Language resource library system based on data security control

Abstract

The invention provides a language resource library system based on data security control, and the system comprises an acquisition module which collects language resources in a manner of manual input and crawler program directional capture, obtains a language data set, and stores the language data set in a database; the processing module develops a software program according to a preset metadata specification, and processes and manually marks the language data set to obtain a standard data set; the management module is used for performing data management, user management and log management in combination with the data access control model and the network security monitoring model; the presentation module is used for providing data retrieval and statistical analysis functions according to the data type and carrying out visualized display of the data; and the product library module is used for developing a language resource platform. By establishing a perfect security mechanism, unauthorized access and data leakage can be effectively prevented, the security of language resources in cross-border transmission and use is improved, and efficient and stable operation of the system is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of language resource library data security management, and in particular to a language resource library system based on data security control. Background Art

[0002] A language resource database refers to a database that stores and organizes language resources. For example, a language resource database for language protection aims to comprehensively record the characteristics and attributes of a language so that it can be reproduced when the language is endangered. For example, the Chinese Language Resource Protection Project took five years to complete a language resource survey and collection at 1,712 survey points, covering 123 languages ​​and all Chinese dialects, and established the world's largest language resource database.

[0003] For another example, there are nine countries in the South China Sea region, including Vietnam, Cambodia, Thailand, Malaysia, Singapore, Indonesia, the Philippines and Brunei, which have close exchanges with China. Countries around the South China Sea have diverse languages, with incomplete statistics showing that there are more than 1,000 languages. Not many of them are currently being paid attention to, and a considerable number have not been recorded and studied. In addition, in the process of cross-border data transmission and sharing, the data security of language resources is becoming increasingly prominent. Summary of the invention

[0004] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide a language resource library system based on data security control, which can effectively prevent unauthorized access and data leakage by establishing a complete security mechanism, improve the security of language resources in cross-border transmission and use, and ensure that the system can operate efficiently and stably.

[0005] To achieve the above object, the present invention provides the following solution: a language resource library system based on data security control, comprising:

[0006] A collection module, used to collect language resources by manual input and directional crawling by crawler programs, obtain a language data set, and store the language data set in a database;

[0007] A processing module, used to develop a software program according to a preset metadata specification, to process and manually annotate the language data set to obtain a standard data set;

[0008] Management module, which is used to combine the data access control model and the network security monitoring model to perform data management, user management and log management;

[0009] The presentation module is used to provide data retrieval and statistical analysis functions according to the data type, and to visualize the data;

[0010] Product library module, used to develop language resource platform using language processing and application tools;

[0011] Wherein, the acquisition module, the processing module, the management module, the presentation module and the product library module are interconnected.

[0012] Optionally, the acquisition module includes:

[0013] Manual input unit, used to manually input paper academic literature and language policy literature;

[0014] Directed crawling unit, used to develop web crawler tools to collect and vertically crawl language resource information on seed websites;

[0015] The storage unit is used to store the collected data into the database system in the form of XML files and text files.

[0016] Optionally, the processing module includes:

[0017] A metadata standard unit, used to design a data item set and a data item definition according to the language data set, obtain a data rule set, and complete the metadata specification;

[0018] The data annotation unit is used to program-annotate the language data set according to a preset metadata specification by using a text segmentation method, a text classification and clustering method, and an automatic summarization method, and then manually proofread the program annotation of the language data set to obtain a standard data set.

[0019] Optionally, the management module includes:

[0020] User management unit, used for user registration, information modification, account deletion and permission adjustment;

[0021] The data management unit is used to add, modify and review data, complete data updates, and then program-mark and manually proofread the updated data to complete the quality control of the newly added data;

[0022] The log management unit is used to perform statistics and analysis on the standard data set.

[0023] Optionally, the management module further includes:

[0024] An access control unit, used to design a role-based access control algorithm according to the data type of the standard data set, perform role management and authority management, and complete the management and control of user access rights; wherein the data type includes a vocabulary, a sentence list, a conversation, a translation text, an audio file, and a language learning material;

[0025] The network security monitoring unit is used to build and train a security detection model using anomaly detection algorithm and isolation forest algorithm, and then use the security detection model to monitor network traffic in real time and detect abnormal behavior data.

[0026] Optionally, the access control algorithm includes:

[0027] The User Component sub-unit includes students for accessing language learning materials, teachers for accessing and modifying language learning materials and assessing the learning of said students, translators for accessing and modifying translated texts, and administrators for managing all users and resources;

[0028] The role assignment component subunit includes a learner role for assigning permissions to the student, an educator role for assigning permissions to the teacher, a translator role for assigning permissions to the translator, and an administrator role for assigning permissions to the administrator;

[0029] Permission subunits, including read permissions for accessing specific resources, write permissions for modifying specific resources, create permissions for adding new resources, and delete permissions for deleting specific resources;

[0030] Resources subunit, including language learning materials, translated texts, and user information

[0031] The logic mapping subunit is used to map the user to one or more roles and map the role to one or more permissions by using the mapping function, and then use the permission checking function to check whether the user and the role have the corresponding permissions.

[0032] Optionally, the expression of the anomaly detection algorithm is:

[0033]

[0034] Among them, S is the safety score, w i is the weight, f i (x) is the detection index function, and n is the number of detection indicators.

[0035] Optionally, the safety detection model includes:

[0036] A feature extraction subunit is used to extract key features from network traffic; the key features include packet size, packet interval time, source IP address, destination IP address and packet protocol type;

[0037] A data preprocessing subunit, used to standardize and normalize the key features

[0038] A model training subunit is used to select an anomaly detection model, and train the selected anomaly detection model using an isolation forest algorithm and the standard data set to obtain a safety detection model;

[0039] an anomaly detection subunit, used to obtain real-time network traffic data, extract traffic feature vectors from the real-time network traffic data, standardize and normalize the traffic feature vectors to obtain standard feature vectors, input the standard feature vectors into an isolation forest model, and calculate using an anomaly score algorithm to obtain an anomaly score;

[0040] The abnormality judgment subunit is used to preset an abnormality threshold and judge whether the abnormality score is greater than the abnormality threshold. If so, it is judged that abnormal behavior exists.

[0041] Optionally, the expression of the anomaly score algorithm is:

[0042]

[0043] H(n)=ln(n)+γ

[0044] Among them, E(h(x)) is the average path length of X in each isolated tree, c(n) is the adjustment factor, H(n) is the harmonic number, and γ is the Euler constant.

[0045] Optionally, the presentation module includes:

[0046] A portal information presentation unit is used to construct an information query system, a portal website and a historical electronic library; wherein the information query system includes accurate and fuzzy search functions by word, search functions by additional components, and search functions by paragraph and by article;

[0047] The intelligent report presentation unit is used to design monitoring report content and monitoring database, and uses data mining and programming to build a resource library statistical analysis system for feedback statistical report information.

[0048] The present invention discloses the following technical effects by providing a language resource library system based on data security control:

[0049] 1. Through the combination of role-based access control (RBAC) and isolated forest anomaly detection technology, a complete set of security mechanisms is constructed, which shows high detection accuracy and response speed in practical applications, significantly improves the security and data processing efficiency of the system, and provides effective protection for the cross-border transmission and use of language resource libraries. This invention not only improves data security, but also ensures the standardization and credibility of language resources in use, providing more powerful data support and technical guarantees for national security and cultural exchanges.

[0050] 2. The language resource library system designed by the present invention comprehensively covers all aspects of data collection, processing, storage, management and application. In the engineering design of the system, all modules are developed and iteratively optimized according to standardized processes to ensure that the system can run efficiently and stably.

[0051] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0053] Figure 1 A schematic diagram of a system architecture provided by an embodiment of the present invention;

[0054] Figure 2 A schematic diagram of system logic provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0055] 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.

[0056] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0057] like Figure 1-2 As shown, the present invention provides a language resource library system based on data security control, taking the South China Sea as an example:

[0058] 1. The present invention includes a collection module, a processing module, a management module, a presentation module and a product library module which are interconnected.

[0059] 1. Collection module

[0060] It is used to collect language resources by manual input and directional crawling of crawler programs, obtain language data sets, and store the language data sets in a database; the collection module includes:

[0061] 1.1 Manual Entry Unit

[0062] Manual input of paper academic literature and language policy literature. Taking the South China Sea as an example, the scope of language resources manually input includes: paper academic literature and language policy literature of countries surrounding the South China Sea.

[0063] 1.2 Directional grabbing unit

[0064] Develop web crawler tools to collect and vertically crawl language resource information on seed websites. Web crawler targeted crawling work is carried out from two aspects: seed website collection and vertical crawling. Seed information includes articles published on websites by major media organizations of countries surrounding the South China Sea in official languages ​​or major common languages, as well as various information texts published on government websites of various countries.

[0065] The seed website collection adopts manual investigation method, and the investigation is carried out in different regions, which is operational and accurate. For example, domestic investigation and collection: manual investigation and collection or small-scale conference investigation and collection are carried out in the country; overseas cooperation base station investigation and collection of Jinan University: Vietnam, Cambodia, Thailand, Malaysia, Singapore, Indonesia, Philippines and Brunei. These countries are the partners of the Chinese Language and Literature College of Jinan University in exchange, school operation and language research. These on-site institutions can complete the information investigation and collection of the corresponding overseas regions.

[0066] Periodically apply thematic vertical search engine technology to capture the most relevant resources in seeds.

[0067] 1.3 Storage Unit

[0068] The collected data is stored in the database system in the form of XML files and text files. The construction of a dynamic resource library for storing language resources can be carried out from multiple aspects. Commercial database systems such as ORACLE and SQLSERVER can be used, and data can also be stored in the form of XML files, text files, etc.

[0069] 2. Processing module

[0070] According to the preset metadata specification, a software program is developed to process and manually annotate the language data set to obtain a standard data set; the processing module includes:

[0071] 2.1 Metadata Standard Unit

[0072] Before processing language resources, it is necessary to design the metadata specification of the resource library. The metadata specification is a collection of all rules that describe specific resource objects. According to the language data set, the data item set and data item definition are designed to obtain the data rule set and complete the metadata specification. After studying and designing the specification, the processing of resources can be carried out in a regular and orderly manner.

[0073] 2.2 Data Annotation Unit

[0074] According to the preset metadata specification, the language data set is program-annotated using a text segmentation method, a text classification and clustering method, and an automatic summarization method, and the program annotations of the language data set are manually proofread to obtain a standard data set.

[0075] Manual proofreading is carried out by management personnel in accordance with operating specifications to proofread the program annotations of resources. After the proofreading is completed, expert intervention is introduced to form a reliable and usable mature resource library to serve the content retrieval and statistical reporting functions of the upper layer of the resource library.

[0076] 3. Management module

[0077] The management module is one of the core modules of the language resource library, which is mainly responsible for user management and permission control. It combines the data access control model and the network security monitoring model to perform data management, user management and log management; the management module includes:

[0078] 3.1 User Management Unit

[0079] It is used to register users, modify information, delete accounts and adjust permissions, etc., to ensure that all types of users operate within the appropriate scope of permissions, thereby ensuring the security of the system and the effective management of resources.

[0080] 3.2 Data Management Unit

[0081] Perform operations such as adding, modifying and reviewing data to continuously update and improve data content, and then perform program annotation and manual proofreading on the updated data to complete the quality control of the newly added data. Data resources that have been edited and reviewed will be marked as standardized resources to ensure the quality and authority of the resources in the library. When using the resource library, users can obtain the necessary permissions through the management system to browse, download, and search for resources.

[0082] 3.3 Log Management Unit

[0083] Statistics and analysis of the standard data set are performed to facilitate users to obtain and utilize comprehensive information.

[0084] 3.4 Access Control Unit

[0085] The cross-border transmission and sharing of language resources involves a large amount of sensitive information, such as scientific research data, translation data, personal information and cultural materials. These data are vulnerable to attacks and theft during transmission and storage. Therefore, advanced encryption and access control technologies must be used to ensure the confidentiality and integrity of the data.

[0086] According to the data type of the standard data set, a role-based access control algorithm (i.e., RBAC model algorithm) is designed to perform role management and authority management to complete the management and control of user access rights; wherein the data types include vocabulary lists, sentence lists, conversations, translated texts, audio files, and language learning materials.

[0087] 3.41 Access Control Matrix

[0088] Access control is an important means of limiting data access rights to ensure that only authorized users can access and operate specific data. This paper adopts the RBAC (Role-Based Access Control) model and makes a detailed design based on the data characteristics of corpus language resources. Its access control matrix is:

[0089] A[i][j]={1,0}

[0090] Where A is the access control matrix, i and j represent the indexes of users and resources respectively. If user i has access rights to resource j, the value is 1, and if user i does not have access rights to resource j, the value is 0.

[0091] 3.42 Access control algorithm (RBAC model algorithm) includes:

[0092] The user component subunit (User, U) includes students (Student) for accessing language learning materials, teachers (Teacher) for accessing and modifying language learning materials and evaluating the learning situation of the students, translators (Translator) for accessing and modifying translated texts, and administrators (Admin) for managing all users and resources.

[0093] The role allocation component subunit (Role, R) includes a learner role (Learner) for allocating permissions to the student, an educator role (Educator) for allocating permissions to the teacher, a translator role (TranslatorRole) for allocating permissions to the translator, and an administrator role (Manager) for allocating permissions to the administrator;

[0094] The permission subunit (Permission, P), including the read permission (Read, r) for accessing a specific resource, the write permission (Write, w) for modifying a specific resource, the create permission (Create, c) for adding a new resource, and the delete permission (Delete, d) for deleting a specific resource;

[0095] Resource subunit (Resource, Res), including language learning materials (LearningMaterials), translation texts (TranslationTexts) and user information (UserInfo). Among them, language learning materials include learning resources such as text, audio, video, etc.; translation texts include texts to be translated and translated; user information system basic information of all users.

[0096] As shown in Table 1:

[0097] Table 1 Permission allocation matrix

[0098] Role resource Read (r) Write (w) Create(c) Delete (d) Learner Language Learning Materials yes no no no Educator Language Learning Materials yes yes yes no TranslatorRole TranslationTexts yes yes yes no Manager All resources yes yes yes yes

[0099] The logic mapping subunit is used to map the user to one or more roles and map the role to one or more permissions by using the mapping function, and then use the permission checking function to check whether the user and the role have the corresponding permissions.

[0100] The logic of the 3.43 algorithm is described as follows:

[0101] User role mapping:

[0102] U R :U→R

[0103] Among them, U is the user set, R is the role set, and the mapping function U R Map users to one or more roles.

[0104] Role permission mapping:

[0105] R P :R→P

[0106] Among them, P is the permission set, and the mapping function R P Map a role to one or more permissions.

[0107] Permission check:

[0108] Assume that user u needs to access resource res to perform operation op, then the permission check function is:

[0109]

[0110] 3.44 The access control algorithm implementation process is:

[0111] (1) Role definition:

[0112] CREATE TABLE Roles(

[0113] RoleID INT PRIMARY KEY,

[0114] RoleName VARCHAR(50) );

[0116] INSERT INTO Roles(RoleID,RoleName)VALUES(1,'Learner');INSERT INTO Roles(RoleID,RoleName)VALUES(2,'Educator');

[0117] INSERT INTO Roles(RoleID,RoleName)VALUES(3,'TranslatorRole');

[0118] INSERT INTO Roles(RoleID,RoleName)VALUES(4,'Manager');

[0119] (2) User-defined

[0120] CREATE TABLE Users(

[0121] UserID INT PRIMARY KEY,

[0122] UserName VARCHAR(50),

[0123] RoleID INT,

[0124] FOREIGN KEY(RoleID) REFERENCES Roles(RoleID) );

[0126] INSERT INTO Users(UserID,UserName,RoleID)VALUES(1,'Alice',1);INSERT INTO Users(UserID,UserName,RoleID)VALUES(2,'Bob',2);INSERT INTO Users(UserID,UserName,RoleID)VALUES(3,'Charlie',3);INSERT INTO Users(UserID,UserName,RoleID)VALUES(4,'Dave',4);(3) Permission definition

[0127] CREATE TABLE Permissions(

[0128] PermissionID INT PRIMARY KEY,

[0129] Resource VARCHAR(50),

[0130] PermissionType VARCHAR(10) );

[0132] INSERT INTO Permissions(PermissionID, Resource, PermissionType) VALUES(1, 'LearningMaterials', 'r');

[0133] INSERT INTO Permissions(PermissionID, Resource, PermissionType) VALUES(2, 'LearningMaterials', 'w');

[0134] INSERT INTO Permissions(PermissionID, Resource, PermissionType) VALUES(3, 'TranslationTexts', 'r');

[0135] INSERT INTO Permissions(PermissionID, Resource, PermissionType) VALUES(4, 'TranslationTexts', 'w');

[0136] INSERT INTO Permissions(PermissionID, Resource, PermissionType) VALUES(5, 'All', 'c');

[0137] INSERT INTO Permissions(PermissionID, Resource, PermissionType) VALUES(6, 'All', 'd');

[0138] (4) Role Permission Assignment

[0139] CREATE TABLE RolePermissions(

[0140] RoleID INT,

[0141] PermissionID INT,

[0142] FOREIGN KEY(RoleID)REFERENCES Roles(RoleID),

[0143] FOREIGN KEY(PermissionID)REFERENCES Permissions(PermissionID) );

[0145] --Assign permissions to learner roles

[0146] INSERT INTO RolePermissions(RoleID,PermissionID)VALUES(1,1);--Learner,Read LearningMaterials

[0147] --Assign permissions to the educator role

[0148] INSERT INTO RolePermissions(RoleID,PermissionID)VALUES(2,1);--Educator,Read LearningMaterials

[0149] INSERT INTO RolePermissions(RoleID,PermissionID)VALUES(2,2);--Educator,Write LearningMaterials

[0150] INSERT INTO RolePermissions(RoleID,PermissionID)VALUES(2,5);--Educator,Create

[0151] --Assign permissions to translation roles

[0152] INSERT INTO RolePermissions(RoleID,PermissionID)VALUES(3,3);--TranslatorRole,Read TranslationTexts

[0153] INSERT INTO RolePermissions(RoleID,PermissionID)VALUES(3,4); -- TranslatorRole,Write TranslationTexts

[0154] -- Assign permissions for the manager role

[0155] INSERT INTO RolePermissions(RoleID,PermissionID)VALUES(4,1); -- Manager,Read LearningMaterials

[0156] INSERT INTO RolePermissions(RoleID,PermissionID)VALUES(4,2); -- Manager,Write LearningMaterials

[0157] INSERT INTO RolePermissions(RoleID,PermissionID)VALUES(4,3); -- Manager,Read TranslationTexts

[0158] INSERT INTO RolePermissions(RoleID,PermissionID)VALUES(4,4); -- Manager,Write TranslationTexts

[0159] INSERT INTO RolePermissions(RoleID,PermissionID)VALUES(4,5); -- Manager,Create

[0160] INSERT INTO RolePermissions(RoleID,PermissionID)VALUES(4,6); -- Manager,Delete

[0161] Through the above design, the access rights of corpus language resources can be effectively managed and controlled to ensure data security and operational compliance. Through the RBAC model, the access rights of different users to resources can be flexibly managed to avoid the complexity and security risks brought by direct assignment of permissions. At the same time, role-based access control has high scalability and can adapt to systems of different scales and requirements. The RBAC model designed in this paper provides a detailed role and permission management mechanism based on the data characteristics of corpus language resources.

[0162] 3.5 Network Security Monitoring Unit

[0163] It is used to build and train a security detection model using anomaly detection algorithms and isolation forest algorithms, and then use the security detection model to monitor network traffic in real time, detect abnormal behavior data, and promptly discover and respond to potential security threats.

[0164] The expression of the anomaly detection algorithm is:

[0165]

[0166] Among them, S is the safety score, w i is the weight, f i (x) is the detection index function, and n is the number of detection indicators.

[0167] The security detection model includes:

[0168] 3.51 Feature Extraction Subunit

[0169] Extract key features from network traffic; the key features include:

[0170] Packet Size (PS): The size of each data packet.

[0171] Inter-Arrival Time (IAT): The time interval between adjacent data packets.

[0172] Source IP address (SIP): The source IP address of the data packet.

[0173] Destination IP address (DIP): The destination IP address of the data packet.

[0174] The data packet uses the protocol type (PT).

[0175] 3.52 Data Preprocessing Subunit

[0176] The key features are standardized and normalized to ensure that the value range of each feature is similar and improve the detection accuracy of the algorithm. The standardization formula is:

[0177]

[0178] Among them, μ is the mean of the feature and σ is the standard deviation of the feature. The normalization formula is:

[0179]

[0180] 3.53 Model Training Subunit

[0181] An anomaly detection model is selected, and the selected anomaly detection model is trained using the isolation forest algorithm and the standard data set to obtain a safety detection model.

[0182] The Isolation Forest algorithm is used for model training. Isolation Forest is an unsupervised learning algorithm based on decision trees, which is suitable for anomaly detection in high-dimensional data. Assume that the Isolation Forest contains n isolated trees, and the training data of each isolated tree is X i (i=1,2,...,n), the path length is h(X i ). The basic principle of Isolation Forest: Isolation Forest isolates sample points by constructing multiple random trees. The shorter the path length required for a sample point to be isolated, the more likely it is an anomaly. Training process: Randomly select sample points from normal network traffic data. Construct multiple isolation trees (IsolationTree), and each tree randomly selects features for splitting. Calculate the anomaly score based on the path length of the sample point in each isolation tree.

[0183] 3.54 Anomaly Detection Subunit

[0184] It is used to obtain real-time network traffic data, extract traffic feature vectors from the real-time network traffic data, standardize and normalize the traffic feature vectors to obtain standard feature vectors, input the standard feature vectors into the isolation forest model, and calculate using the anomaly score algorithm to obtain an anomaly score.

[0185] The calculation formula of anomaly score S(X) is:

[0186]

[0187] Among them, E(h(x)) is the average path length of X in each isolated tree, and c(n) is the adjustment factor:

[0188]

[0189] Where H(n) is the harmonic number:

[0190] H(n)=ln(n)+γ

[0191] Here, γ is the Euler constant.

[0192] 3.55 Abnormal Judgment Subunit

[0193] An abnormality threshold θ is preset, and it is determined whether the abnormality score is greater than the abnormality threshold. If so, it is determined that abnormal behavior exists. The abnormality judgment formula is:

[0194] If S(X)>θthen X is an anomaly

[0195] 4. Presentation module

[0196] It is used to provide data retrieval and statistical analysis functions according to the data type, and to visualize the data; the presentation module includes:

[0197] 4.1 Portal Information Presentation Unit

[0198] Build an information query system, portal website and historical electronic library; among which:

[0199] Information query system: provides various search methods for resources in the resource library, such as accurate and fuzzy search by word, search by additional components, and search by paragraph and by article, so that users can quickly and accurately obtain the required data information.

[0200] Portal website: Resource information can be automatically generated according to the website category content and updated daily to build a calendar-style news portal. Resource briefings, hot topics, etc. can be pushed to users through Weibo, WeChat public accounts, email, etc.

[0201] Diachronic electronic library: archives, stores and manages massive amounts of diachronic information, providing an electronic library platform for future diachronic research.

[0202] 4.2 Intelligent Report Presentation Unit

[0203] The monitoring report content is designed by artificial investigation and expert methods. Research on the design of monitoring report content is carried out to build a monitoring knowledge base, and the content items of the knowledge base can be dynamically added, deleted and modified as needed. And using data mining and program design, a resource library statistical analysis system for feedback of statistical report information is constructed.

[0204] 5. Product library module

[0205] Develop a language resource platform using language processing and application tools. For example, language research in countries around the South China Sea should not only focus on the importance of language data ontology research, but also on the importance of science and technology in language application. In the information age, people's language communication, learning and research have gradually transitioned to language technology products. The application of science and technology in language is generally reflected in the application of intelligent communication equipment and software, such as using smart phones to communicate, using mobile phone text messages, QQ and other social software, WeChat, email, mobile phone APP software, etc. to chat and communicate. Based on the language resource library system around the South China Sea, various language processing and application tools can be developed, as well as various application software and mobile phone APP software to facilitate scientific research and user use, such as mobile phone APP learning software, mobile phone APP language map software, etc.

[0206] 2. Deploy this system for experimental testing to verify that the present invention has significant advantages in ensuring data transmission security and access control effectiveness.

[0207] 1. Access Control Experiment

[0208] The experimental environment is set up as follows: Dataset: a language resource database, containing 1,000 data records. Role settings: administrator, researcher, student. Permission allocation: permission configuration based on the RBAC model. Access request: simulate access requests from different roles to data.

[0209] Experimental steps:

[0210] (1) Role configuration:

[0211] Administrator: has read and write permissions for all data. Researcher: has read and limited write permissions for some data. Student: has only read permissions for data.

[0212] (2) Access request simulation:

[0213] Each role randomly issues 100 access requests and records the result of each request (successful or rejected).

[0214] 4.1.3 Experimental results analysis, as shown in Table 2:

[0215] Table 2 Experimental results of role access control

[0216] Role Total requests Number of successful requests Rejected Requests administrator 100 100 0 Researchers 100 85 15 student 100 70 30

[0217] All requests from administrators were approved, as expected. 85% of the requests from researchers were approved and 15% were rejected, reflecting the limited permissions. 70% of the student requests were approved and 30% were rejected, in line with the permission configuration that students can only read part of the data. The figure above shows the number of successful and rejected requests for different roles (administrators, researchers, students) when accessing data. It can be seen that all the administrator's requests were approved, while a certain proportion of the requests from researchers and students were rejected, which is in line with the permission configuration of the RBAC model.

[0218] 2. Network security monitoring experiment

[0219] The experimental environment is set up as follows: Dataset: Contains normal network traffic and simulated abnormal traffic, totaling 10,000 data records. Feature extraction: packet size, packet interval, source IP address, destination IP address, protocol type. Model training: Use isolation forest and DFA model training. Anomaly detection: Detect real-time traffic and record the detection results.

[0220] Experimental steps:

[0221] (1) Feature extraction and preprocessing: Extract features from network traffic data and perform standardization and normalization.

[0222] (2) Model training: Use the isolation forest and DFA models to train normal traffic data.

[0223] (3) Anomaly detection: Perform real-time detection on test data and record the detection results (normal or abnormal).

[0224] The experimental results are shown in Table 3:

[0225] Table 3 Network security monitoring experimental results

[0226] Data Types Total records Detected normal number Number of anomalies detected Normal traffic 8000 7900 100 Abnormal traffic 2000 150 1850

[0227] The detection accuracy of normal traffic is 98.75% (7900 / 8000), and the detection accuracy of abnormal traffic is 92.5% (1850 / 2000). The detection effect of the combination of Isolation Forest and DFA model is good, and it can effectively identify abnormal behavior in network traffic. The above figure shows the detection results of normal traffic and abnormal traffic in the network security monitoring experiment. The detection accuracy of normal traffic is 98.75%, and the detection accuracy of abnormal traffic is 92.5%, indicating that the detection algorithm combining Isolation Forest and DFA model is effective in identifying abnormal traffic.

[0228] The access control experiment shows that the permission configuration based on the RBAC model can effectively manage the access rights of different roles and ensure data security. The network security monitoring experiment shows that the detection algorithm combined with the isolation forest model has a high accuracy in identifying abnormal traffic, which helps to timely discover and respond to potential security threats. These experimental results verify the effectiveness and feasibility of the access control model and network security monitoring algorithm proposed in this paper. The results of the access control experiment show that the permission configuration based on the RBAC model can effectively manage the access rights of different roles. The results of the network security monitoring experiment show that it has a high accuracy in identifying abnormal traffic. Future research will further optimize the permission allocation strategy and anomaly detection algorithm to improve the security and management efficiency of the system.

[0229] Therefore, the present invention provides a language resource library system based on data security control and establishes a complete security mechanism, which can effectively prevent unauthorized access and data leakage, improve the security of language resources in cross-border transmission and use, and ensure that the system can operate efficiently and stably.

[0230] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0231] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A language resource library system based on data security control, characterized in that: include: A collection module, used to collect language resources by manual input and directional crawling by crawler programs, obtain a language data set, and store the language data set in a database; A processing module, used to develop a software program according to a preset metadata specification, to process and manually annotate the language data set to obtain a standard data set; Management module, which is used to combine the data access control model and the network security monitoring model to perform data management, user management and log management; The presentation module is used to provide data retrieval and statistical analysis functions according to the data type, and to perform visual display of the data; Product library module, used to develop language resource platform using language processing and application tools; Wherein, the acquisition module, the processing module, the management module, the presentation module and the product library module are interconnected.

2. A language resource library system based on data security control according to claim 1, characterized in that: The acquisition module comprises: Manual input unit, used to manually input paper academic literature and language policy literature; Directed crawling unit, used to develop web crawler tools to collect and vertically crawl language resource information on seed websites; The storage unit is used to store the collected data into the database system in the form of XML files and text files.

3. A language resource library system based on data security control according to claim 2, characterized in that: The processing module comprises: A metadata standard unit, used to design a data item set and a data item definition according to the language data set, obtain a data rule set, and complete the metadata specification; The data annotation unit is used to program-annotate the language data set according to a preset metadata specification by using a text segmentation method, a text classification and clustering method, and an automatic summarization method, and then manually proofread the program annotation of the language data set to obtain a standard data set.

4. A language resource library system based on data security control according to claim 3, characterized in that: The management module comprises: User management unit, used for user registration, information modification, account deletion and permission adjustment; The data management unit is used to add, modify and review data, complete data updates, and then program-mark and manually proofread the updated data to complete the quality control of the newly added data; The log management unit is used to perform statistics and analysis on the standard data set.

5. A language resource library system based on data security control according to claim 4, characterized in that: The management module also includes: An access control unit, used to design a role-based access control algorithm according to the data type of the standard data set, perform role management and authority management, and complete the management and control of user access rights; wherein the data type includes a vocabulary, a sentence list, a conversation, a translation text, an audio file, and a language learning material; The network security monitoring unit is used to build and train a security detection model using anomaly detection algorithm and isolation forest algorithm, and then use the security detection model to monitor network traffic in real time and detect abnormal behavior data.

6. A language resource library system based on data security control according to claim 5, characterized in that: The access control algorithm includes: The User Component sub-unit includes students for accessing language learning materials, teachers for accessing and modifying language learning materials and assessing the learning of said students, translators for accessing and modifying translated texts, and administrators for managing all users and resources; The role assignment component subunit includes a learner role for assigning permissions to the student, an educator role for assigning permissions to the teacher, a translator role for assigning permissions to the translator, and an administrator role for assigning permissions to the administrator; Permission subunits, including read permissions for accessing specific resources, write permissions for modifying specific resources, create permissions for adding new resources, and delete permissions for deleting specific resources; Resources subunit, including language learning materials, translated texts, and user information The logic mapping subunit is used to map the user to one or more roles and map the role to one or more permissions by using the mapping function, and then use the permission checking function to check whether the user and the role have the corresponding permissions.

7. A language resource library system based on data security control according to claim 6, characterized in that: The expression of the anomaly detection algorithm is: Among them, S is the safety score, w i is the weight, f i (x) is the detection index function, and n is the number of detection indicators.

8. A language resource library system based on data security control according to claim 7, characterized in that: The safety detection model includes: A feature extraction subunit is used to extract key features from network traffic; the key features include packet size, packet interval time, source IP address, destination IP address and packet protocol type; A data preprocessing subunit, used to standardize and normalize the key features A model training subunit is used to select an anomaly detection model, and train the selected anomaly detection model using an isolation forest algorithm and the standard data set to obtain a safety detection model; an anomaly detection subunit, used to obtain real-time network traffic data, extract traffic feature vectors from the real-time network traffic data, standardize and normalize the traffic feature vectors to obtain standard feature vectors, input the standard feature vectors into an isolation forest model, and calculate using an anomaly score algorithm to obtain an anomaly score; The abnormality judgment subunit is used to preset an abnormality threshold and judge whether the abnormality score is greater than the abnormality threshold. If so, it is judged that abnormal behavior exists.

9. A language resource library system based on data security control according to claim 8, characterized in that: The expression of the anomaly score algorithm is: H(n)=ln(n)+γ Among them, E(h(x)) is the average path length of X in each isolated tree, c(n) is the adjustment factor, H(n) is the harmonic number, and γ is the Euler constant.

10. A language resource library system based on data security control according to claim 9, characterized in that: The presentation module comprises: A portal information presentation unit is used to construct an information query system, a portal website and a historical electronic library; wherein the information query system includes accurate and fuzzy search functions by word, search functions by additional components, and search functions by paragraph and by article; The intelligent report presentation unit is used to design monitoring report content and monitoring database, and uses data mining and programming to build a resource library statistical analysis system for feedback statistical report information.

Images