similarity is greater than the second preset similarity threshold and less than the first preset similarity threshold, a primary alarm instruction is generated; If the mobile terminal receives the secondary alarm instruction and the similarity is less than a second preset similarity threshold, a secondary alarm instruction is generated and a secondary alarm prompt is issued according to the secondary alarm instruction; wherein the second preset similarity threshold is less than the first preset similarity threshold, the secondary alarm prompt is a moderate fraud suspicion alarm prompt, the first alarm prompt is a severe fraud suspicion alarm prompt, the similarity is calculated by the first text information and the second text information, the first text information is information extracted from specific voice information, the specific voice information is obtained by the acquisition module according to the call voice of the other party, and the second text information is the text information of the pre-stored fraudster's voice information; An alarm module, used for issuing a first-level alarm prompt according to the first-level alarm instruction; The pre-stored fraudster's voiceprint is stored in a cloud server, and the analysis module obtains the pre-stored fraudster's voiceprint from the cloud server.

2. An anti-fraud system as claimed in claim 1, characterized in that: The acquisition module is further used to acquire specific voice information according to the call voice of the other party, and the analysis module is further used to match and identify the specific voice information with the pre-stored voice information of the fraudster, and if the specific voice information matches the pre-stored voice information of the fraudster, a first-level alarm instruction is generated; The pre-stored voice information of the fraudster is stored in the cloud server, and the analysis module obtains the pre-stored voice information of the fraudster from the cloud server.

3. An anti-fraud system as claimed in claim 2, characterized in that: The analysis module comprises: A keyword extraction unit, used to extract a plurality of first keywords from the specific voice information and a plurality of second keywords from the pre-stored fraudster voice information; The first matching and identifying unit is used to match and identify the plurality of the first keywords and the plurality of the second keywords. If the matching ratio of the plurality of the first keywords to the plurality of the second keywords is greater than a preset matching threshold, it is determined that the specific voice information matches the pre-stored fraudster voice information and a first-level alarm instruction is generated.

4. An anti-fraud system as claimed in claim 3, characterized in that: The analysis module also includes: A text extraction unit, used to extract the first text information of the specific voice information and the second text information of the pre-stored fraudster voice information; The second matching identification unit is used to calculate the similarity between the first text information and the second text information. If the similarity is greater than a first preset similarity threshold, it is determined that the specific voice information matches the pre-stored fraudster voice information and a first-level alarm instruction is generated.

5. A fraud prevention method, applied to the fraud prevention system as claimed in any one of claims 1 to 4, characterized in that: The steps include: Acquire voiceprint features based on the other party's voice; Matching and identifying the voiceprint feature with the pre-stored voiceprint of the fraudster, and generating a first-level alarm instruction if the voiceprint feature matches the pre-stored voiceprint of the fraudster; A first-level alarm prompt is issued according to the first-level alarm instruction.

6. An anti-fraud method as claimed in claim 5, characterized in that: The following steps are also included: Obtaining specific voice information based on the other party's voice; The specific voice information is matched and identified with the pre-stored voice information of the fraudster. If the specific voice information matches the pre-stored voice information of the fraudster, a first-level alarm instruction is generated.

7. An anti-fraud method as claimed in claim 6, characterized in that: The specific method of matching and identifying the specific voice information with the pre-stored voice information of the fraudster, and if the specific voice information matches the pre-stored voice information of the fraudster, then generating a first-level alarm instruction comprises the following steps: Extracting a plurality of first keywords from the specific voice information and a plurality of second keywords from the pre-stored fraudster voice information; Match and identify multiple first keywords and multiple second keywords. If the matching ratio of multiple first keywords to multiple second keywords is greater than a preset matching threshold, it is determined that the specific voice information matches the pre-stored fraudster voice information and a first-level alarm instruction is generated.

8. An anti-fraud method as claimed in claim 7, characterized in that: The specific method of matching and identifying the specific voice information with the pre-stored voice information of the fraudster, and if the specific voice information matches the pre-stored voice information of the fraudster, then generating a first-level alarm instruction also includes the following steps: Extracting the first text information of the specific voice information and the second text information of the pre-stored fraudster voice information; The similarity between the first text information and the second text information is calculated. If the similarity is greater than a first preset similarity threshold, it is determined that the specific voice information matches the pre-stored fraudster voice information and a first-level alarm instruction is generated.

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 anti-fraud method according to any one of claims 5 to 8.

Images