Internet pushing system based on big data

Abstract

The invention discloses an internet pushing system based on big data, and relates to the technical field of big data analysis, the internet pushing system comprises a management center, the management center is connected with a login acquisition module, a data processing module, a data analysis module and an encryption recommendation module; user registration and login are carried out in the login acquisition module, and behavior data in the Internet service process are acquired; the data processing module processes the collected behavior data and constructs a user access table; analyzing the constructed user access table in a data analysis module to obtain an access keyword; performing character conversion and blocking on the access keyword through an encryption recommendation module to obtain a word sequence segment, performing cardinal number encryption on the obtained word sequence segment to obtain a ciphertext recommendation sequence, and performing decryption, optimization and screening on the obtained ciphertext recommendation sequence to obtain an adjustment recommendation sequence; recommendation accuracy is improved for the user, user privacy protection in the recommendation process is enhanced, and information browsed by the user is prevented from being leaked.

Description

Technical Field

[0001] The present invention relates to the technical field of big data analysis, and specifically to an Internet push system based on big data. Background Art

[0002] Internet push is a system that pushes personalized content to users through the Internet. It analyzes users' behaviors, interests, and preferences, and according to users' needs and habits, it real-time pushes suitable information and content to users; through the Internet push system, users can obtain the content they are interested in, improve the usage experience, and increase user stickiness and activity.

[0003] However, existing push systems often only provide fixed push content, lack personalized customization and diverse content selection, and cannot meet the diverse needs of users; at the same time, due to the lack of in-depth analysis and mining of users' behaviors and preferences in the push system, the push content does not match users' needs, resulting in poor push effects. And existing push systems often lack reasonable planning and control of users' time, resulting in too high or too low push frequencies, affecting users' usage experience; finally, the Internet push system needs to process a large amount of user data, and data security and privacy protection issues cannot be ignored, and effective data encryption and privacy protection measures need to be taken; therefore, optimizing the deficiencies of Internet push based on big data has important theoretical and practical significance.

[0004] How to use big data analysis technology to collect the behavior data of registered and logged-in users on the Internet, match and classify the behavior data and construct a user access table, obtain keyword sorting by analyzing the user access table, perform radix encryption on the keyword sorting and then recommend it to users, and provide it to users for selection after authorized decryption is the problem we need to solve; for this reason, an Internet push system based on big data is provided now. Summary of the Invention

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] An Internet push system based on big data, including a management center, the management center is connected with a login collection module, a data processing module, a data analysis module, and an encryption recommendation module;

[0007] The login collection module is used to register and log in users, obtain networked users, and collect the behavior data of networked users during the Internet service process. The specific process includes:

[0008] Set up a user login port and an identification record end;

[0009] Enter user information through the user login port and set a login password, upload the obtained user information and login password to the management center, and the management center will review the user information and generate a login account for the user who passes the review according to the user information;

[0010] Enter the login account and password to log in, and the user who has completed the login will be marked as an online user;

[0011] The comprehensive data of Internet users during Internet services is collected through the identification and recording terminal, and the collected comprehensive data is identified and recorded to obtain behavior data, which includes access addresses, click behaviors, and dwell time.

[0012] The process of processing the collected behavioral data and obtaining the user access form includes:

[0013] Set the statistical period, count the access addresses within the statistical period, and obtain the number of address categories;

[0014] Classify click behaviors based on statistical cycles to obtain behavior types;

[0015] Matching the obtained access address with the behavior type to obtain the address matching behavior, and associating the obtained address matching behavior with the residence time;

[0016] A user access table is constructed based on the obtained access addresses, address category numbers, address matching behaviors, and dwell time.

[0017] The process of the data analysis module analyzing the obtained user access form includes:

[0018] Obtaining the address behavior degree according to the address matching behavior in the obtained user access table;

[0019] Obtain visit preference based on the obtained address behavior, dwell time, statistical period, and number of address categories;

[0020] Sort the access addresses according to the obtained access popularity to obtain push ranking;

[0021] Set a key extraction end, extract keywords from the access address through the key extraction end according to the push ranking obtained, and obtain the access keyword;

[0022] The obtained access addresses are associated with access keywords, and the access keywords are repeated to obtain the corresponding number of addresses, the obtained corresponding number of addresses are sorted, the keyword sorting is obtained, and the obtained keyword sorting and push sorting are associated with networked users.

[0023] Perform character conversion on the accessed keywords in the sorted keywords obtained, to obtain the original keyword sequence, and perform character conversion on the accessed addresses obtained, to obtain the original address sequence;

[0024] Set the base number and the initial vector according to the obtained original address sequence, and obtain the initial address vector according to the obtained base number and initial vector;

[0025] Insert the obtained initial address vector into the original keyword sequence, to obtain the original key sequence, perform block division on the obtained original key sequence, to obtain the word sequence segments, and mark the initial address vectors in the word sequence segments as decoding segments.

[0026] The process of performing radix encryption on the obtained word sequence segments includes:

[0027] Set the seed radix according to the obtained word sequence segments, and obtain the radix product according to the obtained seed radix;

[0028] Obtain the characteristic function according to the obtained seed radix and radix product;

[0029] Set the convention radix, and obtain the private radix according to the obtained characteristic function and convention radix;

[0030] Obtain the ciphertext word segment according to the obtained convention radix, radix product and word sequence segments;

[0031] Mark the obtained radix product and private radix as the locked key group, associate the obtained locked key group with the login password of the networked user, and upload the obtained locked key group to the management center;

[0032] Combine the ciphertext word segment according to the obtained original key sequence, to obtain the ciphertext key group, and mark the decoding segments in the ciphertext key group as characteristic decoding segments;

[0033] Sort the obtained ciphertext key group according to the keyword sorting, to obtain the ciphertext recommendation sequence.

[0034] Upload the obtained ciphertext recommendation sequence to the networked user, and send a plaintext response to the management center, and send a password matching instruction to the networked user through the management center;

[0035] The networked user inputs the login password to match with the login password associated with the locked key group, and grant the key acquisition permission to the networked user who matches successfully;

[0036] The networked user obtains the locked key group associated with the login password according to the obtained key acquisition permission;

[0037] Obtain the characteristic plaintext segment according to the obtained locked key group and characteristic decoding segment;

[0038] Sort the feature description paragraphs according to the obtained keyword sorting to obtain the preselected address sorting.

[0039] Set up a push window and upload the obtained preselected address sorting to the push window;

[0040] Internet users select the preselected recommended addresses through the push window, mark the selected feature description paragraphs as the selected access addresses, and obtain the favorite behavior degrees of the selected access addresses according to the obtained address behavior degrees;

[0041] Set a favorite degree threshold, and mark the access addresses corresponding to the feature description paragraphs with a favorite behavior degree less than the favorite degree threshold as insensitive access addresses;

[0042] Optimize and sort the obtained insensitive access addresses to obtain the adjusted recommended sorting.

[0043] Compared with the prior art, the beneficial effects of the present invention are: process the collected behavior data to obtain a user access form, obtain the access favorite degree, access keywords and keyword sorting according to the user access form, and perform character conversion on the access keywords and access addresses in the keyword sorting to obtain word sequence segments and decoding segments;

[0044] Perform radix encryption on the obtained word sequence segments and decoding segments to obtain ciphertext key segments, feature decoding segments and locked key groups, aiming to encrypt and protect the obtained user preference types, prevent third-party systems or personnel from directly obtaining the user preference types, ensure that user privacy will not be leaked, and increase confidentiality according to the obtained locked key groups, so that only the user himself has the permission to view the user preference types, improving the security of the Internet recommendation system;

[0045] Decrypt and sort the ciphertext key segments by the obtained locked key groups to obtain the preselected address sorting, provide the obtained preselected address sorting to the user for selection to obtain the selected access addresses, and optimize and screen the selected access addresses to obtain the adjusted recommended sorting; the user preference types can be updated in real time during the push, increasing the push accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0047] Figure 1 It is the schematic diagram of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0048] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative work belong to the scope of protection of the present invention.

[0049] As Figure 1 shown, an Internet push system based on big data includes a management center, and the management center is connected with a login collection module, a data processing module, a data analysis module and an encryption recommendation module;

[0050] The login collection module is used to register and log in users, obtain networked users, and collect the behavior data of networked users during the Internet service process. The specific process includes:

[0051] Set up a user login port and an identification record end;

[0052] The user login port is used to register and log in accounts for Internet users;

[0053] Input user information through the user login port and set a login password;

[0054] Upload the obtained user information and login password to the management center, and the management center audits the user information and generates a login account for the successfully audited users according to the user information;

[0055] It should be further noted that in the specific implementation process, the user information includes name, mobile phone number, and ID number, and the login account is generated according to the mobile phone number;

[0056] Log in by inputting the login account and login password at the user login port, and mark the logged-in users as networked users;

[0057] Collect the comprehensive data of networked users during the Internet service process through the identification record end;

[0058] Identify and record the collected comprehensive data to obtain behavior data, and the behavior data includes access addresses, click behaviors, and stay times;

[0059] It should be further noted that in the specific implementation process, the access address is a record of a specific page opened or accessed by a user during an Internet service. Whenever an Internet-connected user clicks on a link or button to access a specific page of a website or application, a record of the access address is generated; the click behavior represents a record of an Internet-connected user clicking on a certain link, button, picture, or other clickable element in a website or application. Whenever an Internet-connected user clicks on any interactive element, a record of the click behavior is generated.

[0060] The data processing module is used to process the obtained behavior information to obtain a user access form. The specific process includes:

[0061] Set a statistical period and mark the duration of the set statistical period as T i , where i = 1, 2, 3,..., v1, and v1 is a positive integer;

[0062] Statistically count the access addresses within the statistical period to obtain the number of address categories;

[0063] Mark the obtained access address as DJ m , where m represents the number of the access address;

[0064] Mark the obtained number of address categories as DJ m-k , where k represents the number of repetitions of the same access address, and k = 1, 2, 3,..., v2, and v2 is a positive integer. Then m - k represents that the m-th access address has been accessed k times within the statistical period. For example, DJ 2-5 represents that the 2nd access address has been accessed 5 times within the statistical period;

[0065] Classify the click behavior based on the statistical period to obtain the behavior types, and the behavior types include browsing behavior, liking behavior, commenting behavior, and sharing behavior;

[0066] Mark the browsing behavior as LL m-k Mark the liking behavior as DZ m-k Mark the commenting behavior as PL m-k Mark the sharing behavior as FX m-k ;

[0067] Match the obtained access address with the behavior types to obtain the address-matched behavior;

[0068] Associate the obtained address-matched behavior with the residence time and mark the obtained residence time as T m-k ;

[0069] Construct a user access form based on the obtained access address, number of address categories, address-matched behavior, and residence time, as shown in Table 1:

[0070]

[0071] Table 1

[0072] It should be further noted that in the specific implementation process, the address matching behavior is represented by the number 0 or 1 in the user access table. 0 indicates the absence of this behavior, and 1 indicates the presence of this behavior. For example, in DJ 1-2 if there is only a browsing behavior and no other behaviors, then in the user access table, the corresponding table for the browsing behavior is filled with 1, and the corresponding tables for the like behavior, comment behavior, and share behavior are filled with 0.

[0073] The data analysis module is used to analyze the obtained user access table to obtain an analysis result. The specific process includes:

[0074] Obtain the address behavior degree according to the address matching behavior in the obtained user access table, and mark the obtained address behavior degree as XW m-k , where XW m-k =α1*LL m-k +α2*DZ m-k +α3*PL m-k +α4*FX m-k , α1, α2, α3, α4 are proportionality factors, and α1 + α2 + α3 + α4 = 1;

[0075] It should be further noted that in the specific implementation process, the browsing behavior, like behavior, comment behavior, and share behavior are marked with levels, and the address behavior degree can be obtained only by calculating according to the level marks. That is, the browsing level is marked as 1, the like level is marked as 2, the comment level is marked as 3, and the share level is marked as 4. Then XW m-k =α1*1 + α2*2 + α3*3 + α4*4. If it is 0 at the corresponding position in the user access table, the proportionality factor of the corresponding behavior is 0, and the non-zero proportionality factors remain unchanged. For example, in the user behavior degree table, when accessing DJ 1-2 , there is only a browsing behavior and no other behaviors, then LL 1-2 =1, α2 = α3 = α4 = 0, that is, XW m-k =α1*1;

[0076] Obtain the access preference according to the obtained address behavior degree, residence time, duration of the statistical period, and number of address categories, and mark the obtained access preference as XA m , where β1 and β2 represent influence factors, and β1 + β2 = 1;

[0077] Sort the access addresses in descending order according to the obtained access popularity to obtain a push sort order;

[0078] Set a key extraction end. According to the obtained push sort order, extract keywords from the access addresses through the key extraction end to obtain access keywords;

[0079] It should be further noted that in the specific implementation process, the key extraction end can be a website page analysis tool, a search engine optimization tool, and a social media analysis tool. Extract the content corresponding to the access address through the keyword extraction end to obtain keywords;

[0080] Associate the obtained access addresses with the access keywords, and count the repetition degree of the access keywords to obtain the number corresponding to the address. Sort the obtained number corresponding to the address in descending order to obtain a keyword sort order;

[0081] And sort the different access addresses corresponding to the same access keyword in the keyword sort order according to the obtained push sort order to obtain a keyword-address sub-order;

[0082] It should be further noted that in the specific implementation process, the repetition degree statistics is to count the repetition times of the same keyword, and the sorting of different access addresses corresponding to the same keyword in the keyword sort order is the same as the push sort order;

[0083] Associate the obtained keyword sort order and push sort order with the online users, and upload the obtained keyword sort order to an encrypted recommendation module. The encrypted recommendation module is used to perform encrypted push for the online users according to the obtained keyword sort order. The specific process includes:

[0084] Perform character conversion on the access keywords in the obtained keyword sort order to obtain a keyword original sequence, which is composed of several randomly arranged numbers;

[0085] Perform character conversion on the obtained access addresses to obtain an address original sequence;

[0086] Set a base number and an initial vector according to the obtained address original sequence;

[0087] It should be further noted that in the specific implementation process, the base number is a randomly generated integer, the initial vector is a randomly generated 128-bit integer, and the initial vector must be randomly generated, and there can be only one initial vector for the encryption of the same set of keyword sort orders;

[0088] An initial address vector is obtained based on the acquired base number and the initial vector. The initial address vector is a random 128 - bit number generated by taking the base number and the initial vector as inputs through a key generation algorithm, that is, some mathematical functions and a random number generator. This is the initial address vector. That is, according to two given random numbers, another random number is generated, and it is necessary to ensure the randomness of the generated initial address vector.

[0089] Insert the obtained initial address vector at the end of the original keyword sequence to obtain the original key sequence. The obtained original key sequence is segmented to obtain word sequence segments. The obtained word sequence segments are marked as e, and the initial address vectors in the word sequence segments are marked as decoding segments.

[0090] It should be further noted that in the specific implementation process, since one access keyword can correspond to multiple access addresses, but only one initial address vector can be inserted at the end of one access keyword. If the same keyword corresponds to multiple access addresses, multiple original key sequences with the same original keyword sequence are generated.

[0091] In particular, the length of the obtained word sequence segments must be an integer multiple of the length of the initial vector.

[0092] The process of performing radix encryption on the obtained word sequence segments includes:

[0093] Set the seed radix according to the obtained word sequence segments. The seed radix includes a first seed number and a second seed number. Mark the obtained first seed number as p and the obtained second seed number as q, and both p and q are very large prime numbers.

[0094] Obtain the radix product according to the obtained seed radix. Mark the obtained radix product as n, where n = q * p.

[0095] Obtain the characteristic function according to the obtained seed radix and radix product. Mark the obtained characteristic function as ψ(n), where ψ(n)=(q - 1)*(p - 1), and ψ(n) is the Euler's totient function.

[0096] Set the common radix, and mark the obtained common radix as f, where 1 < f < ψ(n), and f and ψ(n) are relatively prime.

[0097] Obtain the private radix according to the obtained characteristic function and common radix. Mark the obtained private radix as d, where d * f = 1 (mod ψ(n)), that is, d is the modular multiplicative inverse of f with respect to ψ(n).

[0098] Encrypt the word sequence segments according to the obtained common radix and radix product to obtain the ciphertext word segments. Mark the obtained ciphertext word segments as a, where a = e f(mod(n)), where mod represents taking the modulus;

[0099] Mark the obtained base product and private base as the locked key group, associate the obtained locked key group with the login password of the networked user, and upload the obtained locked key group to the management center;

[0100] Combine the obtained ciphertext word segments according to the obtained original key sequence to obtain a ciphertext key group, mark the decoded segments in the ciphertext key group as feature decoded segments, and mark the obtained feature decoded segments as g;

[0101] Sort the obtained ciphertext key group according to the keyword sorting to obtain a ciphertext recommendation sequence;

[0102] Upload the obtained ciphertext recommendation sequence to the networked user, send a plaintext response to the management center, and send a password matching instruction to the networked user through the management center;

[0103] The networked user inputs the login password to match with the login password associated with the locked key group, and grants the key acquisition permission to the networked user with successful matching;

[0104] The networked user obtains the locked key group associated with the login password according to the obtained key acquisition permission;

[0105] Decrypt the feature decoded segments in the ciphertext key group according to the obtained locked key group to obtain feature plaintext segments, and mark the obtained feature plaintext segments as c, where c = g d (mod(n));

[0106] Sort the feature plaintext segments according to the obtained keyword sorting and keyword address sub - order to obtain a pre - selected address sorting;

[0107] Set a push window, upload the obtained pre - selected address sorting to the push window. Further, the order uploaded to the push window is the same as the order of the keyword address sub - order;

[0108] The networked user makes a push selection through the push window, marks the selected feature plaintext segments as the selected access addresses, obtains the address matching behavior of the networked user within the selected access addresses, obtains the address behavior degree according to the obtained address matching behavior, and marks the obtained address behavior degree as the favorite behavior degree;

[0109] Set a favorite degree threshold, mark the obtained favorite degree threshold as W0, and mark the access addresses corresponding to the feature plaintext segments with the favorite behavior degree less than the favorite degree threshold as non - sensitive access addresses;

[0110] Lower the sorting of the obtained non - sensitive access addresses to the last position of the keyword address sub - order to obtain an adjusted recommendation sorting;

[0111] Continue to recommend to networked users according to the obtained adjusted recommended ranking, and repeat the process of obtaining the feature clear text segment to adjust the access addresses with a decreasing popularity among networked users, so as to obtain the most favorite adjusted recommended ranking of networked users within the statistical period.

[0112] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation manners described. Obviously, many modifications and changes can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the present invention, so that those skilled in the relevant technical fields can well understand and utilize the present invention. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. An Internet push system based on big data, including a management center, characterized in that, The management center is connected to a login collection module, a data processing module, a data analysis module, and an encryption recommendation module; The login collection module is used to register and log in users, obtain networked users, and collect the behavior data of networked users during the Internet service process. The behavior data includes access addresses, click behaviors, and stay times; The data processing module is used to set a statistical period, statistically classify the collected access addresses within the statistical period to obtain the number of address categories, perform address matching on the click behaviors to obtain address matching behaviors, and construct a user access form based on the obtained access addresses, number of address categories, address matching behaviors, and stay times; The data analysis module is used to obtain the access preference according to the obtained user access form, sort the access addresses according to the access preference and extract keywords to obtain access keywords, and sort the obtained keywords to obtain the keyword sorting; The encryption recommendation module is used to perform character conversion on the access keywords and access addresses to obtain the original keyword sequence and the original address sequence, set the base number and the initial vector according to the original address sequence, obtain the initial address vector according to the base number and the initial vector, insert the initial address vector into the original keyword sequence and perform block division to obtain word sequence segments, perform base encryption on the obtained word sequence segments to obtain the ciphertext recommendation sequence, perform key acquisition permission review on networked users through the management center, decrypt the ciphertext recommendation sequence corresponding to the networked users with successful review to obtain the feature plaintext segments, and perform optimized screening on the obtained feature plaintext segments to obtain the adjusted recommendation sorting.

2. The Internet push system based on big data according to claim 1, characterized in that, Set a user login port and an identification record terminal; Input user information through the user login port and set a login password, upload the obtained user information and login password to the management center, and review the user information through the management center. Generate a login account for the users with successful review according to the user information; Input the login account and login password to log in, and mark the logged-in users as networked users; Collect the comprehensive data of networked users during the Internet service process through the identification record terminal, identify and record the collected comprehensive data to obtain behavior data. The behavior data includes access addresses, click behaviors, and stay times.

3. An Internet push system based on big data according to claim 2, characterized in that, The process of processing the collected behavior data to obtain a user access form includes: Set a statistical period, and statistically analyze the access addresses within the statistical period to obtain the number of address categories; Classify the click behaviors based on the statistical period to obtain the behavior types; Match the obtained access addresses with the behavior types to obtain address matching behaviors, and associate the obtained address matching behaviors with the stay times; Construct a user access form based on the obtained access addresses, number of address categories, address matching behaviors, and stay times.

4. An Internet push system based on big data according to claim 3, characterized in that, The process of the data analysis module analyzing the obtained user access form includes: Obtain the address behavior degree according to the address matching behaviors in the obtained user access form; Obtain the access preference according to the obtained address behavior degree, stay time, statistical period, and number of address categories; Sort the access addresses according to the obtained access preference to obtain a push sorting; Set a key extraction end, and according to the obtained push sorting, extract keywords from the access addresses through the key extraction end to obtain access keywords; Associate the obtained access addresses with the access keywords, count the repetition degree of the access keywords to obtain the corresponding number of addresses, sort the obtained corresponding number of addresses to obtain keyword sorting, and associate the obtained keyword sorting and push sorting with the networked users.

5. An Internet push system based on big data according to claim 4, characterized in that, Perform character conversion on the access keywords in the obtained keyword sorting to obtain a keyword original sequence, and perform character conversion on the obtained access addresses to obtain an address original sequence; Set a base number and an initial vector according to the obtained address original sequence, and obtain an initial address vector according to the obtained base number and initial vector; Insert the obtained initial address vector into the keyword original sequence to obtain an original key sequence, divide the obtained original key sequence into blocks to obtain word sequence segments, and mark the initial address vectors in the word sequence segments as decoding segments.

6. The Internet push system based on big data according to claim 5, characterized in that, The process of performing radix encryption on the obtained word sequence segments includes: Set a seed radix according to the obtained word sequence segments, and obtain a radix product according to the obtained seed radix; Obtain a characteristic function according to the obtained seed radix and radix product; Set a convention radix, and obtain a private radix according to the obtained characteristic function and convention radix; Obtain a ciphertext word segment according to the obtained convention radix, radix product and word sequence segments; Mark the obtained radix product and private radix as a locked key group, associate the obtained locked key group with the login password of the networked user, and upload the obtained locked key group to the management center; Combine the ciphertext word segments according to the obtained original key sequence to obtain a ciphertext key group, and mark the decoding segments in the ciphertext key group as characteristic decoding segments; Sort the obtained ciphertext key group according to the keyword sorting to obtain a ciphertext recommendation sequence.

7. An Internet push system based on big data according to claim 6, characterized in that, Upload the obtained ciphertext recommendation sequence to the networked user, send a plaintext response to the management center, and send a password matching instruction to the networked user through the management center; The networked user inputs the login password to match the login password associated with the locked key group, and grants the key acquisition permission to the networked user with successful matching; The networked user obtains the locked key group associated with the login password according to the obtained key acquisition permission; Obtain a characteristic plaintext segment according to the obtained locked key group and characteristic decoding segment; Sort the characteristic plaintext segment according to the obtained keyword sorting to obtain a preselected address sorting; 8. An Internet push system based on big data according to claim 7, characterized in that, Set a push window, and upload the obtained preselected address sorting to the push window; The networked user selects the preselected recommended addresses through the push window, marks the selected characteristic plaintext segment as the selected access address, and obtains the favorite behavior degree of the selected access address according to the obtained address behavior degree; Set a preference threshold, and mark the access addresses corresponding to the characteristic plaintext segments with a favorite behavior degree less than the preference threshold as insensitive access addresses; Optimize and sort the obtained insensitive access addresses to obtain an adjusted recommendation sorting.

Images