Method for locating target person through association of dark network clues and public opinion events
By obtaining the keywords of dark web target clues and establishing a sensitive account portrait indicator library, we can explore key figures in public opinion events, and use correlation analysis and multi-dimensional portrayal, combined with image search and face recognition technology, we can solve the targets in the dark web The problem of character unrecognition and positioning has achieved the positioning of target characters related to dark web clues and public opinion events, and improved the security of the network environment.
Patent Information
- Application Number
- CN202410271582.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-03-11
- Publication Date
- 2024-11-12
AI Technical Summary
The prior art is difficult to identify and locate organizers and participants of illegal activities in the dark web, and it is impossible to effectively link dark web clues and targeted people in public opinion events.
By obtaining dark web target clue keywords, establishing a database of sensitive account portrait indicators, and discovering key figures in public opinion events, using correlation analysis and multi-dimensional portrayal of target figures, combining picture search and face recognition technology Conduct cross-platform association.
The target person positioning is achieved for the connection between dark web clues and public opinion events, and through big data analysis and multi-dimensional characterization, the security of the network environment is improved.
Smart Images

Figure CN118939870A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of big data mining, and in particular to a method for locating a target person by associating dark web clues with public opinion events. Background Art
[0002] The rapid development of Internet technology has brought convenience to human society while also bringing many new problems. Among them, a large number of illegal activities are usually carried out in the dark web. Due to the encryption and confidentiality of the dark web, and the participants are usually anonymous, it is not easy to track the real geographical location and identity of the user; therefore, the current monitoring and analysis of the dark web can only realize the mining of illegal intelligence clues, and it is impossible to track and locate the organizers and participants of illegal events.
[0003] However, public opinion monitoring based on big data mining technology can achieve real-time monitoring and analysis of posts and other information of specific target accounts on public social platforms, and mine information that may include name, gender, age, mobile phone number, email address, education level, work experience, main remarks, social relations, etc. Therefore, it is necessary to associate the dark web with key accounts on social networks, and use technologies such as image search and face recognition, combined with transaction information mined in the dark web, to achieve a multi-dimensional portrait of the target person to maintain the security of the network environment. Summary of the invention
[0004] The purpose of the present invention is to provide a method for associating dark web clues with public opinion events to locate target persons, and to solve the following technical problems:
[0005] How to solve the problem that the current dark web target clues cannot identify and locate the corresponding key figures;
[0006] The purpose of the present invention can be achieved through the following technical solutions:
[0007] A method for locating a target person by associating dark web clues with public opinion events, the method comprising the following steps:
[0008] S1. Obtain sensitive accounts through dark web target clue keywords;
[0009] S2. Establish a dark web sensitive account portrait indicator library;
[0010] S3. Discover key figures in public opinion events and conduct correlation analysis between key figures’ accounts and sensitive dark web accounts;
[0011] S4. Draw a portrait of the target person’s true identity based on correlation analysis.
[0012] As a further technical solution, the process of obtaining sensitive accounts includes:
[0013] S11. Establish a dark web target clue classification mechanism, and divide each target clue into several categories according to the classification mechanism;
[0014] S12, configuring a keyword library for each type of target clue;
[0015] S13. Based on the target clue keyword library, comprehensive information is collected on different types of dark web sites and sensitive accounts are extracted.
[0016] As a further technical solution, the process of obtaining sensitive accounts also includes:
[0017] S14. Based on target clues, simulate human operations through robot accounts to join groups and collect sensitive accounts.
[0018] As a further technical solution, the process of establishing a dark web sensitive account portrait indicator library includes:
[0019] S21. Use the MD5 algorithm to perform hash operations on sensitive accounts, perform duplication judgment and deduplication processing;
[0020] S22. Monitor and analyze sensitive accounts, establish a dark web sensitive account resource library, and classify dark web sensitive accounts according to the source target clues of dark web sensitive accounts;
[0021] S23. Expand the profile indicators of each sensitive dark web account through continuous monitoring.
[0022] As a further technical solution, the content indicated by the portrait indicator includes the personal information and historical remarks of the corresponding dark web sensitive account.
[0023] As a further technical solution, the process of mining key figures in public opinion events includes:
[0024] S31. Monitor various social media and public websites, and use automatic data clustering algorithms to discover public opinion events
[0025] S32. By setting keywords for major public opinion events, monitor the information of major public opinion events found on various social media and public websites;
[0026] S33. Obtain the key person account corresponding to the key person in the public opinion event through keyword association search.
[0027] As a further technical solution, the methods for obtaining key person accounts include:
[0028] Calculate the interaction volume of social accounts in public opinion events. The interaction volume is the sum of the likes, reposts, and comments of the social accounts in a unit of time. Extract social accounts whose interaction volume exceeds a preset value and mark them as key accounts.
[0029] As a further technical solution, the process of correlation analysis between key person accounts and sensitive dark web accounts includes:
[0030] By comparing the account names of key figures with sensitive accounts on the dark web, as well as the account and person information, a preliminary match is made;
[0031] Use image search and face recognition technology to link cross-platform social accounts, and add the linked key person accounts to the target library;
[0032] Based on the Euclidean distance algorithm, the similarity between the key figures' speech keywords and each type of clue keywords on the dark web is calculated. The calculation results indicate the similarity between the key figures' speech keywords and each type of clue keywords, and a group of key figures' speech keywords whose similarity exceeds the preset standard are selected.
[0033] Beneficial effects of the present invention:
[0034] The present invention mines key social accounts through public opinion events, analyzes the key social accounts, portrays the key social accounts, and associates them with real target persons; at the same time, the public opinion event keywords are clustered, and through similarity calculation, a group of dark web sensitive accounts with the largest similarity value are selected to be associated with the key accounts of public opinion events, and a target library is established and continuously enriched. Through big data analysis and mining technology, valuable information is mined, and through image search, face recognition and other technologies, the real identity portrait of the target person is portrayed in multiple dimensions. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] The present invention will be further described below in conjunction with the accompanying drawings.
[0036] Figure 1 A flow chart of the main steps of the method provided by the present invention for associating dark web clues with public opinion events to locate target persons. DETAILED DESCRIPTION
[0037] 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.
[0038] See also Figure 1 As shown, the present invention is a method for associating dark web clues with public opinion events to locate a target person, and the method comprises the following steps:
[0039] S1. Obtain sensitive accounts through dark web target clue keywords;
[0040] S2. Establish a dark web sensitive account portrait indicator library;
[0041] S3. Discover key figures in public opinion events and conduct correlation analysis between key figures’ accounts and sensitive dark web accounts;
[0042] S4. Draw a portrait of the target person’s true identity based on correlation analysis.
[0043] Through the above technical solution, this embodiment can combine the dark web with public opinion events, and lock in the target person by analyzing and judging the correlation between key figures in public opinion events and sensitive accounts on the dark web, and comprehensively characterize the sensitive accounts.
[0044] The process of obtaining sensitive accounts includes:
[0045] S11. Establish a dark web target clue classification mechanism, and divide each target clue into several categories according to the classification mechanism;
[0046] S12, configuring a keyword library for each type of target clue;
[0047] S13. Based on the target clue keyword library, comprehensive information is collected on different types of dark web sites and sensitive accounts are extracted.
[0048] The process of obtaining sensitive accounts also includes:
[0049] S14. Based on target clues, simulate human operations through robot accounts to join groups and collect sensitive accounts.
[0050] Through the above technical scheme, this embodiment provides a process for obtaining sensitive accounts. Specifically, first, a dark web target clue classification mechanism is established, and each target clue is divided into several categories according to the classification mechanism, and then a keyword library is configured for each type of illegal clue. Since the types of dark web information are diverse, the language types are many, and the colloquial expressions and argot are rich, a threat clue corpus and an argot knowledge base are first established and synchronized with each type of intelligence clue keyword library in real time to ensure the accuracy of the establishment and classification of the clue keyword library, improve the accuracy of weight calculation and account screening, and according to the target clue keyword library, for different types of dark web sites, the sites are comprehensively collected by directional, crawling, meta-search, etc. to extract sensitive accounts. The specific process of collecting sensitive accounts can be based on the target clues, through robot accounts to simulate human operations to join the group and collect sensitive accounts.
[0051] The process of building a dark web sensitive account portrait indicator library includes:
[0052] S21. Use the MD5 algorithm to perform hash operations on sensitive accounts, perform duplication judgment and deduplication processing;
[0053] S22. Monitor and analyze sensitive accounts, establish a dark web sensitive account resource library, and classify dark web sensitive accounts according to the source target clues of dark web sensitive accounts;
[0054] S23. Expand the profile indicators of each sensitive dark web account through continuous monitoring.
[0055] Through the above technical solution, this embodiment provides a process for establishing a dark web sensitive account portrait index library. Specifically, first, the sensitive account is hashed through the MD5 algorithm to perform repeatability judgment and deduplication processing, and then the sensitive account is monitored and analyzed to establish a dark web sensitive account resource library. According to the source target clue type of the dark web sensitive account, the dark web sensitive account is correspondingly classified, and finally, through continuous monitoring, the portrait index of each dark web sensitive account is expanded. Through long-term monitoring and continuous enrichment of sensitive dark web account portrait indicators, a sensitive account portrait index library can be constructed.
[0056] The content indicated by the portrait indicators includes the personal information and historical comments of the corresponding dark web sensitive account. The personal information includes the account registration time, the person's age, gender and the content of interest.
[0057] The process of discovering key figures in public opinion events includes:
[0058] S31. Monitor various social media and public websites, and use automatic data clustering algorithms to discover public opinion events
[0059] S32. By setting keywords for major public opinion events, monitor the information of major public opinion events found on various social media and public websites;
[0060] S33. Obtain the key person account corresponding to the key person in the public opinion event through keyword association search.
[0061] Through the above technical solution, this embodiment provides a process for mining key figures in public opinion events. Specifically, first, by continuously improving the sensitive keyword rule library related to economy, education, medical care, public health, employment, culture, etc., various social media and public websites are monitored in all directions, and the data automatic clustering algorithm is used to discover public opinion events. Then, by setting keywords for major public opinion events, the information of the major public opinion events discovered on various social media and public websites is monitored, and finally, the key person accounts corresponding to the key figures in the public opinion events are obtained through keyword association retrieval.
[0062] Ways to obtain key person accounts include:
[0063] Calculate the interaction volume of social accounts in public opinion events. The interaction volume is the sum of the likes, reposts, and comments of the social accounts in a unit of time. Extract social accounts whose interaction volume exceeds a preset value and mark them as key accounts.
[0064] Through the above technical solution, this embodiment provides a method for obtaining key person accounts. Specifically, the interaction volume of social accounts in public opinion events can be calculated. The interaction volume is the sum of the likes, reposts and comments of social accounts in a unit time. Social accounts with interaction volume exceeding the preset value are extracted and marked as key accounts. For the opinions, speeches and articles published in the event, sentiment judgment is performed through the sentiment vocabulary, and the interaction volume of opinions, speeches and articles judged as negative is ranked, and the top accounts that publish negative opinions, speeches and articles are extracted. In addition, the comments under the promoters, detonating points and articles with more interactions of the event can be analyzed, and the number of likes of the comments is ranked, and the comment accounts with the number of likes reaching the preset value are extracted. It should be noted that a promoter refers to a person whose fans of the reposter are more than 100,000, but the number of reposts is less than 10,000, and a detonating point refers to a forwarded article that is reposted by 10,000 people.
[0065] The process of analyzing the correlation between key person accounts and sensitive dark web accounts includes:
[0066] By comparing the account names of key figures with sensitive accounts on the dark web, as well as the account and person information, a preliminary match is made;
[0067] Use image search and face recognition technology to link cross-platform social accounts, and add the linked key person accounts to the target library;
[0068] Based on the Euclidean distance algorithm, the similarity between the key figures' speech keywords and each type of clue keywords on the dark web is calculated. The calculation results indicate the similarity between the key figures' speech keywords and each type of clue keywords, and a group of key figures' speech keywords whose similarity exceeds the preset standard are selected.
[0069] Through the above technical solution, in this embodiment, a preliminary match is made by comparing the account names of key person accounts and sensitive accounts on the dark web, as well as the account and person information;
[0070] Through image search and face recognition technology, cross-platform social accounts are associated, and the associated key figures are added to the target library. Finally, the similarity between the key figures' speech keywords and each type of clue keywords in the dark web is calculated based on the Euclidean distance algorithm. The calculation results indicate the similarity between the key figures' speech keywords and each type of clue keywords. A group of key figures' speech keywords with similarity exceeding the preset standard are selected. In order to effectively associate the dark web account portrait with the key accounts in the public opinion events, the naive Bayes classification algorithm can be used. This algorithm is based on the Bayesian theorem and performs probabilistic processing on the data, which can effectively perform classification and association analysis. By training the naive Bayes classifier, the dark web account portrait indicators are integrated with the key indicators of the accounts in major public opinion events, and an association model is established. The model can automatically identify the key accounts in the public opinion that are highly correlated with the dark web accounts, providing support for analysis and monitoring.
[0071] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.
Claims
1. A method for locating a target person by associating dark web clues with public opinion events, characterized in that: The method comprises the following steps: S1. Obtain sensitive accounts through dark web target clue keywords; S2. Establish a dark web sensitive account portrait indicator database; S3. Discover key figures in public opinion events and conduct correlation analysis between key figures’ accounts and sensitive dark web accounts; S4. Draw a portrait of the target person’s true identity based on correlation analysis.
2. According to claim 1, a method for locating a target person by associating dark web clues with public opinion events is characterized in that: The process of obtaining sensitive accounts includes: S11. Establish a dark web target clue classification mechanism, and divide each target clue into several categories according to the classification mechanism; S12, configuring a keyword library for each type of target clue; S13. Based on the target clue keyword library, comprehensive information is collected on different types of dark web sites and sensitive accounts are extracted.
3. According to claim 2, a method for locating a target person by associating dark web clues with public opinion events is characterized in that: The process of obtaining sensitive accounts also includes: S14. Based on target clues, simulate human operations through robot accounts to join groups and collect sensitive accounts.
4. According to claim 3, a method for locating a target person by associating dark web clues with public opinion events is characterized in that: The process of building a dark web sensitive account portrait indicator library includes: S21. Use the MD5 algorithm to perform hash operations on sensitive accounts, perform duplication judgment and deduplication processing; S22. Monitor and analyze sensitive accounts, establish a dark web sensitive account resource library, and classify dark web sensitive accounts according to the source target clues of dark web sensitive accounts; S23. Expand the profile indicators of each sensitive dark web account through continuous monitoring.
5. According to claim 4, a method for locating a target person by associating dark web clues with public opinion events is characterized in that: The content indicated by the portrait indicators includes the personal information and historical remarks of the corresponding sensitive dark web account.
6. According to claim 5, a method for locating a target person by associating dark web clues with public opinion events is characterized in that: The process of discovering key figures in public opinion events includes: S31. Monitor various social media and public websites, and use automatic data clustering algorithms to discover public opinion events S32. By setting keywords for major public opinion events, monitor the information of major public opinion events found on various social media and public websites; S33. Obtain the key person account corresponding to the key person in the public opinion event through keyword association search.
7. According to claim 6, a method for locating a target person by associating dark web clues with public opinion events is characterized in that: Ways to obtain key person accounts include: Calculate the interaction volume of social accounts in public opinion events. The interaction volume is the sum of the likes, reposts, and comments of the social accounts in a unit of time. Extract social accounts whose interaction volume exceeds a preset value and mark them as key accounts.
8. According to claim 7, a method for locating a target person by associating dark web clues with public opinion events is characterized in that: The process of analyzing the correlation between key person accounts and sensitive dark web accounts includes: By comparing the account names of key figures with sensitive accounts on the dark web, as well as the account and person information, a preliminary match is made; Use image search and face recognition technology to link cross-platform social accounts, and add the linked key person accounts to the target library; Based on the Euclidean distance algorithm, the similarity between the key figures' speech keywords and each type of clue keywords on the dark web is calculated. The calculation results indicate the similarity between the key figures' speech keywords and each type of clue keywords, and a group of key figures' speech keywords whose similarity exceeds the preset standard are selected.
Citation Information
Cited By
Bad netizen tracing method, device, equipment and medium
CN121125313A
