Database digital watermark adaptive selection method, system, device and storage medium

Abstract

The present invention discloses a method, system, device, and storage medium for adaptively selecting digital watermarks for a database. The method comprises: S1, importing watermark algorithms, detection, and extraction algorithms supported by a database digital watermark system into an algorithm library; S2, reading the attributes of each tuple in a database table to be watermarked, and constructing a mapping relationship between the attribute, data usage, and watermark algorithm; S3, reading the attribute characteristics of the database table, and based on the data usage, priority, or usage ratio of different usages input by the user, after reading the watermark algorithm mapping relationship in the mapping relationship library, constructing a training set and forming a decision tree. The decision tree is used to determine the type of watermark embedding algorithm that should be selected for the database table in the current business scenario, and the algorithm is output as the algorithm to be embedded; S4, based on the algorithm to be embedded, calling the corresponding watermark algorithm in the algorithm library to embed the watermark in the database table. The embedding algorithm is automatically selected based on the database table characteristics and business requirements, thereby improving ease of use and configuration efficiency.

Description

Technical Field

[0001] The present invention relates to the field of data security, and in particular to a method, system, device and storage medium for adaptively selecting a digital watermark for a database. Background Art

[0002] Database digital watermarking technology verifies database ownership by adding secret information to the database without affecting normal user access. This technology plays a key role in determining database ownership, data distribution, tracing data leaks, and determining whether data has been tampered with during data transfer. It is currently a key technology for database data ownership verification and traceability.

[0003] Database digital watermarking technology includes three processes: digital watermark embedding, detection and watermark extraction. There are many current database digital watermark embedding algorithms, such as pseudo-row algorithm, pseudo-column algorithm, and some distortion algorithms that modify data.

[0004] However, each algorithm has a certain impact on the application of database tables. For example, pseudo-row watermarks are generally not used as query targets, while other row data can be queried normally. However, this may affect data statistics. Pseudo-column watermarks have no impact on data statistics, but may affect accurate data queries. Distortion watermarks have no impact on the table structure. Inserting special characters in specific cells or modifying certain data can affect precise queries on the watermarked column, but not fuzzy queries.

[0005] In practical applications, it is necessary to select the appropriate database digital watermarking algorithm based on the database's business scenario or attributes. Currently, technical personnel manually configure the database digital watermark embedding and extraction algorithms for a database table. This is cumbersome in some complex scenarios, and operations and maintenance personnel or software installers are sometimes unclear about the most appropriate watermark embedding algorithm. Summary of the Invention

[0006] The purpose of the present invention is to provide a method, system, device and storage medium for adaptive selection of database digital watermarks, which automatically selects the digital watermark embedding algorithm through automated database table feature recognition and business demand judgment, greatly reducing the workload of manual configuration, improving the usability of the database digital watermark system, and improving the efficiency and accuracy of the use of database digital watermarks.

[0007] To solve the above technical problems, an embodiment of the present invention provides a method for adaptively selecting a digital watermark from a database, comprising:

[0008] S1, import the watermark algorithm, detection and extraction algorithm supported by the database digital watermark system into the algorithm library;

[0009] S2, read the attributes of each tuple in the database table to be embedded with watermark, and build a mapping relationship between attributes, data usage and watermark algorithm;

[0010] S3, reading the attribute characteristics of the database table, and based on the data usage, usage priority, or usage ratio of different usages input by the user, after reading the watermark algorithm mapping relationship in the mapping relationship library, constructing a training set and forming a decision tree, using the decision tree to determine the watermark embedding algorithm type that should be selected for the database table in the current business scenario, and outputting it as the algorithm to be embedded;

[0011] S4, according to the algorithm to be embedded, calling the corresponding watermark algorithm in the algorithm library to embed the watermark in the database table.

[0012] Wherein, the S2 includes:

[0013] According to historical data, the attribute characteristics, usage scenarios and applicable watermark algorithms of the database table are mapped to form an initial mapping relationship library;

[0014] Performing feature learning on the attribute features of the database table, and outputting features of a predetermined type of the database table;

[0015] After collecting data on the usage of the database table, based on the attribute characteristics and usage scenarios of the database table, the database digital watermark algorithm that can be used for the database is inferred by studying the initial mapping relationship library to improve the initial mapping relationship library.

[0016] Wherein, after S2, the method further includes:

[0017] After obtaining the database attributes for which no mapping relationship is established in the database table, the characteristics of the database data and data usage corresponding to the database attributes are learned, and combined with the historical preference of the watermark algorithm, a training set of data attributes, usage and watermark algorithm is constructed to form an inferred mapping relationship.

[0018] Wherein, after S2, the method further includes:

[0019] Manual correction is performed on the mapping relationship.

[0020] The watermark algorithm includes a pseudo-row algorithm, a pseudo-column algorithm, a distortion algorithm based on valid bits, a distortion algorithm based on invisible characters, a simulation watermark algorithm, and a document database watermark algorithm.

[0021] In addition, the embodiment of the present application also provides a database digital watermark adaptive selection system, including:

[0022] The watermark algorithm library module is used to import the watermark algorithm, detection and extraction algorithm supported by the database digital watermark system into the algorithm library;

[0023] The mapping relationship library module is used to read the attributes of each tuple in the database table to be embedded with the watermark, and to build a mapping relationship between the attribute, data usage and watermark algorithm;

[0024] A decision-making module is used to read the attribute characteristics of the database table and, based on the data usage, usage priority or usage ratio of different usages input by the user, after reading the watermark algorithm mapping relationship in the mapping relationship library, construct a training set to form a decision tree, use the decision tree to determine the type of watermark embedding algorithm that should be selected for the database table in the current business scenario, and output it as the algorithm to be embedded;

[0025] The watermark embedding and extraction module is used to call the corresponding watermark algorithm in the algorithm library according to the algorithm to be embedded to embed the watermark in the database table.

[0026] The mapping relationship library module includes an initial mapping relationship library unit, a feature learning unit, and a mapping relationship improvement unit;

[0027] The initial mapping relationship library unit is used to make corresponding mapping relationships between the attribute characteristics, usage scenarios and applicable watermark algorithms of the database table according to historical data to form an initial mapping relationship library;

[0028] The feature learning unit is used to perform feature learning on the attribute features of the database table and output features of a predetermined type of the database table;

[0029] After collecting data on the usage of the database table, the mapping relationship improvement unit infers the database digital watermark algorithm that can be used for the database by learning the initial mapping relationship library based on the attribute characteristics and usage scenarios of the database table, and uses it to improve the initial mapping relationship library.

[0030] Among them, it also includes a mapping relationship inference module connected to the mapping relationship library module, which is used to learn the characteristics of database data and data usage corresponding to the database attributes after obtaining the database attributes for which no mapping relationship is established in the database table, and to construct a training set of data attributes, usage and watermark algorithm in combination with the historical preference of the watermark algorithm to form an inferred mapping relationship.

[0031] In addition, the embodiment of the present application also provides a device for a database digital watermark adaptive selection system, including:

[0032] Memory for storing computer programs;

[0033] A processor is used to execute the computer program to implement the steps of any one of the above methods for adaptively selecting a database digital watermark.

[0034] In addition, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. The computer program is executed by a processor to implement the steps of the database digital watermark adaptive selection method as described in any one of the above items.

[0035] The method, system, device, and storage medium for adaptively selecting digital watermarks from a database provided by the embodiments of the present invention have the following advantages over the prior art:

[0036] The disclosed database digital watermark adaptive selection method, system, device, and storage medium first imports the watermark algorithms, detection, and extraction algorithms supported by the database digital watermark system into an algorithm library for subsequent retrieval. The system then reads the attributes of each tuple in the database table and constructs a mapping relationship between the attribute, data usage, and watermark algorithm. The system then reads the attribute characteristics of the database table and, based on the user-entered data usage, usage priority, or usage ratio of different usages, reads the watermark algorithm mapping relationship in the mapping relationship library, constructs a training set, and forms a decision tree. The decision tree is used to determine the watermark embedding algorithm type that should be selected for the database table in the current business scenario and outputs it as the algorithm to be embedded. Finally, based on the algorithm to be embedded, the corresponding watermark algorithm in the algorithm library is called to embed the watermark in the database table. Throughout the entire process, the digital watermark embedding algorithm is automatically selected through automated database table feature recognition and business requirement determination, significantly reducing the workload of manual configuration, improving the usability of the database digital watermark system, and enhancing the efficiency and accuracy of database digital watermarking. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0038] Figure 1 A schematic diagram of a specific implementation of the method for adaptively selecting a digital watermark from a database provided in an embodiment of the present application;

[0039] Figure 2 A schematic diagram of the overall structural components of an embodiment of the method for adaptively selecting a digital watermark from a database provided in an embodiment of the present application;

[0040] Figure 3This is a structural diagram of a specific implementation of the database digital watermark adaptive selection method system provided in the embodiment of the present application.

[0041] Figure 4 This is a schematic diagram of a decision tree judgment in an embodiment of the database digital watermark adaptive selection method system provided by the embodiment of the invention of this application. DETAILED DESCRIPTION

[0042] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0043] Please refer to Figures 1 to 4 , Figure 1 A schematic diagram of a specific implementation of the method for adaptively selecting a digital watermark from a database provided in an embodiment of the present application; Figure 2 A schematic diagram of the overall structural components of an embodiment of the method for adaptively selecting a digital watermark from a database provided by an embodiment of the present invention; Figure 3 This is a structural diagram of a specific implementation of the database digital watermark adaptive selection method system provided in the embodiment of the present application. Figure 4 A schematic diagram of a decision tree judgment in an embodiment of the database digital watermark adaptive selection method system provided in an embodiment of the present application.

[0044] In a specific embodiment, the method for adaptively selecting a digital watermark from a database includes:

[0045] S1, import the watermark algorithm, detection and extraction algorithm supported by the database digital watermark system into the algorithm library; after adding all the watermark algorithms supported by the database digital watermark system into the algorithm library, it will be convenient for subsequent retrieval and provide a range of choices.

[0046] S2, reads the attributes of each tuple in the database table to be embedded with the watermark, and builds a mapping relationship library of attributes, data usage and watermark algorithm; constructs the mapping relationship based on the attribute reading to form a mapping relationship library, which is convenient for subsequent decision-making based on user needs and can improve decision-making efficiency and accuracy.

[0047] This application does not specifically limit the mapping relationship. In one embodiment, for example, user names, full names, ID numbers, and mobile phone numbers are generally used for query services, and pseudo-row algorithms are generally more suitable for queries, so a mapping relationship between user names, queries, and pseudo-row algorithms is constructed. Data attributes such as amounts and grades are often used for statistics or queries, and pseudo-column algorithms are generally used for statistics, so a mapping relationship between amounts, statistics, and pseudo-column algorithms is constructed.

[0048] It should be pointed out that in this application, the attribute characteristics of statistics and queries may be possessed at the same time, and both statistics and query services may be required. In this case, either a watermark algorithm with both functions may be adopted, or the user may be asked to provide whether more query scenarios or more statistical scenarios are required, and a priority list may be provided to solve the problem of selecting the watermark algorithm. For example, for precise statistics, a pseudo-column algorithm is suitable, while for rough statistics, pseudo-rows and pseudo-columns may be used, and such a priority list may be provided.

[0049] S3, reads the attribute characteristics of the database table, and according to the data usage, usage priority or usage ratio of different usages input by the user, after reading the watermark algorithm mapping relationship in the mapping relationship library, constructs a training set and forms a decision tree, uses the decision tree to determine the watermark embedding algorithm type that should be selected for the database table in the current business scenario, and outputs it as the algorithm to be embedded; in this application, after reading the attribute characteristics of the database table, a training set is constructed to form a decision tree based on the obtained user needs, and then the final required algorithm type is determined and output for the final embedding needs.

[0050] S4. Based on the algorithm to be embedded, the corresponding watermark algorithm in the algorithm library is called to embed the watermark in the database table. This process can use a matching method, where the algorithm obtained in the previous step is matched with the possible algorithm in this step, and then the watermark is embedded based on the matching result. Alternatively, after obtaining the algorithm to be embedded, the algorithm is called from the algorithm library and then embedded in the database table, or other watermark embedding methods are used, which are not limited in this application.

[0051] The database digital watermarking system first imports the watermark algorithms, detection, and extraction algorithms supported by the system into an algorithm library for subsequent access. The system then reads the attributes of each tuple in the database table and constructs a mapping relationship between the attribute, data usage, and watermark algorithm. The system then reads the database table's attribute characteristics and, based on the user-entered data usage, usage priority, or usage ratio of different usages, determines user needs. After reading the watermark algorithm mapping relationships in the mapping library, a training set is constructed to form a decision tree. This decision tree is used to determine the watermark embedding algorithm type that should be selected for the database table in the current business scenario and outputs it as the algorithm to be embedded. Finally, based on the algorithm to be embedded, the corresponding watermark algorithm in the algorithm library is called to embed the watermark in the database table. Throughout this process, the digital watermark embedding algorithm is automatically selected through automated database table feature recognition and business requirement determination, significantly reducing the workload of manual configuration, improving the usability of the database digital watermarking system, and enhancing the efficiency and accuracy of database digital watermarking.

[0052] This application does not limit the formation of a specific mapping relationship library. In one embodiment, S2 includes:

[0053] According to historical data, the attribute characteristics, usage scenarios and applicable watermark algorithms of the database table are mapped to form an initial mapping relationship library;

[0054] Performing feature learning on the attribute features of the database table, and outputting features of a predetermined type of the database table;

[0055] After collecting data on the usage of the database table, based on the attribute characteristics and usage scenarios of the database table, the database digital watermark algorithm that can be used for the database is inferred by studying the initial mapping relationship library to improve the initial mapping relationship library.

[0056] By first forming an initial mapping relationship library, then performing feature learning to obtain the feature type of the database table, and finally collecting data on its use, an algorithm can be selected for learning and inference of the initial mapping relationship library to achieve the improvement of the initial mapping relationship library. After multiple improvements, the relationship library can be optimized, and the optimization can be performed regularly, or before each watermark embedding, or other optimization methods can be used.

[0057] In actual applications, there may be a situation where no mapping relationship is established. In order to facilitate the formation of a subsequent decision tree and decision-making, in one embodiment, after S2, the following steps are further included:

[0058] After obtaining the database attributes for which no mapping relationship is established in the database table, the characteristics of the database data and data usage corresponding to the database attributes are learned, and combined with the historical preference of the watermark algorithm, a training set of data attributes, usage and watermark algorithm is constructed to form an inferred mapping relationship.

[0059] When the mapping relationship cannot be obtained directly, the characteristics of database data and data usage corresponding to the database attributes are used for learning, and a training set of data attributes, usage and watermark algorithm is constructed to achieve inferred mapping relationship and obtain mapping relationship for all database tables.

[0060] Due to this inferred mapping relationship, errors may occur. To improve the accuracy, in one embodiment, after S2, the following steps are further included:

[0061] Manual correction is performed on the mapping relationship.

[0062] For this inferred mapping relationship, manual correction and modification are performed regularly to continuously improve the mapping relationship library and enhance the accuracy of subsequent decisions.

[0063] This application does not specifically limit the type of watermark algorithm. The watermark algorithm includes a pseudo-row algorithm, a pseudo-column algorithm, a distortion algorithm based on valid bits, a distortion algorithm based on invisible characters, a simulation watermark algorithm, a document database watermark algorithm, and may also include other algorithms, which are not limited in this application.

[0064] The overall implementation of the present invention includes a database digital watermark system server, a user terminal, and an original database server. The specific architecture is as follows: Figure 2 The business process is as follows.

[0065] In the device structure and business process of one embodiment of the present application, the specific business process of a user requesting to embed a database digital watermark is as follows:

[0066] ① The user sends a request to the database digital watermark server through the terminal device, requesting to embed a watermark in a table of the original database, and stating the usage scenario of the database table. If there are multiple usage scenarios, the priority of the usage is stated.

[0067] ②After receiving the request, the database digital watermark server sends a request to the source database to read the database table;

[0068] ③The source database feeds back the requested database table;

[0069] ④The database digital watermark server conducts feature learning on the database table fed back, obtains attribute information and user scenario usage information, and uses the decision tree algorithm to make decisions based on the mapping relationship between attributes, usage and watermark algorithm in the relationship mapping library. After outputting the watermark algorithm to be used, it calls the corresponding algorithm in the watermark algorithm library to embed the watermark operation on the database table and marks the embedded algorithm on the database table.

[0070] ⑤The database digital watermark server will feed back the database table with the watermark embedded into it to the user terminal device.

[0071] In addition, the embodiment of the present application also provides a database digital watermark adaptive selection system, including:

[0072] The watermark algorithm library module 10 is used to import the watermark algorithm, detection and extraction algorithm supported by the database digital watermark system into the algorithm library;

[0073] The mapping relationship library module 20 is used to read the attributes of each tuple in the database table to be embedded with the watermark, and to construct a mapping relationship between the attribute, the data usage and the watermark algorithm;

[0074] The decision-making module 30 is used to read the attribute characteristics of the database table and, based on the data usage, usage priority, or usage ratio of different usages input by the user, after reading the watermark algorithm mapping relationship in the mapping relationship library, construct a training set to form a decision tree, use the decision tree to determine the type of watermark embedding algorithm that should be selected for the database table in the current business scenario, and output it as the algorithm to be embedded;

[0075] The watermark embedding and extraction module 40 is used to embed the watermark in the database table by calling the corresponding watermark algorithm in the algorithm library based on the algorithm to be embedded. The watermark embedding and extraction module is an auxiliary module of this application. It determines the selected watermark algorithm based on the decision tree, calls the relevant algorithm in the algorithm library, and performs the watermark embedding or extraction operation, completing the task of loading or extracting the watermark into the database table.

[0076] Since the database digital watermark adaptive selection system is a system corresponding to the database digital watermark adaptive selection method and has the same beneficial effects, this application will not elaborate on it.

[0077] This application does not limit the specific structure of the mapping relationship library and the formation of the mapping relationship. Its main purpose is to form a subsequent decision tree and achieve efficient automatic decision-making. In one embodiment, the mapping relationship library module includes an initial mapping relationship library unit, a feature learning unit, and a mapping relationship improvement unit;

[0078] The initial mapping relationship library unit is used to make corresponding mapping relationships between the attribute characteristics, usage scenarios and applicable watermark algorithms of the database table according to historical data to form an initial mapping relationship library;

[0079] The feature learning unit is used to perform feature learning on the attribute features of the database table and output features of a predetermined type of the database table;

[0080] After collecting data on the usage of the database table, the mapping relationship improvement unit infers the database digital watermark algorithm that can be used for the database by learning the initial mapping relationship library based on the attribute characteristics and usage scenarios of the database table, and uses it to improve the initial mapping relationship library.

[0081] The mapping relationship library module is one of the core modules of this application. It mainly forms the mapping relationship between the three based on the module developers' existing database development experience, the general usage scenarios of different databases and the corresponding more suitable watermark algorithms.

[0082] However, in some cases, a mapping relationship may not be formed directly. In one embodiment, the database digital watermark adaptive selection system also includes a mapping relationship inference module connected to the mapping relationship library module, which is used to learn the characteristics of database data and data usage corresponding to the database attributes after obtaining the database attributes for which no mapping relationship has been established in the database table, and to construct a training set of data attributes, usage and watermark algorithm in combination with the historical preference of the watermark algorithm to form an inferred mapping relationship.

[0083] By learning the attribute characteristics of database data, the characteristics of data usage, and the preferences of watermark algorithms used in the past, feature training, learning and inference are carried out, and some mapping relationships between attributes, usage and watermark algorithms are automatically formed. If necessary, these mapping relationships can be manually corrected.

[0084] Then, based on the table attribute characteristics and table usage scenarios, by learning the configured mapping relationships, we infer the database digital watermarking algorithm that can be used for the database, and continuously improve the mapping relationship library. The implementation of this module is automatically generated when the database digital watermarking system is deployed in the actual environment.

[0085] The following Table 1 is an example of a mapping relationship table.

[0086]

[0087] Table 1

[0088] The decision-making module in this application is the core module of this module and is the key step in selecting the adaptive watermark algorithm. When the user sends the watermark embedding instruction (including database table information, database table usage scenario information, usage scenario priority information, special requirements, etc.) to the database digital watermark system, the system uses this module to make the decision on which watermark algorithm to select. The specific implementation method of the module is as follows:

[0089] This module reads and intelligently learns the database table feature information fed back, collects table usage scenarios, priorities, special requirements and other information in user instruction information, and reads the relevant mapping relationships in the mapping relationship library based on this information, constructs a training set based on table feature information, table usage scenario information, usage priority information, special requirements, and mapping relationships, and forms a decision tree. Based on the decision tree, a decision is made on the selection of the watermark algorithm. The specific example is shown in the figure below. Figure 3 shown.

[0090] In addition, the embodiment of the present application also provides a device for a database digital watermark adaptive selection system, including:

[0091] Memory for storing computer programs;

[0092] A processor is used to execute the computer program to implement the steps of any one of the above methods for adaptively selecting a database digital watermark.

[0093] This application does not limit the type of device for the database digital watermark adaptive selection system.

[0094] In addition, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. The computer program is executed by a processor to implement the steps of the database digital watermark adaptive selection method as described in any one of the above items.

[0095] Similarly, since the computer program in the computer-readable storage medium is executed by the processor to implement the steps of any of the above methods for adaptively selecting database digital watermarks, the same beneficial effects are achieved, and this application does not limit this.

[0096] In summary, the database digital watermark adaptive selection method, system, device, and storage medium provided by the embodiments of the present invention first imports the watermark algorithm, detection, and extraction algorithm supported by the database digital watermark system into an algorithm library for subsequent retrieval, then reads the attributes of each tuple in the database table and constructs a mapping relationship between the attribute, data usage, and watermark algorithm. Subsequently, the attribute characteristics of the database table are read, and the user's needs are obtained based on the data usage, usage priority, or usage ratio of different usages entered by the user. After reading the watermark algorithm mapping relationship in the mapping relationship library, a training set is constructed to form a decision tree. The decision tree is used to determine the type of watermark embedding algorithm to be selected for the database table in the current business scenario and output as the algorithm to be embedded. Finally, based on the algorithm to be embedded, the corresponding watermark algorithm in the algorithm library is called to embed the watermark in the database table. Throughout the entire process, the digital watermark embedding algorithm is automatically selected through automated database table feature recognition and business requirement judgment, greatly reducing the workload of manual configuration, improving the ease of use of the database digital watermark system, and improving the efficiency and accuracy of database digital watermarking.

[0097] The above describes in detail the method, system, device, and storage medium for adaptively selecting digital watermarks for databases provided by the present invention. This article uses specific examples to illustrate the principles and implementation methods of the present invention. The description of the above examples is intended only to facilitate understanding of the method and core concepts of the present invention. It should be noted that those skilled in the art will be able to make various improvements and modifications to the present invention without departing from the principles of the present invention, and such improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. A database digital watermark adaptive selection method, characterized by: include: S1, import the watermark algorithm, detection and extraction algorithm supported by the database digital watermark system into the algorithm library; S2, read the attributes of each tuple in the database table to be embedded with watermark, and build a mapping relationship between attributes, data usage and watermark algorithm; S3, reading the attribute characteristics of the database table, and based on the data usage, usage priority, or usage ratio of different usages input by the user, after reading the watermark algorithm mapping relationship in the mapping relationship library, constructing a training set and forming a decision tree, using the decision tree to determine the watermark embedding algorithm type that should be selected for the database table in the current business scenario, and outputting it as the algorithm to be embedded; S4, based on the algorithm to be embedded, calling the corresponding watermark algorithm in the algorithm library to embed the watermark in the database table; The S2 includes: According to historical data, the attribute characteristics, usage scenarios and applicable watermark algorithms of the database table are mapped to form an initial mapping relationship library; Performing feature learning on the attribute features of the database table, and outputting features of a predetermined type of the database table; After collecting data on the usage of the database table, based on the attribute characteristics and usage scenarios of the database table, by learning the initial mapping relationship library, inferring a database digital watermark algorithm that can be used for the database, and using it to improve the initial mapping relationship library; After S2, the method further includes: After obtaining database attributes for which no mapping relationship is established in the database table, characteristics of database data and data usage corresponding to the database attributes are learned, and a training set of data attributes, usage, and watermarking algorithms is constructed in combination with historical preferences of the watermarking algorithm to form an inferred mapping relationship; after S2, the following further includes: Manual correction is performed on the mapping relationship; the watermark algorithm includes a pseudo-row algorithm, a pseudo-column algorithm, a distortion algorithm based on valid bits, a distortion algorithm based on invisible characters, a simulation watermark algorithm, and a document database watermark algorithm.

2. A database digital watermark adaptive selection system, characterized by: include: The watermark algorithm library module is used to import the watermark algorithm, detection and extraction algorithm supported by the database digital watermark system into the algorithm library; The mapping relationship library module is used to read the attributes of each tuple in the database table to be embedded with the watermark, and to build a mapping relationship between the attribute, data usage and watermark algorithm; A decision-making module is used to read the attribute characteristics of the database table and, based on the data usage, usage priority or usage ratio of different usages input by the user, after reading the watermark algorithm mapping relationship in the mapping relationship library, construct a training set and form a decision tree. The decision tree is used to determine the type of watermark embedding algorithm that should be selected for the database table in the current business scenario, and the algorithm is output as the algorithm to be embedded; A watermark embedding and extraction module is used to call the corresponding watermark algorithm in the algorithm library to embed the watermark in the database table according to the algorithm to be embedded. The mapping relationship library module includes an initial mapping relationship library unit, a feature learning unit, and a mapping relationship improvement unit; The initial mapping relationship library unit is used to make corresponding mapping relationships between the attribute characteristics, usage scenarios and applicable watermark algorithms of the database table according to historical data to form an initial mapping relationship library; The feature learning unit is used to perform feature learning on the attribute features of the database table and output features of a predetermined type of the database table; The mapping relationship improvement unit, after collecting data on the usage of the database table, infers the database digital watermark algorithm that can be selected for the database by learning the initial mapping relationship library based on the attribute characteristics and usage scenarios of the database table, and is used to improve the initial mapping relationship library. It also includes a mapping relationship inference module connected to the mapping relationship library module, which is used to learn the characteristics of the database data and data usage corresponding to the database attributes after obtaining the database attributes for which no mapping relationship is established in the database table, and construct a training set of data attributes, usage and watermark algorithm in combination with the historical preference of using the watermark algorithm to form an inferred mapping relationship.

3. A device for a database digital watermark adaptive selection system, characterized in that: include: Memory for storing computer programs; A processor is used to execute the computer program to implement the steps of the database digital watermark adaptive selection method as claimed in claim 1.

4. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which is executed by a processor to implement the steps of the database digital watermark adaptive selection method according to claim 1.

Images