System and method for preventing junk calls of smart phone

Abstract

The invention discloses a system and a method for preventing junk calls of a smart phone. The smart phone is accessed to a telephone network and the Internet so as to receive and send communication signals; the incoming call management assistant operates on the smart phone and cooperates with the junk incoming call prevention center to identify whether the strange incoming call is a junk incoming call; the junk call prevention center runs at a cloud end, and comprises the following steps: an intelligent voice processing engine obtains voiceprint information and semantic information of an incoming call user by analyzing incoming call voice data, and simulates the voice of an owner based on the voiceprint information of the user to carry out voice speaking; the big data processing module counts spam incoming call information reported by the incoming call management assistant, ranks based on spam incoming call frequency reported by each region, updates a TOPN spam incoming call number table, and periodically synchronizes the spam incoming call number table to the currently online incoming call management assistant; and judging whether the incoming call voice is a junk call. According to the invention, junk calls can be intercepted more effectively, the user experience is greatly improved, and the communication security of the smart phone is enhanced.

Description

Technical Field

[0001] The invention belongs to the technical field of communication applications, and in particular relates to a system and method for preventing spam calls from smart phones. Background Art

[0002] With the popularization of mobile communication technology, smartphones, as the main communication tool for people's daily life and work, provide convenience for various social applications, but also bring personal data leakage, resulting in a sharp increase in various junk calls for individual users, such as harassment calls, advertising calls, fraud calls, etc. Many mobile phone users are overwhelmed by these calls, and a large number of people are deceived and suffer huge property losses. There are also some call blocking software on the market, but they are all implemented by setting various blocking rules and constantly updating various blocking rules. The blocking method is relatively single and cannot cope with various constantly updated junk calls.

[0003] At present, smartphones mainly adopt judgment rules based on the statistical results of caller numbers in terms of preventing spam calls. After installing software similar to firewalls on mobile terminals, they deal with it by constantly adding blacklists and whitelists. This method is very passive and cumbersome. Mobile phone users need to constantly update various lists to deal with various changing phone numbers. Therefore, it is necessary to propose a more effective anti-harassment fraud solution for mobile phone users. Summary of the invention

[0004] The technical problem to be solved by the present invention is to address the deficiencies of the above-mentioned prior art and to provide a system and method for preventing spam calls in a smart phone. The system and method can use intelligent voice processing technology and big data statistical capabilities to identify caller information such as caller numbers, caller voiceprints, and caller semantics, without affecting the normal use of the device by the owner and without distracting the owner's daily attention, so as to more actively, accurately, intelligently, and effectively intercept fraud calls, advertising calls, and other types of spam calls that are invalid or even harmful to the owner, so that the spam calls are intercepted in the early stages of communication, greatly improving the user experience and enhancing the communication security of smart phones.

[0005] In order to achieve the above technical objectives, the technical solution adopted by the present invention is:

[0006] A system for preventing spam calls on a smart phone, comprising a smart phone, a call management assistant and a spam call prevention center;

[0007] Among them, smartphones are used to access telephone networks and the Internet to receive and send communication signals;

[0008] The call management assistant runs on the smartphone and works with the spam call prevention center to identify whether an unfamiliar call is a spam call;

[0009] The spam call prevention center runs on the cloud, including an intelligent voice processing engine and a big data processing module;

[0010] Intelligent voice processing engine, used to obtain the voiceprint information and semantic information of the caller through analysis of the incoming call voice data, and to imitate the voice of the phone owner to speak based on the user's voiceprint information;

[0011] The big data processing module is used to count the junk call information reported by the call management assistant, rank and update the TOPN junk call number table based on the frequency of junk calls reported by each region, and regularly synchronize it to the currently online call management assistant; and determine whether the incoming call voice is a junk call.

[0012] To optimize the above technical solutions, the specific measures taken also include:

[0013] The above-mentioned incoming call management assistant calls the underlying network interface of the smartphone operating system when the owner is on a call, and only allows the data flow of the incoming call application to be sent to the communication network. The data flows of other applications in the mobile phone sent to the communication network are temporarily prohibited from passing. This prevents malicious programs lurking in the mobile phone from recording the owner's call content and sending it to the external network. While ensuring that the owner's communication is not eavesdropped on, it also reduces the bombardment and harassment of various junk messages and junk calls due to the leakage of personal information from the source.

[0014] The above-mentioned spam call prevention center provides a whitelist number registration service for real-name authenticated personnel who provide public services such as express delivery, food delivery, and taxi hailing. Phone numbers in the whitelist number database will not be identified as spam calls.

[0015] The aforementioned spam call prevention center also includes a terminal association and data backup module, which provides a cloud backup function for key data for the call management assistant;

[0016] The data locally managed by the call management assistant include: the local address book, the local whitelist caller number table, the local blacklist caller number table, the local whitelist caller voiceprint ID table, the local blacklist caller voiceprint ID table, and the local preset blacklist caller semantic library. When the locally managed data is updated, it is synchronously updated to the terminal association and data backup module of the spam call prevention center so that the owner can still use the existing relevant spam call information after changing the mobile phone.

[0017] The above-mentioned smartphones allow the owner to answer unfamiliar calls and then, based on his or her own subjective judgment, set and report historical calls as spam calls in the call management assistant. For call records reported as spam calls, the call management assistant will upload the owner's number, the reported caller number, the reported call recording data and the caller voiceprint ID information to the spam call prevention center, and at the same time add the corresponding caller number to the local blacklist caller number table and the caller voiceprint ID to the local blacklist caller voiceprint ID table.

[0018] After receiving the report information from the call management assistant, the above-mentioned spam call prevention center first sends the relevant call recording to the intelligent voice processing engine for voiceprint recognition and semantic processing. Then, based on the processing results of the intelligent voice processing engine, it conducts big data statistics of spam calls in the area where the owner is located, updates the TOPN table of spam call numbers in the area, identifies the reported call voiceprint ID as a spam call, and updates the semantic library of spam calls in the entire network;

[0019] The big data processing module regularly ranks the frequency of junk call reports reported by each region, updates the TOPN junk call number table, and then synchronously updates it to the call management assistant in each smart phone.

[0020] A method for preventing spam calls on a smart phone, comprising the following steps:

[0021] Step 1. Smartphone accesses the mobile network for voice calls and Internet access;

[0022] Step 2. The smartphone starts the call management assistant. After starting, the call management assistant first performs data initialization. After the initialization is completed, it enters the call monitoring state. When the call management assistant monitors an incoming call, it compares the incoming call number with the whitelist caller number table and the blacklist caller number table for identification;

[0023] Step 3. If the incoming call number monitored by the call management assistant is identified as being in the whitelist phone number table, the smartphone is allowed to remind the owner to answer the call in a normal incoming call reminder manner;

[0024] Step 4. If the incoming call number monitored by the call management assistant is in the blacklist, the call is directly rejected or hung up after answering;

[0025] Step 5. If the incoming call number monitored by the call management assistant is identified as an unfamiliar call number, the mobile phone will answer the call silently and then call the intelligent voice processing engine to first imitate the owner's voice to speak, then keep answering the call silently and collect and record the audio signal of the caller, and at the same time send the audio data stream to the intelligent voice processing engine of the spam call prevention center, and obtain the current caller voiceprint ID, caller semantics and semantic keywords by calling the intelligent voice processing engine and obtain whether it is marked as a "spam call" through the big data processing module. Then the call management assistant compares the caller voiceprint ID, caller semantics and semantic keywords with the whitelist caller voiceprint ID table, blacklist caller voiceprint ID table, and the blacklist caller semantic library information preset by the local machine, and judges whether the call is a spam call based on whether it is marked as a "spam call" and takes corresponding actions.

[0026] The data initialization described in step 2 above specifically includes: reading the local address book and the local call history record of the mobile phone, and requesting the spam call prevention center to obtain the blacklist and whitelist call information table, including:

[0027] (1) A whitelist of incoming call numbers, including a list of phone numbers in the phone's address book, phone numbers in the phone's call history with a call duration of more than N seconds, a list of public service phone numbers, and a list of public service personnel phone numbers in the owner's area preset by the call prevention center;

[0028] (2) Blacklist caller number list, including the blacklist phone number list set by the phone owner and the top 10 junk caller number list in the area where the phone owner is located;

[0029] (3) The local whitelist caller voiceprint ID table is a list of voiceprint IDs of each contact in the local address book; the local whitelist caller voiceprint ID table is established by cloud collaborative collection, and the specific steps are as follows:

[0030] The call management assistant collects the data of the incoming call voice audio signal and uploads it to the intelligent voice processing engine of the spam call prevention center; the intelligent voice processing engine uses voiceprint recognition technology to extract voiceprint information from the audio signal data and share it across the network. When the caller number corresponding to the voiceprint ID appears as a contact in the address books of N machine owners, the voiceprint ID of the contact corresponding to the caller number in the address books of these N machine owners is completed, thereby quickly establishing a voiceprint ID table of the contacts in the address book of the machine owner, that is, the whitelist caller voiceprint ID table of the machine owner; wherein, the call management assistant collects the data of the incoming call voice audio signal and uploads it to the intelligent voice processing engine of the spam call prevention center, including: when the call management assistant is started for the first time, it reminds the machine owner to use the call management assistant to actively record the voice clip read by the machine owner himself and upload it to the intelligent voice processing engine of the spam call prevention center; when the user A using the call management assistant answers the call of the incoming caller user B, the voice signal of the user B is recorded and uploaded to the intelligent voice processing engine of the spam call prevention center;

[0031] (4) The local blacklist caller voiceprint ID table is the voiceprint ID of the corresponding user from the blacklist phone number table set by the local device;

[0032] (5) The machine has a preset blacklist call semantic library, including common keywords for advertising, promotion, and fraud.

[0033] In the above step 5, the intelligent voice processing engine uses voiceprint recognition technology to extract voiceprint information and caller semantics and semantic keywords from the audio signal data, and uses the big data processing module to determine whether the incoming call voice is a spam call, and obtains the caller voiceprint ID, caller semantics and semantic keywords, and whether it is marked as a "spam call". The specific steps are as follows:

[0034] (1) The call management assistant sends the collected and recorded call audio data, caller ID, phone number of the owner, and the owner's reporting information to the intelligent voice processing engine;

[0035] (2) The intelligent voice processing engine extracts the voiceprint information of the incoming call and the caller semantics and semantic keywords of the incoming call from the received audio data; the voiceprint information includes the following data:

[0036] The caller number that sent this voiceprint information;

[0037] MFCC feature vector: a vector representing the energy distribution of the speech signal in the Mel frequency domain;

[0038] Fundamental frequency spectrum: It is an image showing the change of fundamental frequency spectrum over time, where the horizontal axis represents time, the vertical axis represents frequency, and the depth of color represents the strength of energy;

[0039] Speech envelope curve: a curve that describes the change of speech signal amplitude over time, used in conjunction with the fundamental frequency spectrogram;

[0040] Formant frequency list: a series of values ​​representing the frequency of formants in the speech signal, which change over time to form a trajectory;

[0041] (3) The intelligent voice processing engine sends the extracted voiceprint information, caller semantics and semantic keywords, and owner report information to the big data processing module;

[0042] (4) After receiving the data sent by the intelligent voice processing engine, the big data processing module further determines whether it is a spam call, and specifically makes the following judgment:

[0043] 1) Compare and match the voiceprint information with the whole network voiceprint database maintained by the spam call management center. If the similarity with any four of the five types of voiceprint feature information described in 1)-5) above reaches 80%, it means the match is successful, otherwise it means failure; if it does not match the data of all voiceprint IDs, then the whole network voiceprint database does not have this voiceprint information, and a new voiceprint ID is generated for this voiceprint information and this voiceprint data and its caller number are added to the whole network voiceprint database; if it matches the data of a certain voiceprint ID, check whether this voiceprint ID has been identified as a spam call voiceprint. If it has been identified as a spam call voiceprint, add this caller number to the spam caller number library. If this caller number already exists in the spam caller number library, add 1 to the statistical value of this spam caller number, and then reply this voiceprint ID and the identification result of "spam call voiceprint" to the intelligent voice processing engine to end this operation;

[0044] 2) Match the semantic keywords of the incoming call with the semantic database of spam calls in the entire network. If the similarity exceeds the threshold, the incoming call is judged as a spam call, and the voiceprint ID is marked as "spam call voiceprint" in the voiceprint database of the entire network. At the same time, the incoming call number is added to the database of spam calls in the entire network, and the voiceprint ID and the "spam call" identification result are then returned to the intelligent voice processing engine to end this operation;

[0045] 3) If the phone owner reports the call as a spam call and it is the first time for the phone owner to report, then the "Number of spam call reports" field of this voiceprint ID in the whole network voiceprint database is increased by 1, and the phone number of the phone owner is added to the "Junk call reporter" field, and the text content of the semantics of this call is saved in the "Semantic content of the call when reporting" field under this voiceprint ID entry; when the number of spam reports exceeds N digits, this voiceprint ID is marked as "spam call voiceprint" in the whole network voiceprint database, and the caller number is added to the whole network spam call number library, and the analysis of the semantic keywords of this spam call is started, and the analyzed keywords are added to the whole network spam call semantic library;

[0046] (5) The intelligent voice processing engine replies to the call management assistant with the caller ID, caller voiceprint ID, caller semantics and semantic keywords, and whether the call is marked as a "spam call."

[0047] In the above step 5, the call management assistant compares the caller voiceprint ID, caller semantics and semantic keywords with the whitelist caller voiceprint ID table, blacklist caller voiceprint ID table, and the local preset blacklist caller semantic library information to determine whether the call is a spam call and take corresponding actions, including:

[0048] 5.1) If the caller's voiceprint ID is in the whitelist caller voiceprint ID table, the call management assistant will play a normal call prompt tone or a user-defined call prompt tone to prompt the owner to answer the call;

[0049] 5.2) If the caller's voiceprint ID has been identified as a "spam call" by the intelligent voice processing engine of the spam call prevention center, the call management assistant will directly hang up the call;

[0050] 5.3) If the incoming caller's voiceprint ID is in the blacklisted incoming caller's voiceprint ID table, the incoming call management assistant will hang up the call directly;

[0051] 5.4) If the incoming caller's voiceprint ID is neither in the local whitelist incoming caller voiceprint ID table nor in the local blacklist incoming caller voiceprint ID table, but the incoming caller's semantic keywords have been included in the local preset blacklist incoming caller semantic library, the incoming call management assistant will directly hang up the call, and at the same time add the incoming caller number to the local blacklist incoming caller number table and the incoming caller's voiceprint ID to the local blacklist incoming caller voiceprint ID table;

[0052] 5.5) If the voiceprint ID of the incoming call does not belong to the above 5.1)-5.4), the call management assistant will play a normal incoming call prompt tone or a user-defined incoming call prompt tone to prompt the owner to answer the call.

[0053] The present invention has the following beneficial effects:

[0054] 1. The system of the present invention is based on the technical architecture of end-cloud collaborative processing, which fully exploits the advantages of high-performance voice data processing capabilities in the cloud, making the voice processing results and junk call judgment more comprehensive and accurate. The present invention can not only determine whether the incoming call is a junk call based on the big data statistics of the incoming call number, but also determine whether the unfamiliar call is a junk call by extracting the caller's voiceprint information (sound features) and the semantics of the call (the semantic content expressed by the call). In terms of the specific implementation method of the judgment, it is not limited to the voice information processing capabilities of the terminal itself, but coordinates the capabilities of terminal voice collection and processing, cloud-based intelligent voice processing, and cloud-based big data processing based on regional characteristics, thereby obtaining a more comprehensive, more accurate, and faster junk call judgment result; the system architecture uses end-cloud collaborative technology to process and identify junk calls. The terminal side is mainly responsible for the collection of incoming call audio signals, and the intelligent voice processing engine deployed on the cloud is responsible for processing the collected audio data and extracting the user-specific voiceprint information data and incoming call semantic text information. Thanks to the development of stronger and richer cloud computing power and intelligent voice processing technology, especially voice processing technology based on large model technology, the intelligent voice processing engine of the present invention can run more advanced and accurate voice processing algorithms and large model semantic analysis algorithms, which can quickly and efficiently remove noise interference from incoming call voice data, and accurately extract the caller's unique voiceprint information and incoming call semantic text information.

[0055] 2. The present invention is based on the system architecture of voice feature extraction and centralized maintenance of voiceprint data combined with end-cloud, and proposes an end-cloud collaborative establishment method for a whitelist caller voiceprint ID table (voiceprint information of mobile phone address book users). It is not necessary for each contact in the address book to call the owner of the machine. As long as a certain contact has called other owners, the owner of the machine can also obtain his voiceprint ID information, so that the voiceprint ID table of each contact in the owner's address book can be quickly established. Specifically, the voiceprint information of a single incoming call voice can be reused for several owner users, and then a relatively complete whitelist caller voiceprint ID table can be quickly collected for users who have newly installed the call management assistant, so that the voiceprint information obtained from a single incoming call voice collection can be used by several owners (the user to whom the voiceprint belongs is the contact in the address book of these owners), so that each contact in the owner's address book does not need to call the owner himself before obtaining the voiceprint information of each contact. The present invention can collect relatively complete voiceprint data of address book users. It will take a long time to accumulate the voiceprint information of the caller only by relying on the call management assistant installed by the owner during the call process. By centrally maintaining the voiceprint information of all incoming calls on the cloud (spam call prevention center) and then sharing it with relevant phone owners, as long as user A (including users who do not use the call management assistant) has voice communication with user B who uses the call management assistant, the spam call prevention center can collect the voiceprint information of the mobile phone user, so as to quickly obtain the voiceprint information of all incoming call users on the network, and thus quickly collect more complete voiceprint information of users in the address book. With more complete voiceprint information of users in the address book (the owner's whitelist voiceprint ID table), when a contact in the address book calls the owner's phone using a new number, the call management assistant will immediately remind the owner to answer the call, thereby avoiding being misjudged as a spam call and losing calls from acquaintances.

[0056] 3. The whitelist caller number collection and editing function provided by the spam call prevention center of the present invention can realize the whitelist caller number service capability of one-point acceptance and full network service through the centralized acceptance capability of the cloud. The spam call prevention center can provide whitelist number registration services for real-name authenticated personnel who provide public services such as express delivery, takeout, and taxi. Phone numbers in the whitelist number database will not be marked as spam calls, which can effectively avoid the phenomenon that staff providing public services frequently call different mobile phone users due to work needs and are mistakenly judged as spam calls. For example, many express delivery personnel and takeout personnel’s phones are often reminded as "advertising and marketing calls" in the called party’s mobile phone, further improving the accuracy of spam call identification of this technical solution.

[0057] 4. The call management assistant of the present invention can provide the owner with an anti-eavesdropping function. The call management assistant can accurately identify the data flow of the call application and can call the communication control capability of the underlying network of the mobile phone to temporarily disable all other application flows except the current call flow from entering and leaving the data network (including the Internet). Specifically, when the owner is on a call, the call management assistant calls the underlying network interface of the smartphone operating system to only allow the data flow of the call application to be sent to the communication network. The data flows of other applications in the mobile phone to the communication network are temporarily prohibited from passing, preventing malicious programs lurking in the mobile phone from recording the owner's call content and sending it to the external network. It can prevent malicious software from eavesdropping on the communication content and stealing user privacy information during the owner's call, thereby greatly improving the accuracy of junk call identification. While ensuring that the owner's communication is not eavesdropped, it can reduce the bombardment and harassment of various junk advertising information and junk calls caused by the leakage of personal information and eavesdropping on the call content from the source, more effectively reduce junk call harassment, and greatly improve the user experience of mobile phone users.

[0058] 5. The present invention avoids the deficiency of relying solely on the caller number to identify spam calls by processing and counting big data of multi-dimensional spam call information such as spam caller numbers, caller voiceprints, and the semantics of caller speech. The matching based on voiceprint data can effectively prevent spam callers who frequently change their caller numbers, and the matching based on speech semantics can effectively prevent spam software that uses voice changing software to make outbound calls, thereby achieving more intelligent judgment and more accurate interception of spam calls.

[0059] 6. The junk call prevention center of the present invention can provide a cloud backup function for key data for the call management assistant. The relevant data managed locally by the call management assistant include: the local address book, the local blacklist caller number table, the local whitelist caller voiceprint ID table, the local blacklist caller voiceprint ID table, the local preset blacklist caller semantic library, etc. When these local data are updated, they will be synchronized to the junk call prevention center, which can facilitate the owner to use the existing relevant junk call information after changing the phone. When the user switches to a new phone, the data of the old phone can be synchronized to the current phone, so as to obtain the same junk call prevention effect, further enhancing the user's experience of using smart phones. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 This is a flow chart of a method for preventing spam calls on a smart phone according to an embodiment of the present invention;

[0061] Figure 2 The present invention is a system structure diagram of a smart phone preventing junk calls in an embodiment of the present invention. DETAILED DESCRIPTION

[0062] 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 the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0063] Although the steps in the present invention are arranged with numbers, they are not used to limit the order of the steps. Unless the order of the steps is clearly stated or the execution of a certain step requires other steps as a basis, the relative order of the steps can be adjusted. It is understood that the term "and / or" used in this article involves and covers any and all possible combinations of one or more of the associated listed items.

[0064] The system of the present invention for preventing spam calls from smart phones has a structure as follows: Figure 2 As shown, it includes: a smart phone (hardware), a call management assistant (APP) running on the mobile phone, and a spam call prevention center (business software) running on the cloud.

[0065] Smartphones are used to access telephone networks and the Internet, and to receive and send communication signals;

[0066] The call management assistant is an APP software running in the mobile phone. It works with the spam call prevention center to identify whether unfamiliar calls are spam calls;

[0067] The spam call prevention center has two major capabilities: an intelligent voice processing engine and a big data processing module;

[0068] The intelligent voice processing engine can obtain the voiceprint information and speech semantic text (the speech content expressed by the caller) of the caller by analyzing the incoming call sound data, and can imitate the user's voice based on the user's voiceprint information;

[0069] The big data module is responsible for counting the junk call information reported by the call management assistant, ranking the junk calls based on the frequency of junk calls reported by each region, and regularly synchronizing the top junk call number list to the currently online call management assistant.

[0070] The system of the present invention is based on the technical architecture of end-cloud collaborative processing, fully exploring the advantages of high-performance voice data processing capabilities in the cloud. It can not only determine whether an incoming call is a spam call based on the caller number, but also determine whether an unfamiliar call is a spam call by extracting the caller's voiceprint information (sound features) and call semantics (semantic content expressed by the call). In addition, it can temporarily block the network traffic of other applications in the mobile phone during the phone owner's call to prevent the call content from being stolen, thereby greatly improving the accuracy of spam call identification and reducing various spam information caused by the leakage of owner information from the source, thereby more effectively reducing spam call harassment and greatly improving the user experience of mobile phone users.

[0071] The system architecture of the present invention uses end-cloud collaborative technology to process and identify spam calls. The terminal side is mainly responsible for collecting audio signals of incoming calls, and the intelligent voice processing engine deployed on the cloud is responsible for processing the collected audio data and extracting the user's unique voiceprint information data and the semantic text information of the incoming call. Thanks to the stronger and richer cloud computing power, the intelligent voice processing engine of the present invention can run more advanced and accurate voice processing algorithms and large model semantic analysis algorithms, which can quickly and efficiently remove noise interference from incoming call voice data and accurately extract the unique voiceprint information and semantic text information of the incoming caller.

[0072] A method for preventing spam calls from smart phones of the present invention comprises the following implementation steps:

[0073] Step 1. The smartphone is connected to the mobile network to make voice calls and access the Internet.

[0074] Step 2. The smartphone starts the Call Management Assistant APP. After the APP is started, data is initialized first. After the initialization is completed, it enters the call monitoring state.

[0075] In the embodiment, the data initialization work includes: reading the local address book of the mobile phone, reading the local call history record, and requesting the spam call prevention center to obtain the black and white list caller information table, which includes: white list caller number table, black list caller number table, local white list caller voiceprint ID table, local black list caller voiceprint ID table, and local preset black list caller semantic library (such as advertising promotion, fraud common keywords). In this article, the voiceprint ID represents the voiceprint identification code obtained by a certain caller.

[0076] in:

[0077] (1) The whitelist caller number table includes: the phone list of the local address book contacts, the phone numbers in the local call history with call duration exceeding N seconds (this N value can be set, the default is 60 seconds), the public service phone list such as 12315, 10000, 120, 110, and the phone number list of public service personnel (such as couriers) in the owner's area preset by the call prevention center.

[0078] (2) The blacklist caller number table includes: the blacklist phone number table set by the phone owner, and the TOP N (N value can be set, the default is TOP500) junk caller number table in the area where the phone owner is located.

[0079] (3) The local whitelist caller voiceprint ID table is a list of the voiceprint IDs of each contact in the local address book.

[0080] (4) The local blacklist caller voiceprint ID table is the voiceprint ID of the user corresponding to the blacklist phone number table set by the local device.

[0081] Among them, the local whitelist caller voiceprint ID table is obtained and established through the cloud collaborative collection method. The specific steps for obtaining the whitelist caller voiceprint ID table are: the call management assistant running on the mobile phone is responsible for the data collection of the incoming call audio signal and uploads it to the intelligent voice processing engine of the spam call prevention center; the intelligent voice processing engine uses voiceprint recognition technology to extract its characteristic data, namely voiceprint information (identified by voiceprint ID), from the audio data. This voiceprint information is shared by the entire network, that is, when the caller number corresponding to this voiceprint ID appears as a contact in N owner's address books, the contact voiceprint ID corresponding to this caller number in these N owner's address books is completed, so that the voiceprint ID table of the owner's address book contacts (that is, the owner's whitelist caller voiceprint ID table) can be quickly established.

[0082] The method for establishing a whitelist caller voiceprint ID table proposed in the present invention does not require every contact in the address book to call the local owner. As long as a certain contact has called other owners, the local owner can also obtain their voiceprint ID information, thereby quickly establishing a voiceprint ID table for each contact in the local owner's address book.

[0083] The audio data collected by the call management assistant comes from the following two sources:

[0084] (1) Mobile phone users use the Call Management Assistant to actively record the voice clips of the phone owner reading aloud and upload them to the intelligent voice processing engine of the spam call prevention center. This operation is usually prompted to the phone owner when the Call Management Assistant APP is first launched.

[0085] (2) When user A (corresponding to a certain number) using the call management assistant answers a call from user B, the call management assistant simultaneously records the voice signal of user B. The recorded audio data is then uploaded by the call management assistant to the intelligent voice processing engine of the spam call prevention center.

[0086] In the embodiment, when the call management assistant monitors an incoming call, it first compares the incoming call number with the whitelist caller number table and the blacklist caller number table, and enters the following related processing flow based on the comparison result.

[0087] Step 3. If the incoming call number monitored by the Call Management Assistant APP is in the whitelist phone number table, the smartphone is allowed to remind the owner to answer the call in the normal way (screen display when the phone rings, vibrates or is silent, etc.).

[0088] Step 4. If the incoming call number monitored by the call management assistant is in the blacklist, directly refuse to answer the call or hang up directly after answering.

[0089] Step 5. If the incoming call number monitored by the call management assistant is an unfamiliar number, the mobile phone will first imitate the owner's voice to say "Hello, how are you?" after answering the call silently, then keep answering the call silently and collect and record the caller's audio signal, and at the same time send the audio data stream (it is recommended to provide audio data of at least 10 seconds of voice clips) to the intelligent voice processing engine of the spam call prevention center, and then call the intelligent voice processing engine to obtain the voiceprint ID of the current call, the semantics and semantic keywords of the call, whether it is marked as a "spam call" and other information, and then compare the caller's voiceprint ID, caller semantic keywords with the whitelist caller voiceprint ID table, blacklist caller voiceprint ID table, the local preset blacklist caller semantic library and other information to determine whether the call is a spam call and take corresponding actions such as notifying the owner to answer the call or directly hanging up the call.

[0090] in:

[0091] An unfamiliar incoming call number refers to a phone number that is neither in the whitelist nor in the blacklist.

[0092] Imitating the owner's voice means that the call management assistant uses the TTS (text-to-speech function module) provided by the intelligent voice processing engine based on the owner's voiceprint data to imitate the owner's voice and say "Hello, how are you?"

[0093] The intelligent voice processing engine uses voiceprint recognition technology to extract the characteristic data of the incoming call, namely voiceprint information and the semantic text information of the incoming call, and uses the big data processing module to determine whether the incoming call is a spam call. The implementation steps are as follows:

[0094] (1) The call management assistant sends the collected and recorded call audio data, caller ID and phone number of the phone owner, and the owner's report information (this field is optional and is used for the owner's post-judgment after answering the call. If the owner believes that a certain answered call is a spam call, he or she can report the answered call as a spam call through the call management assistant) and other information to the intelligent voice processing engine.

[0095] (2) An intelligent voice processing engine that extracts the voiceprint information and semantic text of the incoming call from the audio data.

[0096] The voiceprint information includes the following data:

[0097] ◆The caller number that sent this voiceprint information.

[0098] ◆MFCC feature vector: A vector representing the energy distribution of a speech signal in the Mel-frequency domain.

[0099] ◆ Fundamental frequency spectrum: An image showing the fundamental frequency spectrum changing over time, where the horizontal axis represents time, the vertical axis represents frequency, and the depth of color represents the strength of energy.

[0100] ◆Speech envelope curve: A curve that describes the change of speech signal amplitude over time, usually used in conjunction with the fundamental frequency spectrogram.

[0101] ◆Formant frequency list: A series of numerical values ​​representing the frequency of occurrence of formants in the speech signal. These values ​​will change over time to form a trajectory.

[0102] The above voiceprint data features together constitute the caller's unique "voiceprint information". Based on these features, it can be determined whether a certain voice is spoken by a specific speaker.

[0103] (3) The intelligent voice processing engine sends the extracted voiceprint information data, semantic text and semantic keywords, and the owner's report information (whether it is a spam call) to the big data processing module, and asks the big data processing module to perform further processing and determine whether it is a spam call.

[0104] (4) After receiving the above data, the big data processing module performs the following processing steps:

[0105] 1) Compare and match this new voiceprint information with the network-wide voiceprint database maintained by the spam call management center. A successful match is indicated when the similarity with any four of the five types of voiceprint feature information described in 1)-5) above reaches 80%, otherwise it is unsuccessful. If there is no match with the data of all voiceprint IDs, there is no such voiceprint information in the network-wide voiceprint database (no matching voiceprint database). A new voiceprint ID is generated for this voiceprint information and this voiceprint data (including the caller number that generated this voiceprint) is added to the network-wide voiceprint database.

[0106] 2) If it matches the data of a certain voiceprint ID, check whether this voiceprint ID has been identified as a spam voiceprint. If it is a spam voiceprint, add the caller number to the spam caller number library (if the caller number already exists in the spam caller number library, add 1 to the statistical value of the spam caller number), then reply this voiceprint ID and the identification result of "spam voiceprint" to the intelligent voice processing engine, and end this operation.

[0107] 3) Match the semantic keywords of the incoming call with the semantic database of spam calls in the entire network. If the similarity is high (the semantic keywords of the incoming call are spam call keywords), the incoming call is judged as a spam call, and the voiceprint ID is marked as "spam call voiceprint" in the voiceprint database of the entire network. At the same time, the incoming call number is added to the database of spam call numbers in the entire network, and then the voiceprint ID and the "spam call" identification result are returned to the intelligent voice processing engine, ending this operation;

[0108] 4) If the owner reports the call as a spam call and it is the first time for the owner to report, then the "Number of spam call reports" field of this voiceprint ID in the whole network voiceprint database will be increased by 1, and the owner's phone number will be added to the "spam call reporter" field (to determine whether the reporter is reporting for the first time), and the text content of the semantics of this call will be saved in the "Semantic content of the call when reporting" field under this voiceprint ID entry. When the number of spam reports exceeds N digits (N value is configurable, default N = 3), this voiceprint ID will be marked as a "spam call voiceprint" in the whole network voiceprint database, and the caller number will be added to the whole network spam call number library, and the analysis of the semantic keywords of this spam call will be started, and the analyzed keywords will be added to the whole network spam call semantic library.

[0109] (5) The intelligent voice processing engine replies to the call management assistant with information such as “caller number, caller voiceprint ID, whether it is a spam call, and caller semantic text.”

[0110] The call management assistant compares the caller voiceprint ID, caller semantics and semantic keywords with the whitelist caller voiceprint ID table, blacklist caller voiceprint ID table, and the local preset blacklist caller semantic library information to determine whether the call is a spam call and take corresponding actions, including:

[0111] 5.1 If the incoming call voiceprint ID is in the whitelisted incoming call voiceprint ID table (in this scenario, a contact in the address book calls the owner using an unfamiliar phone number), the incoming call management assistant will issue a normal incoming call prompt tone or a user-defined incoming call prompt tone (such as "Please answer the call") to prompt the owner to answer the call.

[0112] 5.2 If the caller's voiceprint ID has been identified as a "spam call" by the intelligent voice processing engine of the spam call prevention center, the call management assistant will directly hang up the call.

[0113] 5.3 If the caller's voiceprint ID is in the blacklist caller's voiceprint ID table of this machine (this scenario is that the caller has called the owner before, and was set or reported as a blacklist caller by the owner, and then called the machine with another phone), the call management assistant will directly hang up the call.

[0114] 5.4 If the incoming caller's voiceprint ID is neither in the local whitelist incoming caller voiceprint ID table nor in the local blacklist incoming caller voiceprint ID table, but the incoming caller's semantic keywords have been included in the local preset blacklist incoming caller semantic library, the incoming call management assistant will directly hang up the call, and at the same time add the incoming caller number to the local blacklist incoming caller number table and the incoming caller's voiceprint ID to the local blacklist incoming caller voiceprint ID table.

[0115] 5.5 If the voiceprint ID of the incoming call does not belong to the above 5.1-5.4, the call management assistant will play a normal incoming call prompt tone or a user-defined incoming call prompt tone (such as "Please answer the call") to prompt the owner to answer the call.

[0116] Furthermore, after answering an unfamiliar call, the owner can set and report a certain historical call or multiple historical calls as junk calls in the call management assistant based on his / her subjective judgment. For the call records reported as junk calls, the call management assistant will upload the owner's number (reporter's number), the reported caller number, the reported call recording data, and the caller's voiceprint ID to the junk call prevention center, and at the same time add the caller number to the local blacklist caller number table and the caller's voiceprint ID to the local blacklist caller voiceprint ID table.

[0117] Furthermore, after receiving the report information from the call management assistant, the spam call prevention center will first send the relevant call recording to the intelligent voice processing engine for voiceprint recognition and semantic processing, and then based on the processing results of the intelligent voice processing engine, conduct big data statistics of spam calls in the area where the owner is located, such as updating the TOPN table of spam call numbers in the area, marking the reported caller voiceprint ID as spam calls, and updating the semantic library of spam calls in the entire network. The big data processing module regularly updates the ranking based on the frequency of spam call reports reported by each region (ranking information is updated every certain period of time, such as 30 minutes). In order to avoid misjudging normal unfamiliar calls as spam calls, reported calls with less frequency (such as less than 3 times) are not marked as "spam calls" for the time being. After the TOPN spam call number table is updated, it will be synchronously updated to the call management assistant in each smartphone.

[0118] In the embodiment, after the TOPN junk caller number list is updated, it will be synchronously updated to the call management assistant in each smart phone.

[0119] By processing and counting the big data of junk call information in multiple dimensions, such as the caller ID, the voiceprint of the caller, and the semantics of the caller's speech, we can avoid the deficiency of relying solely on the caller ID to identify junk calls. The matching based on voiceprint data can effectively prevent junk callers who frequently change their caller IDs, and the matching based on speech semantics can effectively prevent junk callers who use voice-changing software to make outbound calls, thereby achieving smarter judgment and more accurate interception of junk calls.

[0120] Furthermore, by providing whitelist number collection and editing functions, public service calls can be effectively prevented from being misjudged as spam calls.

[0121] The spam call prevention center can provide a whitelist number registration service for real-name authenticated personnel who provide public services such as express delivery, food delivery, and taxi hailing. Phone numbers in the whitelist number database will not be identified as spam calls. This further improves the accuracy of spam call identification of this technical solution and prevents public service calls from being mistakenly identified as spam calls.

[0122] Furthermore, the spam call prevention center can provide the call management assistant with a cloud backup function for key data. The relevant data managed locally by the call management assistant include: the local address book, the local blacklist caller number list, the local whitelist caller voiceprint ID list, the local blacklist caller voiceprint ID list, the local preset blacklist caller semantic library, etc. When these local data are updated, they will be synchronized to the spam call prevention center, so that the owner can still use the existing relevant spam call information after changing the phone.

[0123] When users switch to a new phone, they can synchronize the data of the old phone to the current phone to obtain the same spam call prevention effect, further enhancing the user's experience of using smartphones.

[0124] Furthermore, the call management assistant can provide the owner with an anti-eavesdropping function. When the owner is on a call, the call management assistant calls the underlying network interface of the smartphone operating system to only allow the data flow of the incoming call application to be sent to the communication network. The data flows of other applications in the mobile phone sent to the communication network are temporarily prohibited from passing. This prevents malicious programs lurking in the mobile phone from recording the owner's call content and sending it to the external network. While ensuring that the owner's communication is not eavesdropped, it can also reduce the bombardment and harassment of various junk messages and junk calls due to the leakage of personal information from the source.

[0125] During the call process, the owner only allows the current call application to use the data network, which can prevent malicious software in the phone from stealing the user's private data and call content, thereby reducing various spam call harassment caused by the owner's personal information being collected by advertising platforms.

[0126] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0127] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A system for preventing spam calls in a smart phone, characterized in that: Includes smartphones, call management assistants and spam call prevention centers; Among them, smartphones are used to access telephone networks and the Internet to receive and send communication signals; The call management assistant runs on the smartphone and works with the spam call prevention center to identify whether an unfamiliar call is a spam call; The spam call prevention center runs on the cloud, including an intelligent voice processing engine and a big data processing module; Intelligent voice processing engine, used to obtain the voiceprint information and semantic information of the caller through analysis of the incoming call voice data, and to imitate the voice of the phone owner to speak based on the user's voiceprint information; The big data processing module is used to count the junk call information reported by the call management assistant, rank and update the TOPN junk call number table based on the frequency of junk calls reported by each region, and regularly synchronize it to the currently online call management assistant; and determine whether the incoming call voice is a junk call.

2. A system for preventing spam calls in a smart phone according to claim 1, characterized in that: When the owner is on a call, the call management assistant calls the underlying network interface of the smartphone operating system to only allow the data flow of the incoming call application to be sent to the communication network. The data flows of other applications in the mobile phone sent to the communication network are temporarily prohibited from passing. This prevents malicious programs lurking in the mobile phone from recording the owner's call content and sending it to the external network. While ensuring that the owner's communication is not eavesdropped on, it also reduces the bombardment and harassment of various junk messages and junk calls due to the leakage of personal information from the source.

3. A system for preventing spam calls in a smart phone according to claim 1, characterized in that: The spam call prevention center provides a whitelist number registration service for real-name authenticated personnel who provide public services such as express delivery, food delivery, and taxi hailing. Phone numbers in the whitelist number database will not be identified as spam calls.

4. The system for preventing spam calls in a smart phone according to claim 1, characterized in that: The spam call prevention center also includes a terminal association and data backup module, which provides a key data cloud backup function for the call management assistant; The data locally managed by the call management assistant include: the local address book, the local whitelist caller number table, the local blacklist caller number table, the local whitelist caller voiceprint ID table, the local blacklist caller voiceprint ID table, and the local preset blacklist caller semantic library. When the locally managed data is updated, it is synchronously updated to the terminal association and data backup module of the spam call prevention center so that the owner can still use the existing relevant spam call information after changing the mobile phone.

5. A system for preventing spam calls in a smart phone according to claim 4, characterized in that: The smartphone allows the owner to answer unfamiliar calls and then, based on his or her subjective judgment, set and report historical calls as spam calls in the call management assistant. The call records reported as spam calls will be uploaded to the spam call prevention center by the call management assistant, including the owner's number, the reported caller number, the reported call recording data and the caller voiceprint ID information. At the same time, the corresponding caller number will be added to the local blacklist caller number table, and the caller voiceprint ID will be added to the local blacklist caller voiceprint ID table.

6. The system for preventing spam calls in a smart phone according to claim 1, characterized in that: After receiving the report information from the call management assistant, the spam call prevention center first sends the relevant call recording to the intelligent voice processing engine for voiceprint recognition and semantic processing, and then based on the processing results of the intelligent voice processing engine, conducts big data statistics of spam calls in the area where the owner is located, updates the TOPN table of spam call numbers in the area, identifies the reported call voiceprint ID as a spam call, and updates the semantic library of spam calls in the entire network; The big data processing module regularly ranks the frequency of junk call reports reported by each region, updates the TOPN junk call number table, and then synchronously updates it to the call management assistant in each smart phone.

7. A method for preventing spam calls on a smart phone implemented by the system according to any one of claims 1 to 6, characterized in that: The specific steps include: Step 1. Smartphone accesses the mobile network for voice calls and Internet access; Step 2. The smartphone starts the call management assistant. After starting, the call management assistant first performs data initialization. After the initialization is completed, it enters the call monitoring state. When the call management assistant monitors an incoming call, it compares the incoming call number with the whitelist caller number table and the blacklist caller number table for identification; Step 3. If the incoming call number monitored by the call management assistant is identified as being in the whitelist phone number table, the smartphone is allowed to remind the owner to answer the call in a normal incoming call reminder manner; Step 4. If the incoming call number monitored by the call management assistant is in the blacklist, the call is directly rejected or hung up after answering; Step 5. If the incoming call number monitored by the call management assistant is identified as an unfamiliar number, the mobile phone will answer the call silently and call the intelligent voice processing engine to first imitate the owner's voice to speak, then keep answering the call silently and collect and record the audio signal of the caller, and at the same time send the audio data stream to the intelligent voice processing engine of the spam call prevention center. By calling the intelligent voice processing engine, the current incoming call voiceprint ID, incoming call semantics and semantic keywords are obtained, and the big data processing module is used to obtain whether the call is marked as a "spam call". Then the call management assistant compares the incoming call voiceprint ID, incoming call semantics and semantic keywords with the whitelist incoming call voiceprint ID table, the blacklist incoming call voiceprint ID table, and the blacklist incoming call semantic library information preset by the local machine, and based on whether it is marked as a "spam call" information, it determines whether the call is a spam call and takes corresponding actions.

8. A method for preventing spam calls by a smart phone implemented by the system according to claim 7, characterized in that: The data initialization in step 2 specifically includes: reading the local address book and the local call history of the mobile phone, and requesting the spam call prevention center to obtain the blacklist and whitelist call information table, including: (1) A whitelist of incoming call numbers, including a list of phone numbers in the phone's address book, phone numbers in the phone's call history with a call duration of more than N seconds, a list of public service phone numbers, and a list of public service personnel phone numbers in the owner's area preset by the call prevention center; (2) Blacklist caller number list, including the blacklist phone number list set by the phone owner and the top N junk caller number list in the area where the phone owner is located; (3) The local whitelist caller voiceprint ID table is a list of voiceprint IDs of each contact in the local address book; the local whitelist caller voiceprint ID table is established by cloud collaborative collection, and the specific steps are as follows: The call management assistant collects the data of the incoming call voice audio signal and uploads it to the intelligent voice processing engine of the spam call prevention center; the intelligent voice processing engine uses voiceprint recognition technology to extract voiceprint information from the audio signal data and share it across the network. When the caller number corresponding to the voiceprint ID appears as a contact in the address books of N machine owners, the voiceprint ID of the contact corresponding to the caller number in the address books of these N machine owners is completed, thereby quickly establishing a voiceprint ID table of the contacts in the address book of the machine owner, that is, the whitelist caller voiceprint ID table of the machine owner; wherein, the call management assistant collects the data of the incoming call voice audio signal and uploads it to the intelligent voice processing engine of the spam call prevention center, including: when the call management assistant is started for the first time, it reminds the machine owner to use the call management assistant to actively record the voice clip read by the machine owner himself and upload it to the intelligent voice processing engine of the spam call prevention center; when the user A using the call management assistant answers the call of the incoming caller user B, the voice signal of the user B is recorded and uploaded to the intelligent voice processing engine of the spam call prevention center; (4) The local blacklist caller voiceprint ID table is the voiceprint ID of the corresponding user from the blacklist phone number table set by the local device; (5) The machine has a preset blacklist call semantic library, including common keywords for advertising, promotion, and fraud.

9. A method for preventing spam calls by a smart phone implemented by the system according to claim 7, characterized in that: In step 5, the intelligent voice processing engine uses voiceprint recognition technology to extract voiceprint information and caller semantics and semantic keywords from the audio signal data, and uses the big data processing module to determine whether the incoming call voice is a spam call, and obtains the caller voiceprint ID, caller semantics and semantic keywords, and whether it is marked as a "spam call". The specific steps are as follows: (1) The call management assistant sends the collected and recorded call audio data, caller ID, phone number of the owner, and the owner's reporting information to the intelligent voice processing engine; (2) The intelligent voice processing engine extracts the voiceprint information of the incoming call and the caller semantics and semantic keywords of the incoming call from the received audio data; the voiceprint information includes the following data: 1) The caller number that sent the voiceprint information; 2) MFCC feature vector: a vector representing the energy distribution of the speech signal in the Mel frequency domain; 3) Fundamental frequency spectrum: It is an image showing the change of fundamental frequency spectrum over time, where the horizontal axis represents time, the vertical axis represents frequency, and the depth of color represents the strength of energy; 4) Speech envelope curve: a curve that describes the change of speech signal amplitude over time, used in conjunction with the fundamental frequency spectrogram; 5) Formant frequency list: a series of values ​​representing the frequency of formants in the speech signal, which change over time to form a trajectory; (3) The intelligent voice processing engine sends the extracted voiceprint information, caller semantics and semantic keywords, and owner report information to the big data processing module; (4) After receiving the data sent by the intelligent voice processing engine, the big data processing module further determines whether it is a spam call, and specifically makes the following judgment: 1) Compare and match the voiceprint information with the whole network voiceprint database maintained by the spam call management center. If the similarity with any four of the five types of voiceprint feature information described in 1)-5) above reaches 80%, it means the match is successful, otherwise it means failure; if it does not match the data of all voiceprint IDs, then the whole network voiceprint database does not have this voiceprint information, and a new voiceprint ID is generated for this voiceprint information and this voiceprint data and its caller number are added to the whole network voiceprint database; if it matches the data of a certain voiceprint ID, check whether this voiceprint ID has been identified as a spam call voiceprint. If it has been identified as a spam call voiceprint, add this caller number to the spam caller number library. If this caller number already exists in the spam caller number library, add 1 to the statistical value of this spam caller number, and then reply this voiceprint ID and the identification result of "spam call voiceprint" to the intelligent voice processing engine to end this operation; 2) Match the semantic keywords of the incoming call with the semantic database of spam calls in the entire network. If the similarity exceeds the threshold, the incoming call is judged as a spam call, and the voiceprint ID is marked as "spam call voiceprint" in the voiceprint database of the entire network. At the same time, the incoming call number is added to the database of spam calls in the entire network, and the voiceprint ID and the "spam call" identification result are then returned to the intelligent voice processing engine to end this operation; 3) If the phone owner reports the call as a spam call and it is the first time for the phone owner to report, then the "Number of spam call reports" field of this voiceprint ID in the whole network voiceprint database is increased by 1, and the phone number of the phone owner is added to the "Junk call reporter" field, and the text content of the semantics of this call is saved in the "Semantic content of the call when reporting" field under this voiceprint ID entry; when the number of spam reports exceeds N digits, this voiceprint ID is marked as "spam call voiceprint" in the whole network voiceprint database, and the caller number is added to the whole network spam call number database, and the analysis of the semantic keywords of this spam call is started, and the analyzed keywords are added to the whole network spam call semantic database; (5) The intelligent voice processing engine replies to the call management assistant with the caller ID, caller voiceprint ID, caller semantics and semantic keywords, and whether the call is marked as "spam".

10. A method for preventing spam calls by a smart phone implemented by the system according to claim 7, characterized in that: In step 5, the call management assistant compares the caller voiceprint ID, caller semantics and semantic keywords with the whitelist caller voiceprint ID table, blacklist caller voiceprint ID table, and the local preset blacklist caller semantic library information to determine whether the call is a spam call and take corresponding actions, including: 5.1) If the caller's voiceprint ID is in the whitelist caller voiceprint ID table, the call management assistant will play a normal call prompt tone or a user-defined call prompt tone to prompt the owner to answer the call; 5.2) If the caller's voiceprint ID has been identified as a "spam call" by the intelligent voice processing engine of the spam call prevention center, the call management assistant will directly hang up the call; 5.3) If the incoming caller's voiceprint ID is in the blacklisted incoming caller's voiceprint ID table, the incoming call management assistant will directly hang up the call; 5.4) If the incoming caller's voiceprint ID is neither in the local whitelist incoming caller voiceprint ID table nor in the local blacklist incoming caller voiceprint ID table, but the incoming caller's semantic keywords have been included in the local preset blacklist incoming caller semantic library, the incoming call management assistant will directly hang up the call, and at the same time add the incoming caller number to the local blacklist incoming caller number table and the incoming caller's voiceprint ID to the local blacklist incoming caller voiceprint ID table; 5.5) If the voiceprint ID of the incoming call does not belong to the above 5.1)-5.4), the call management assistant will play a normal incoming call prompt tone or a user-defined incoming call prompt tone to prompt the owner to answer the call.

Images