A spam text message shielding method based on big data

Abstract

The invention relates to the technical field of SMS shielding, and discloses a method for shielding spam SMS based on big data, the method comprising the following steps: first, identifying a target user for whom spam SMS shielding is required, and obtaining the SMS received by the target user and the sender; using a pre-shielding strategy for preliminary screening, if not completely shielded, requesting access to the user's address book; adopting different SMS shielding strategies according to whether the user agrees to access the address book; if the user does not agree to access, partially or completely shielding the SMS by setting a threshold based on the number of times other users mark the spam SMS; if the user agrees to access, calculating the trust according to the address book information and the user interaction frequency, and processing the SMS accordingly; in addition, by setting a shielding word library and its pinyin variants to identify the SMS content, the pre-shielding strategy is further improved to achieve accurate interception.

Description

Technical Field

[0001] The present invention relates to the technical field of SMS shielding, and in particular to a method for shielding junk SMS based on big data. Background Art

[0002] With the rapid development of information technology, mobile phone text messages have become one of the main ways of daily communication for people; however, the problem of the proliferation of spam text messages is becoming increasingly serious, especially various fraudulent text messages, advertising text messages, etc., which have brought great troubles and security risks to users; spam text messages not only occupy the user's text message inbox and affect normal communication, but also may endanger the user's personal privacy and property safety through phishing links, false information, etc.; therefore, how to effectively block spam text messages has become an urgent problem to be solved in the field of mobile communications; at present, the existing spam text message blocking methods mostly rely on a single judgment standard, lacking flexibility and diversity; especially when the text message content is relatively complex and the sender information is difficult to determine, the blocking effect of the existing technology is still unsatisfactory; for example, some spam text messages may avoid the existing recognition standards by changing the pinyin, resulting in the inability of the blocking system to effectively identify and block these spam text messages; the existing technology mostly adopts a single blocking rule, and it is impossible to accurately block spam text messages; in addition, users can blacklist the numbers of certain senders, thereby automatically blocking text messages from these numbers; although this method is effective, it requires active maintenance by users and is difficult to deal with newly emerging spam text message senders.

[0003] In summary, there is an urgent need for a spam message blocking method based on big data to improve the shortcomings of spam message blocking technology in blocking accuracy. Summary of the invention

[0004] The present invention provides a spam text message shielding method based on big data, which helps solve the problems mentioned in the above background technology.

[0005] The present invention provides the following technical solution: a method for shielding spam messages based on big data, specifically comprising:

[0006] Obtain the users who need to block spam messages and record them as target users;

[0007] Get the SMS received by the target user and record it as the target SMS;

[0008] Get the sender of the target SMS, recorded as the sending user;

[0009] Determine whether to block all target SMS messages according to the pre-blocking strategy;

[0010] If it is determined according to the pre-screening strategy that the target SMS is not completely screened, a request to access the address book of the target user is initiated;

[0011] If the target user agrees to access the address book, the address book of the target user is obtained, and whether to block the target SMS is determined according to the first SMS blocking strategy;

[0012] If the target user does not agree to access the address book, whether to block the target SMS is determined according to the second SMS blocking strategy, including:

[0013] Set the first count threshold;

[0014] The first number threshold is used to determine whether to block all target text messages;

[0015] Obtain the number of times that the text messages sent by the sending user are marked as spam text messages by other users, and record it as the first marking number;

[0016] If the first marking number is greater than or equal to the first number threshold, all target text messages are blocked, and the sending user is marked as a harassing contact. After that, all text messages sent by the harassing contact are blocked.

[0017] If the first marking number is less than the first number threshold, then set the second number threshold;

[0018] The second number threshold is less than the first number threshold;

[0019] The second number threshold is used to determine whether to partially block the target short message.

[0020] Optionally, if the target user does not agree to access the address book, determining whether to block the target SMS according to the second SMS blocking strategy specifically includes:

[0021] If the first marking number is less than the second marking number threshold, the SMS sent by the sending user will not be partially blocked, and the entire content of the target SMS will be pushed to the target user;

[0022] If the second number threshold ≤ the first marking number < the first number threshold, the SMS sent by the sending user is partially blocked, specifically:

[0023] Get the target SMS content;

[0024] Determine whether the target SMS content contains network links and contact information;

[0025] If the target SMS contains a network link and contact information, the network link and contact information in the target SMS will be blocked, and the blocked SMS will be recorded as the SMS to be judged;

[0026] Push the SMS to be judged to the target user, and let the target user choose whether to block the entire content of the SMS to be judged;

[0027] If the target user chooses not to block all the contents of the SMS to be judged after viewing the SMS to be judged, the unblocked SMS corresponding to the SMS to be judged will be pushed to the target user;

[0028] If the target user chooses to block all the contents of the SMS to be judged after viewing it, all the SMS to be judged will be blocked, and the sending user will be marked as a harassing contact. Thereafter, all SMS sent by the harassing contact will be blocked.

[0029] Optionally, if the target user agrees to access the address book, obtaining the address book of the target user, and determining whether to block the target SMS according to the first SMS blocking strategy specifically includes:

[0030] If the target user's address book does not contain the sending user;

[0031] Then set the third number threshold;

[0032] The third digital threshold is used to determine whether to block all target text messages;

[0033] Obtain the number of times that the text messages sent by the sending user are marked as spam text messages by other users, and record it as the second marking number;

[0034] If the second marking number is greater than or equal to the third marking number threshold, all target text messages will be blocked, and the sending user will be marked as a harassing contact. All text messages sent by the harassing contact will be blocked thereafter.

[0035] If the second marking times is less than the third times threshold, then the fourth times threshold is set;

[0036] The fourth number threshold is less than the third number threshold;

[0037] The fourth number threshold is used to determine whether to partially block the target short message.

[0038] Optionally, if the target user agrees to access the address book, obtaining the address book of the target user, and determining whether to block the target SMS according to the first SMS blocking strategy specifically includes:

[0039] If the second marking number is less than the fourth number threshold, the SMS sent by the sending user is not partially blocked, and the entire content of the target SMS is pushed to the target user;

[0040] If the fourth number threshold ≤ the second marking number < the third number threshold, the text message sent by the sending user is partially blocked, specifically:

[0041] Get the target SMS content;

[0042] Determine whether the target SMS content contains network links and contact information;

[0043] If the target SMS contains a network link and contact information, the network link and contact information in the target SMS will be blocked, and the blocked SMS will be recorded as the SMS to be judged;

[0044] Push the SMS to be judged to the target user, and let the target user choose whether to block the entire content of the SMS to be judged;

[0045] If the target user chooses not to block all the contents of the SMS to be judged after viewing the SMS to be judged, the unblocked SMS corresponding to the SMS to be judged will be pushed to the target user;

[0046] If the target user chooses to block all the contents of the SMS to be judged after viewing it, all the SMS to be judged will be blocked, and the sending user will be marked as a harassing contact. Thereafter, all SMS sent by the harassing contact will be blocked.

[0047] Optionally, if the target user agrees to access the address book, obtaining the address book of the target user, and determining whether to block the target SMS according to the first SMS blocking strategy specifically includes:

[0048] If the target user's address book contains the sending user;

[0049] Then obtain the number of times the sending user sends SMS to the target user, which is recorded as the first number;

[0050] Get the number of times the target user replies to the text message sent by the user, recorded as the second number;

[0051] Calculate the second number ÷ the first number and record the result as the response ratio;

[0052] Setting a trust threshold, where the trust threshold is used to classify the sending user;

[0053] If the reply ratio is ≥ the trust threshold, the sending user will be listed as a first-level trusted person;

[0054] If the reply ratio is less than the trust threshold, the sending user will be listed as a second-level trust person;

[0055] In addition, the target user can freely categorize the contacts in the address book, and the target user has the highest level of categorizing the contacts in the address book;

[0056] For the SMS sent by the first-level trusted person to the target user, push the entire content of the SMS to the target user;

[0057] For text messages sent by second-level trusted persons to target users, all the contents of the text messages are blocked.

[0058] Optionally, judging whether to completely block the target short message according to the pre-blocking strategy specifically includes:

[0059] Setting a number of blocked words, wherein the blocked words are used to determine whether to block the target text message;

[0060] All blocked words constitute a blocked word database;

[0061] For a blocked word in the blocked word database, obtain the pinyin of the blocked word and record it as the original pinyin;

[0062] The original pinyin is processed as follows:

[0063] Keep the initial consonant and final of each Chinese character in the original pinyin unchanged;

[0064] For the tone of a Chinese character in the original pinyin, the tone is adjusted to be a light tone, a first tone, a second tone, a third tone and a fourth tone respectively;

[0065] By the above method, 5 to the power of N pinyins are formed, where N refers to the number of Chinese characters in the blocked word;

[0066] The 5 to the power of N pinyins formed in the above manner constitute a suspicious pinyin set;

[0067] Obtain the suspicious pinyin set of all blocked words in the blocked word database and form a determination set.

[0068] Optionally, judging whether to completely block the target short message according to the pre-blocking strategy further includes:

[0069] Get the target SMS content;

[0070] The text message content includes Chinese characters, numbers, letters, punctuation marks, spaces and emoticons;

[0071] If the target SMS content is all Chinese characters;

[0072] For the Chinese characters in the target text message content, the pinyins of the Chinese characters in the target text message content are obtained in sequence according to the order of the Chinese characters, and are recorded as the first pinyin, the second pinyin, ... the Xth pinyin, respectively, where X refers to the number of Chinese characters in the target text message content;

[0073] Arrange the first pinyin, the second pinyin, ... the Xth pinyin in the order of the first pinyin, the second pinyin, ... the Xth pinyin to form a pinyin string;

[0074] If the pinyins that appear continuously in the pinyin string are the same as a pinyin in the judgment set, the target SMS will be completely blocked;

[0075] If the pinyins that appear continuously in the pinyin string are different from each pinyin in the determination set, the target SMS will not be completely blocked.

[0076] Optionally, judging whether to completely block the target short message according to the pre-blocking strategy further includes:

[0077] If the target SMS content is not all Chinese characters;

[0078] Then all Chinese characters are extracted from the target text message content, and the pinyins of the Chinese characters in the target text message content are obtained in sequence according to the order in the target text message, and are recorded as the first target pinyin, the second target pinyin, ... the Yth target pinyin, where Y refers to the number of Chinese characters in the target text message content;

[0079] Arrange the first target pinyin, the second target pinyin, ... the Yth target pinyin in the order of the first target pinyin, the second target pinyin, ... the Yth target pinyin to form a pinyin string;

[0080] If the pinyins that appear continuously in the pinyin string are the same as a pinyin in the judgment set, the target SMS will be completely blocked;

[0081] If the pinyins that appear continuously in the pinyin string are different from each pinyin in the determination set, the target SMS will not be completely blocked.

[0082] The present invention has the following beneficial effects:

[0083] 1. This method of spam SMS shielding based on big data can identify and remove those contents that obviously meet the characteristics of spam SMS at the first time by adopting a pre-screening strategy to preliminarily screen the target SMS. The advantage of this strategy is that it can automatically identify and shield common types of spam SMS, such as advertising or malicious promotion SMS, saving users' time and energy. By setting shielding rules in advance, the pre-screening strategy can effectively reduce the misjudgment rate and ensure the accuracy of shielding spam SMS. Furthermore, this strategy can also make judgments quickly in real time according to predetermined rules, avoiding users from frequently manually screening and shielding SMS, thereby improving the user experience. It provides a higher response speed for the system and reduces the possibility of users being harassed by spam SMS. This method is particularly suitable for scenarios where users receive a large number of SMS or frequently receive spam SMS in a short period of time, and can significantly improve the daily use efficiency of users. In addition, the pre-screening strategy has strong scalability and can be adjusted in time according to the new characteristics of future spam SMS. With the improvement of big data analysis technology, the intelligence and flexibility of the strategy will continue to increase, further improving the screening ability of the system.

[0084] 2. For the spam message blocking method based on big data, in the pre-blocking strategy, by setting up a blocked word library and meticulously processing the pinyin in the blocked word library to generate a judgment set, the spam message blocking system can achieve a higher-precision filtering, especially when facing spam messages that avoid the conventional blocked word library through pinyin deformation. During the process of converting Chinese characters to pinyin, the system meticulously adjusts the tones of each pinyin to generate different pinyin variants. Compared with the original pinyin, although these pinyin variants may vary literally, they express the same or similar content. For example, the pinyin corresponding to the Chinese character "骚" is "sāo", but by adjusting the tone, the system can generate multiple pinyin deformations such as "sao", "sāo", "sǎo", etc., and these variants can all point to the same spam message content. In this way, although spam messages may attempt to avoid system detection through minor changes in pinyin, the system can identify these potential spam messages through this comprehensive pinyin deformation analysis.

[0085] 3. For the spam message blocking method based on big data, by judging whether the content of the target message is all Chinese characters, two different blocking strategies are adopted. If the content of the target message is all Chinese characters, then directly combine all the Chinese character pinyins in the message in sequence to form a pinyin string. Each pinyin string will be compared with the pinyins in the judgment set. If the consecutive pinyins in the pinyin string match a certain pinyin string in the judgment set, then this message will be determined as a spam message, thus triggering the blocking action. If the content of the target message is not all Chinese characters, then first extract the Chinese characters from the target message, and then combine all the Chinese character pinyins in sequence to form a pinyin string. Each pinyin string will be compared with the pinyins in the judgment set. If the consecutive pinyins in the pinyin string match a certain pinyin string in the judgment set, then this message will be determined as a spam message, thus triggering the blocking action. Through the above method, it can identify that spam message senders bypass the detection of ordinary spam message keywords through methods such as pinyin deformation, punctuation or spaces mixed in Chinese characters. For example, some spam messages may insert punctuation in Chinese characters, making it easier to bypass the identification of traditional keyword filtering systems.

[0086] 4. This is a spam SMS shielding method based on big data. It executes different shielding strategies by initiating a request to access the address book of the target user. This step enables the system to accurately filter spam SMS according to the address book information of the target user. By accessing the address book, the system can determine whether the sender is a contact familiar to the target user, which can avoid misjudging SMS that may be from a trusted source and improve the intelligence and accuracy of the system. If the user chooses not to allow access to the address book, the system can continue to execute according to other shielding strategies, thereby ensuring the flexibility of system operation. While ensuring privacy protection, the system can use the address book information to screen spam SMS more accurately, thereby reducing interference with important user information and improving user satisfaction with the service.

[0087] 5. This is a spam message blocking method based on big data. Setting the first number threshold can realize dynamic screening of spam messages, avoiding over-blocking or missed judgment caused by simple rule application. By setting the threshold, the system can more accurately identify senders who have been marked as spam messages for many times, so as to effectively block their messages. This feedback mechanism based on user behavior greatly improves the accuracy of spam message detection, and avoids users who frequently receive non-spam messages from being mistakenly blocked due to accidental mislabeled messages. In this way, the system can establish a database of spam message senders and mark them as "harassment contacts". Once the sender is determined to be a harassment contact, the system can fully automatically block all its messages, improving the efficiency of blocking spam messages.

[0088] 6. This is a spam message shielding method based on big data. It determines whether to partially shield the target message by setting a second number threshold. If the first marking number is less than the second number threshold, it means that the target message may not be a spam message, so it is pushed to the target user for viewing. However, in order to ensure the safety of the target user, the network link and contact information in the message will be shielded when it is pushed. After the user has viewed the message, the user decides whether to view the complete message or shield the message. By partially shielding and allowing users to participate in the selection, the system avoids excessive shielding of all message content that may be considered spam.

[0089] 7. A spam SMS shielding method based on big data. If the target user agrees to access the address book, then for the users in the address book, by obtaining the reply ratio of each contact, all contacts are classified into first-level contacts and second-level contacts. At the same time, the user can classify the contacts by himself, and the user's classification level is the highest. For the SMS sent to the target user by the first-level trust person, the entire content of the SMS is pushed to the target user; for the SMS sent to the target user by the second-level trust person, the entire content of the SMS is shielded; by introducing a trust level evaluation mechanism based on the interaction history between the user and the sender, the accuracy and personalization of spam SMS shielding are significantly improved. The system assigns a trust level to the sender based on data such as interaction frequency and reply rate, thereby dynamically adjusting the SMS shielding strategy. This method effectively reduces the occurrence of false shielding, ensures that important user information is not missed, and reduces the interference of harassing SMS. By giving users the autonomy to classify the address book, the flexibility of the system and the sense of user control are further improved, the overall user experience is enhanced, and the spam SMS screening algorithm is optimized through continuous interaction data feedback, achieving long-term and effective spam SMS protection. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0091] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0092] Example, see Figure 1 , a spam SMS shielding method based on big data, specifically comprising:

[0093] Obtain the users who need to block spam messages and record them as target users;

[0094] Get the SMS received by the target user and record it as the target SMS;

[0095] Get the sender of the target SMS, recorded as the sending user;

[0096] Determine whether to block all target SMS messages according to the pre-blocking strategy;

[0097] This is a spam SMS shielding method based on big data. By using a pre-screening strategy to preliminarily screen the target SMS, it can identify and remove those contents that clearly meet the characteristics of spam SMS in the first place. The advantage of this strategy is that it can automatically identify and shield common types of spam SMS, such as advertising or malicious promotion SMS, saving users' time and energy. By setting shielding rules in advance, the pre-screening strategy can effectively reduce the misjudgment rate and ensure the accuracy of shielding spam SMS. Furthermore, this strategy can also make quick judgments in real time according to predetermined rules, avoiding users from frequently manually screening and shielding SMS, thereby improving the user experience. It provides a higher response speed for the system and reduces the possibility of users being harassed by spam SMS. This method is particularly suitable for scenarios where users receive a large number of SMS or frequently receive spam SMS in a short period of time, and can significantly improve the daily use efficiency of users. In addition, the pre-screening strategy has strong scalability and can be adjusted in time according to the new characteristics of future spam SMS. With the improvement of big data analysis technology, the intelligence and flexibility of the strategy will continue to increase, further improving the screening ability of the system;

[0098] If it is determined according to the pre-screening strategy that the target SMS is not completely screened, a request to access the address book of the target user is initiated;

[0099] If the target user agrees to access the address book, the address book of the target user is obtained, and whether to block the target SMS is determined according to the first SMS blocking strategy;

[0100] The junk SMS shielding method based on big data executes different shielding strategies by initiating a request to access the address book of the target user. This step enables the system to accurately filter junk SMS according to the address book information of the target user. By accessing the address book, the system can determine whether the sender is a familiar contact of the target user, so as to avoid misjudging SMS that may be from a trusted source, thereby improving the intelligence and accuracy of the system. If the user chooses not to allow access to the address book, the system can continue to execute according to other shielding strategies, thereby ensuring the flexibility of system operation. While ensuring privacy protection, the system can use the address book information to screen junk SMS more accurately, thereby reducing interference with important user information and improving user satisfaction with the service.

[0101] If the target user does not agree to access the address book, whether to block the target SMS is determined according to the second SMS blocking strategy, including:

[0102] Set the first count threshold;

[0103] The first number threshold is used to determine whether to block all target text messages;

[0104] Obtain the number of times that the text messages sent by the sending user are marked as spam text messages by other users, and record it as the first marking number;

[0105] If the first marking number is greater than or equal to the first number threshold, all target text messages are blocked, and the sending user is marked as a harassing contact. After that, all text messages sent by the harassing contact are blocked.

[0106] If the first marking number is less than the first number threshold, then set the second number threshold;

[0107] The second number threshold is less than the first number threshold;

[0108] The second number threshold is used to determine whether to partially block the target text message;

[0109] The invention discloses a spam message shielding method based on big data. Setting the first number threshold can realize dynamic screening of spam messages, avoiding over-shielding or missed judgment caused by simple rule application. By setting the threshold, the system can more accurately identify senders who have been marked as spam messages for many times, so as to effectively shield their messages. The feedback mechanism based on user behavior greatly improves the accuracy of spam message detection, and avoids users who frequently receive non-spam messages from being erroneously intercepted due to accidentally mislabeled messages. In this way, the system can establish a database of spam message senders and mark them as "nuisance contacts". Once the sender is determined to be a nuisance contact, the system can fully automatically shield all its messages, thus improving the efficiency of blocking spam messages.

[0110] If the target user does not agree to access the address book, determining whether to block the target SMS according to the second SMS blocking strategy specifically includes:

[0111] If the first marking number is less than the second marking number threshold, the SMS sent by the sending user will not be partially blocked, and the entire content of the target SMS will be pushed to the target user;

[0112] If the second number threshold ≤ the first marking number < the first number threshold, the SMS sent by the sending user is partially blocked, specifically:

[0113] Get the target SMS content;

[0114] Determine whether the target SMS content contains network links and contact information;

[0115] If the target SMS contains a network link and contact information, the network link and contact information in the target SMS will be blocked, and the blocked SMS will be recorded as the SMS to be judged;

[0116] Push the SMS to be judged to the target user, and let the target user choose whether to block the entire content of the SMS to be judged;

[0117] If the target user chooses not to block all the contents of the SMS to be judged after viewing the SMS to be judged, the unblocked SMS corresponding to the SMS to be judged will be pushed to the target user;

[0118] If the target user chooses to block all the contents of the SMS to be judged after viewing it, all the SMS to be judged will be blocked, and the sending user will be marked as a harassing contact. Thereafter, all SMS sent by the harassing contact will be blocked.

[0119] If the target user agrees to access the address book, the address book of the target user is obtained, and according to the first message blocking strategy, it is determined whether to block the target message, which specifically includes:

[0120] If the target user's address book does not contain the sending user;

[0121] Then set the third number threshold;

[0122] The third digital threshold is used to determine whether to block all target text messages;

[0123] Obtain the number of times that the text messages sent by the sending user are marked as spam text messages by other users, and record it as the second marking number;

[0124] If the second marking number is greater than or equal to the third marking number threshold, all target text messages will be blocked, and the sending user will be marked as a harassing contact. All text messages sent by the harassing contact will be blocked thereafter.

[0125] If the second marking times is less than the third times threshold, then the fourth times threshold is set;

[0126] The fourth number threshold is less than the third number threshold;

[0127] The fourth number threshold is used to determine whether to partially block the target text message;

[0128] This is a spam message shielding method based on big data. It determines whether to partially shield the target message by setting a second number threshold. If the first marking number is less than the second number threshold, it means that the target message may not be a spam message, so it is pushed to the target user for viewing. However, in order to ensure the safety of the target user, the network link and contact information in the message will be shielded when it is pushed. After the user has viewed the message, it is up to the user to decide whether to view the complete message or shield the message. By partially shielding and allowing the user to participate in the selection, the system avoids excessive shielding of all message contents that may be considered spam.

[0129] If the target user agrees to access the address book, the address book of the target user is obtained, and according to the first message blocking strategy, it is determined whether to block the target message, which specifically includes:

[0130] If the second marking number is less than the fourth number threshold, the SMS sent by the sending user is not partially blocked, and the entire content of the target SMS is pushed to the target user;

[0131] If the fourth number threshold ≤ the second marking number < the third number threshold, the text message sent by the sending user is partially blocked, specifically:

[0132] Get the target SMS content;

[0133] Determine whether the target SMS content contains network links and contact information;

[0134] If the target SMS contains a network link and contact information, the network link and contact information in the target SMS will be blocked, and the blocked SMS will be recorded as the SMS to be judged;

[0135] Push the SMS to be judged to the target user, and let the target user choose whether to block the entire content of the SMS to be judged;

[0136] If the target user chooses not to block all the contents of the SMS to be judged after viewing the SMS to be judged, the unblocked SMS corresponding to the SMS to be judged will be pushed to the target user;

[0137] If the target user chooses to block all the contents of the SMS to be judged after viewing it, all the SMS to be judged will be blocked, and the sending user will be marked as a harassing contact. Thereafter, all SMS sent by the harassing contact will be blocked.

[0138] If the target user agrees to access the address book, the address book of the target user is obtained, and according to the first message blocking strategy, it is determined whether to block the target message, which specifically includes:

[0139] If the target user's address book contains the sending user;

[0140] Then obtain the number of times the sending user sends SMS to the target user, which is recorded as the first number;

[0141] Get the number of times the target user replies to the text message sent by the user, recorded as the second number;

[0142] Calculate the second number ÷ the first number and record the result as the response ratio;

[0143] Setting a trust threshold, where the trust threshold is used to classify the sending user;

[0144] If the reply ratio is ≥ the trust threshold, the sending user will be listed as a first-level trusted person;

[0145] If the reply ratio is less than the trust threshold, the sending user will be listed as a second-level trust person;

[0146] In addition, the target user can freely categorize the contacts in the address book, and the target user has the highest level of categorizing the contacts in the address book;

[0147] For the SMS sent by the first-level trusted person to the target user, push the entire content of the SMS to the target user;

[0148] For the text messages sent by the second-level trusted person to the target user, all the contents of the text messages are blocked;

[0149] This is a spam SMS shielding method based on big data. If the target user agrees to access the address book, then for the users in the address book, by obtaining the reply ratio of each contact, all contacts are classified into first-level contacts and second-level contacts. At the same time, the user can classify the contacts by himself, and the user's classification level is the highest. For the SMS sent to the target user by the first-level trust person, the entire content of the SMS is pushed to the target user; for the SMS sent to the target user by the second-level trust person, the entire content of the SMS is shielded; by introducing a trust level evaluation mechanism based on the interaction history between the user and the sender, the accuracy and personalization of spam SMS shielding are significantly improved. The system assigns trust levels to senders based on data such as interaction frequency and reply rate, thereby dynamically adjusting the SMS shielding strategy. This method effectively reduces the occurrence of misshielding, ensures that important user information is not missed, and reduces the interference of harassing SMS. By giving users the autonomy to classify the address book, the flexibility of the system and the sense of user control are further improved, the overall user experience is enhanced, and the spam SMS screening algorithm is optimized through continuous interaction data feedback, achieving long-term and effective spam SMS protection.

[0150] The determining whether to completely block the target SMS according to the pre-blocking strategy specifically includes:

[0151] Setting a number of blocked words, wherein the blocked words are used to determine whether to block the target text message;

[0152] All blocked words constitute a blocked word database;

[0153] For a blocked word in the blocked word database, obtain the pinyin of the blocked word and record it as the original pinyin;

[0154] The original pinyin is processed as follows:

[0155] Keep the initial consonant and final of each Chinese character in the original pinyin unchanged;

[0156] For the tone of a Chinese character in the original pinyin, the tone is adjusted to be a light tone, a first tone, a second tone, a third tone and a fourth tone respectively;

[0157] By the above method, 5 to the power of N pinyins are formed, where N refers to the number of Chinese characters in the blocked word;

[0158] The 5 to the power of N pinyins formed in the above manner constitute a suspicious pinyin set;

[0159] Obtain the suspicious pinyin set of all blocked words in the blocked word database and form a determination set.

[0160] The step of judging whether to completely block the target short message according to the pre-blocking strategy also includes:

[0161] Get the target SMS content;

[0162] The text message content includes Chinese characters, numbers, letters, punctuation marks, spaces and emoticons;

[0163] If the target SMS content is all Chinese characters;

[0164] For the Chinese characters in the target text message content, the pinyins of the Chinese characters in the target text message content are obtained in sequence according to the order of the Chinese characters, and are recorded as the first pinyin, the second pinyin, ... the Xth pinyin, respectively, where X refers to the number of Chinese characters in the target text message content;

[0165] Arrange the first pinyin, the second pinyin, ... the Xth pinyin in the order of the first pinyin, the second pinyin, ... the Xth pinyin to form a pinyin string;

[0166] If the pinyins that appear continuously in the pinyin string are the same as a pinyin in the judgment set, the target SMS will be completely blocked;

[0167] If the pinyins that appear continuously in the pinyin string are different from each pinyin in the determination set, the target SMS will not be completely blocked.

[0168] The step of judging whether to completely block the target short message according to the pre-blocking strategy also includes:

[0169] If the target SMS content is not all Chinese characters;

[0170] Then all Chinese characters are extracted from the target text message content, and the pinyins of the Chinese characters in the target text message content are obtained in sequence according to the order in the target text message, and are recorded as the first target pinyin, the second target pinyin, ... the Yth target pinyin, where Y refers to the number of Chinese characters in the target text message content;

[0171] Arrange the first target pinyin, the second target pinyin, …, the Y-th target pinyin in the order of the first target pinyin, the second target pinyin, …, the Y-th target pinyin to form a pinyin string;

[0172] If the pinyins that appear continuously in the pinyin string are the same as a pinyin in the determination set, then the target text message is completely blocked;

[0173] If the pinyins that appear continuously in the pinyin string are not the same as each pinyin in the determination set, then the target text message is not completely blocked;

[0174] In the pre-screening strategy of this method for blocking spam text messages based on big data, by setting up a blocked word library and carefully processing the pinyins in the blocked word library to generate a determination set, the spam text message blocking system can achieve a higher-precision filtering, especially when facing those spam text messages that avoid the conventional blocked word library through pinyin deformation; during the process of converting Chinese characters to pinyin, the system carefully adjusts the tones of each pinyin to generate different pinyin variants; compared with the original pinyin, although these pinyin variants may vary literally, the actual content expressed is the same or similar; for example, the pinyin corresponding to the Chinese character "骚" is "sāo", but by adjusting the tone, the system can generate multiple pinyin deformations such as "sao", "sāo", "sǎo", etc., and these variants can all point to the same spam text message content; in this way, although spam text messages may try to avoid system detection through minor changes in pinyin, the system can identify these potential spam text messages through this comprehensive pinyin deformation analysis;

[0175] This method for blocking spam text messages based on big data adopts two different blocking strategies by judging whether the content of the target text message is all Chinese characters. If the content of the target text message is all Chinese characters, then directly combine all the Chinese character pinyins in the text message in order to form a pinyin string; each pinyin string will be compared with the pinyins in the determination set; if the consecutive pinyins in the pinyin string match a certain pinyin string in the determination set, then this text message will be determined as a spam text message, thus triggering the blocking action; if the content of the target text message is not all Chinese characters, then first extract the Chinese characters from the target text message, and then combine all the Chinese character pinyins in order to form a pinyin string; each pinyin string will be compared with the pinyins in the determination set; if the consecutive pinyins in the pinyin string match a certain pinyin string in the determination set, then this text message will be determined as a spam text message, thus triggering the blocking action; through the above method, it is possible to identify that spam text message senders bypass the detection of ordinary spam text message keywords through pinyin deformation, punctuation or spaces mixed in Chinese characters, etc.; for example, some spam text messages may insert punctuation in Chinese characters, making it easier to bypass the recognition of traditional keyword filtering systems.

[0176] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0177] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A spam SMS shielding method based on big data, characterized in that: Specifically include: Obtain the users who need to block spam messages and record them as target users; Get the SMS received by the target user and record it as the target SMS; Get the sender of the target SMS, recorded as the sending user; Determine whether to block all target SMS messages according to the pre-blocking strategy; If it is determined according to the pre-screening strategy that the target SMS is not completely screened, a request to access the address book of the target user is initiated; If the target user agrees to access the address book, the address book of the target user is obtained, and whether to block the target SMS is determined according to the first SMS blocking strategy; If the target user does not agree to access the address book, whether to block the target SMS is determined according to the second SMS blocking strategy, including: Set the first count threshold; The first number threshold is used to determine whether to block all target text messages; Obtain the number of times that the text messages sent by the sending user are marked as spam text messages by other users, and record it as the first marking number; If the first marking number is greater than or equal to the first number threshold, all target text messages are blocked, and the sending user is marked as a harassing contact. After that, all text messages sent by the harassing contact are blocked. If the first marking number is less than the first number threshold, then set the second number threshold; The second number threshold is less than the first number threshold; The second number threshold is used to determine whether to partially block the target text message; Specifically: determine whether the target SMS content contains network links and contact information; if the target SMS content contains network links and contact information, block the network connection links and contact information in the target SMS content, and record the blocked SMS as the SMS to be judged; push the SMS to be judged to the target user, and the target user chooses whether to block the entire content of the SMS to be judged.

2. The method for blocking spam messages based on big data according to claim 1, characterized in that: If the target user does not agree to access the address book, determining whether to block the target SMS according to the second SMS blocking strategy specifically includes: If the first marking number is less than the second marking number threshold, the SMS sent by the sending user will not be partially blocked, and the entire content of the target SMS will be pushed to the target user; If the second number threshold ≤ the first marking number < the first number threshold, the SMS sent by the sending user is partially blocked, specifically: Get the target SMS content; Determine whether the target SMS content contains network links and contact information; If the target SMS contains a network link and contact information, the network link and contact information in the target SMS will be blocked, and the blocked SMS will be recorded as the SMS to be judged; Push the SMS to be judged to the target user, and let the target user choose whether to block the entire content of the SMS to be judged; If the target user chooses not to block all the contents of the SMS to be judged after viewing the SMS to be judged, the unblocked SMS corresponding to the SMS to be judged will be pushed to the target user; If the target user chooses to block all the contents of the SMS to be judged after viewing it, all the SMS to be judged will be blocked, and the sending user will be marked as a harassing contact. Thereafter, all SMS sent by the harassing contact will be blocked.

3. The method for blocking spam messages based on big data according to claim 1, characterized in that: If the target user agrees to access the address book, the address book of the target user is obtained, and according to the first message blocking strategy, it is determined whether to block the target message, which specifically includes: If the target user's address book does not contain the sending user; Then set the third number threshold; The third digital threshold is used to determine whether to block all target text messages; Obtain the number of times that the text messages sent by the sending user are marked as spam text messages by other users, and record it as the second marking number; If the second marking number is greater than or equal to the third marking number threshold, all target text messages will be blocked, and the sending user will be marked as a harassing contact. All text messages sent by the harassing contact will be blocked thereafter. If the second marking times is less than the third times threshold, then the fourth times threshold is set; The fourth number threshold is less than the third number threshold; The fourth number threshold is used to determine whether to partially block the target short message.

4. The method for blocking spam messages based on big data according to claim 3 is characterized in that: If the target user agrees to access the address book, the address book of the target user is obtained, and according to the first message blocking strategy, it is determined whether to block the target message, which specifically includes: If the second marking number is less than the fourth number threshold, the SMS sent by the sending user is not partially blocked, and the entire content of the target SMS is pushed to the target user; If the fourth number threshold ≤ the second marking number < the third number threshold, the text message sent by the sending user is partially blocked, specifically: Get the target SMS content; Determine whether the target SMS content contains network links and contact information; If the target SMS contains a network link and contact information, the network link and contact information in the target SMS will be blocked, and the blocked SMS will be recorded as the SMS to be judged; Push the SMS to be judged to the target user, and let the target user choose whether to block the entire content of the SMS to be judged; If the target user chooses not to block all the contents of the SMS to be judged after viewing the SMS to be judged, the unblocked SMS corresponding to the SMS to be judged will be pushed to the target user; If the target user chooses to block all the contents of the SMS to be judged after viewing it, all the SMS to be judged will be blocked, and the sending user will be marked as a harassing contact. Thereafter, all SMS sent by the harassing contact will be blocked.

5. The method for blocking spam messages based on big data according to claim 4 is characterized in that: If the target user agrees to access the address book, the address book of the target user is obtained, and according to the first message blocking strategy, it is determined whether to block the target message, which specifically includes: If the target user's address book contains the sending user; Then obtain the number of times the sending user sends SMS to the target user, which is recorded as the first number; Get the number of times the target user replies to the text message sent by the user, recorded as the second number; Calculate the second number ÷ the first number and record the result as the response ratio; Setting a trust threshold, where the trust threshold is used to classify the sending user; If the reply ratio is ≥ the trust threshold, the sending user will be listed as a first-level trusted person; If the reply ratio is less than the trust threshold, the sending user will be listed as a second-level trust person; In addition, the target user can freely categorize the contacts in the address book, and the target user has the highest level of categorizing the contacts in the address book; For the SMS sent by the first-level trusted person to the target user, push the entire content of the SMS to the target user; For text messages sent by second-level trusted persons to target users, all the contents of the text messages are blocked.

6. The method for blocking spam messages based on big data according to claim 1, characterized in that: The determining whether to completely block the target SMS according to the pre-blocking strategy specifically includes: Setting a number of blocked words, wherein the blocked words are used to determine whether to block the target text message; All blocked words constitute a blocked word database; For a blocked word in the blocked word database, obtain the pinyin of the blocked word and record it as the original pinyin; The original pinyin is processed as follows: Keep the initial consonant and final of each Chinese character in the original pinyin unchanged; For the tone of a Chinese character in the original pinyin, the tone is adjusted to be a light tone, a first tone, a second tone, a third tone and a fourth tone respectively; By the above method, 5 to the power of N pinyins are formed, where N refers to the number of Chinese characters in the blocked word; The 5 to the power of N pinyins formed in the above manner constitute a suspicious pinyin set; Obtain the suspicious pinyin set of all blocked words in the blocked word database and form a determination set.

7. The method for blocking spam messages based on big data according to claim 6 is characterized by: The step of judging whether to completely block the target short message according to the pre-blocking strategy also includes: Get the target SMS content; The text message content includes Chinese characters, numbers, letters, punctuation marks, spaces and emoticons; If the target SMS content is all Chinese characters; For the Chinese characters in the target text message content, the pinyins of the Chinese characters in the target text message content are obtained in sequence according to the order of the Chinese characters, and are recorded as the first pinyin, the second pinyin, ... the Xth pinyin, respectively, where X refers to the number of Chinese characters in the target text message content; Arrange the first pinyin, the second pinyin, ... the Xth pinyin in the order of the first pinyin, the second pinyin, ... the Xth pinyin to form a pinyin string; If the pinyins that appear continuously in the pinyin string are the same as a pinyin in the judgment set, the target SMS will be completely blocked; If the pinyins that appear continuously in the pinyin string are different from each pinyin in the determination set, the target SMS will not be completely blocked.

8. The method for shielding spam messages based on big data according to claim 7, characterized in that: The step of judging whether to completely block the target short message according to the pre-blocking strategy also includes: If the target SMS content is not all Chinese characters; Then all Chinese characters are extracted from the target text message content, and the pinyins of the Chinese characters in the target text message content are obtained in sequence according to the order in the target text message, and are recorded as the first target pinyin, the second target pinyin, ... the Yth target pinyin, where Y refers to the number of Chinese characters in the target text message content; Arrange the first target pinyin, the second target pinyin, ... the Yth target pinyin in the order of the first target pinyin, the second target pinyin, ... the Yth target pinyin to form a pinyin string; If the pinyins that appear continuously in the pinyin string are the same as a pinyin in the judgment set, the target SMS will be completely blocked; If the pinyins that appear continuously in the pinyin string are different from each pinyin in the determination set, the target SMS will not be completely blocked.

Images