A method and system for establishing a book library based on multi-version annotations

Abstract

The present invention relates to the technical field of database establishment, and in particular to a method and system for establishing a book library based on multi-version annotations, comprising the following steps: data collection; data preprocessing; using a text comparison algorithm to compare and analyze the contents of books of different versions, and generating a version difference report; preliminary annotation association; tracing and associating annotations between versions, providing a multi-version annotation comparison function, and a user can simultaneously view annotations of the same position in different versions, which is convenient for comparison and analysis; data storage; designing a user interface, and providing book version selection, content browsing and annotation viewing functions; intelligent annotation recommendation and optimization; authority control; and data update. The present invention, through the intelligent annotation recommendation and optimization functions, allows users to efficiently manage annotation information, avoids duplication of work, and improves the utilization and management efficiency of annotations.

Description

Technical Field

[0001] The present invention relates to the technical field of database establishment, and in particular to a method and system for establishing a book library based on multi-version annotations. Background Art

[0002] At present, there is no electronic book library system on the market that can effectively manage multi-version books and provide comprehensive annotation functions. Most of the existing technologies focus on the storage and display of book content, while ignoring the importance of version differences and annotation management. Specifically, the existing technologies have the following major problems:

[0003] Lack of multi-version management: Existing electronic book libraries usually only provide a single version of the book content, and cannot effectively manage and display the difference information of multiple versions of books. When users check and compare the content of different versions of books, they often need to manually perform tedious comparison work, which is inefficient and prone to errors.

[0004] Limitations of annotation function: The annotation function of existing electronic book libraries is relatively simple, usually only providing basic annotation adding and viewing functions, lacking intelligent annotation recommendation and redundant annotation optimization functions. When adding and managing annotations, users need to manually handle a large number of redundant and repeated annotations, which is labor-intensive and inefficient.

[0005] Insufficient annotation traceability and version association: In the existing technology, the traceability and version association functions of annotations are not perfect. When users switch between different versions, it is difficult to accurately understand the history and evolution of annotations. This is particularly inconvenient in academic research and teaching, as it is difficult for users to track the source and changes of annotations.

[0006] Therefore, designing a method for establishing a book library based on multi-version annotations can effectively solve the above problems. Summary of the invention

[0007] Based on the above objectives, the present invention provides a method and system for establishing a book library based on multi-version annotations.

[0008] A method for establishing a book library based on multi-version annotations includes the following steps:

[0009] S1, data collection: obtaining book data of different versions, wherein the book data includes book content, version information and author notes;

[0010] S2, data preprocessing: preprocessing the book data, including format conversion, deduplication and segmentation processing;

[0011] S3, multi-version comparison: using a text comparison algorithm to compare and analyze the contents of different versions of books and generate a version difference report, which includes new content, deleted content, and modified content;

[0012] S4, preliminary annotation association: according to the version difference report, the existing annotation information is preliminarily associated with the corresponding book content and displayed by version;

[0013] S5, annotation traceability and association between versions: In the data preprocessing stage, a unique traceability identifier is generated for each annotation, recording the initial version and modification history of the annotation; the annotation traceability function is provided, and users click on the annotation to view its history and changes in different versions; the annotation version tree is displayed through a graphical interface to intuitively display the evolution process of the annotation and the association between different versions; a multi-version annotation comparison function is provided, and users can simultaneously view the annotations of the same location in different versions for easy comparison and analysis;

[0014] S6, data storage: storing the processed book data and its annotation information in a database, and establishing a version management and annotation management unit;

[0015] S7, User Interface Design: Design the user interface to provide book version selection, content browsing and annotation viewing functions, so that users can switch between different versions and view the corresponding annotations;

[0016] S8, Intelligent Annotation Recommendation and Optimization: Utilizes natural language processing technology and machine learning models to analyze book content and existing annotations, and recommend relevant annotation content to users; automatically analyzes the redundancy and repetition of annotations, and prompts users to merge or delete redundant annotations;

[0017] S9, permission control: set user permissions and control user access rights to different versions of book content and annotation information;

[0018] S10, Data Update: Update book data and annotation information regularly or irregularly to ensure the timeliness and accuracy of the library content.

[0019] Furthermore, the S1 specifically includes:

[0020] S11, obtaining authorized book version data from an authorized data source, wherein the authorized data source includes a publisher database, an academic resource platform, and an e-book supplier;

[0021] S12, using optical character recognition (OCR) technology to scan paper books and convert them into electronic data, identify and extract book content, version information and author notes;

[0022] S13, the user uploads book data. The user manually uploads the electronic book file through the system interface and supplements the version information and author notes.

[0023] Furthermore, the multi-version comparison in S3 specifically includes:

[0024] S31, normalize the text of different versions of books, including removing spaces, special symbols and formatting text to ensure the accuracy of comparative analysis;

[0025] S32, using the longest common subsequence (LCS) algorithm to compare and analyze the contents of different versions of books paragraph by paragraph, and determine the similarities and differences of the texts by identifying the longest common subsequence in the texts of two versions;

[0026] S33, identifying new content, deleted content, and modified content in different versions of the book content by using the longest common subsequence algorithm, and generating a version difference report, wherein the report records the content changes between each version;

[0027] S34, mark the version differences in the book content, highlight the new content, indicate the deleted content with a strikethrough, and mark the modified content with different colors to distinguish the content before and after the modification.

[0028] Furthermore, the longest common subsequence algorithm determines the similarity and difference of the text by identifying the longest common subsequence in the two versions of the text, specifically including:

[0029] Define two strings and ;

[0030] Create a two-dimensional array ,in Represents a string Before Characters and strings Before The length of the longest common subsequence of characters;

[0031] Construct a recursive relationship:

[0032] = ,in, Indicates conditional judgment, that is, if the condition is met, the corresponding operation is performed, or means or, that is, when one of the conditions i = 0 or j = 0 is met, the corresponding operation can be performed;

[0033] at last, It's a string and The length of the longest common subsequence of ;

[0034] according to The array is backtracked to find the longest common subsequence and mark the added, deleted and modified content;

[0035] Difference identification: Use the longest common subsequence algorithm to identify new content, deleted content, and modified content in different versions of books, and generate a version difference report that records the content changes between each version in detail;

[0036] New content recognition: exists in The parts that do not exist in the are marked as new content;

[0037] Deleting Content Identification: exists in The parts that do not exist in the content are marked as deleted.

[0038] Modify content identification: and Parts that exist in both but have different contents are marked as modified content.

[0039] Furthermore, the preliminary annotation association in S4 specifically includes:

[0040] S41, version difference location: According to the generated version difference report, determine the location of the addition, deletion and modification of the book content, generate a location mapping table, and and version , define the location mapping table ,in Indicates version Middle Positions in version The corresponding position in the , generation method includes using the result of the longest common subsequence algorithm, traversing the content of the two versions of the book, and recording the corresponding position of each character or word, expressed as: ;

[0041] S42, annotation location: Use the location mapping table in the version difference report to locate the existing annotation information to the specific location of the book content, generate a preliminary annotation association table, and set the annotation table Contains the comment location and annotation content , generate a preliminary annotation association table :

[0042] , according to the location mapping table , the annotation table Each annotation position in Map to the location of the new version and generate a preliminary annotation association table ;

[0043] S43, annotation classification: according to the version information of the annotation, the annotation is classified and stored by version for subsequent display;

[0044] S44, annotation display: in the user interface, according to the book version selected by the user, the annotation information of the corresponding version is displayed, and the annotation information includes the version identifier and the annotation content;

[0045] S45, synchronous update of annotations: when the book content is updated, the position mapping of the annotations is automatically adjusted according to the new version difference report to ensure the relevance of the annotation information to the book content.

[0046] Furthermore, the annotation traceability and association between versions in S5 specifically include:

[0047] S51, traceability identification generation: In the data preprocessing stage, a unique traceability identification is generated for each comment. The traceability identification is generated by combining the hash value of the comment content and the comment position, and recording the initial version and modification history of the comment. Suppose the comment content is , the annotation position is , the initial version is , generate traceability identification , expressed as:

[0048] ;

[0049] S52, annotation traceability record: records the initial version and modification history of the annotation, including the timestamp and modification content of each modification. The format of each annotation record is ,in, is the timestamp, The modified annotation content;

[0050] S53, annotation traceability function: Provide annotation traceability function, users click on the annotation to view its history and changes in different versions, according to the traceability mark , retrieve the corresponding annotation history in the database, generate a history list and display it:

[0051] ;

[0052] S54, annotation version tree display: Based on the annotation traceability record, an annotation version tree is constructed, and a tree diagram is used to represent the evolution of annotations. Each node represents a version of the annotation, and the lines between nodes represent the inheritance relationship between versions.

[0053] S55, Multi-version annotation comparison: In the user interface, the multi-version switching and comparison functions are provided. After the user selects different versions, the annotation contents of multiple versions are displayed at the same time, and the annotations of different versions are distinguished by color and annotations: .

[0054] Furthermore, in S8, a pre-trained language model based on the Transformer architecture, namely the BERT model, is trained to understand the relationship between the book content and the annotations, specifically including:

[0055] Collecting datasets of book content and annotations ,in For the content of the book, For the corresponding annotations;

[0056] Use the pre-trained BERT model and fine-tune it on the dataset:

[0057] ;

[0058] For each pair of input , sentence embedding is obtained through the BERT model :

[0059] ;

[0060] Add a fully connected layer for classification to predict whether the annotation is relevant:

[0061] ;

[0062] Use the cross entropy loss function to train the model: ;

[0063] Based on the trained BERT model, recommend relevant annotations for new book content.

[0064] Furthermore, the method of recommending relevant annotations for new book content based on the trained BERT model specifically includes:

[0065] New book content and comments in the database Calculate similarity:

[0066] ;

[0067] ;

[0068] Calculate the similarity score:

[0069] ,in, For book content With annotations similarity;

[0070] Sort by similarity score and generate a list of recommended annotations :

[0071] ,in, is a collection of annotations, is the recommended threshold.

[0072] Furthermore, the redundancy and repetitiveness of the annotations are automatically analyzed, and the user is prompted to merge or delete redundant annotations based on a redundant similarity analysis algorithm, which is expressed as: ;

[0073] Redundancy prompt: When the similarity exceeds the set threshold , prompts the user to merge or delete redundant comments.

[0074] A system for establishing a book library based on multi-version annotations, used to implement the above-mentioned method for establishing a book library based on multi-version annotations, includes the following modules:

[0075] Data collection module: used to obtain book data of different versions, including book content, version information and author notes, and the collection methods include OCR technology and user upload;

[0076] Data preprocessing module: preprocessing the book data, including format conversion, deduplication and segmentation processing;

[0077] Version comparison module: Use the longest common subsequence algorithm to compare and analyze the contents of different versions of books and generate a version difference report;

[0078] Preliminary annotation association module: Based on the version difference report, the existing annotation information is preliminarily associated with the corresponding book content;

[0079] Inter-version annotation traceability module: Generates a unique traceability identifier for each annotation, records the initial version and modification history of the annotation, provides annotation traceability function, displays the annotation version tree, displays the evolution process of the annotation and the relationship between different versions, and provides annotation comparison function to view the annotations in the same location in different versions;

[0080] Data storage module: stores the processed book data and its annotation information in the database, and establishes version management and annotation management units;

[0081] Interaction module: Design the user interface to provide book version selection, content browsing and annotation viewing functions;

[0082] Intelligent annotation recommendation and optimization module: uses natural language processing technology and the BERT model to analyze book content and existing annotations, recommend relevant annotation content to users, and prompt users to merge or delete redundant annotations;

[0083] Permission control module: set user permissions and control user access rights to different versions of book content and annotation information;

[0084] Data update module: updates book data and annotation information.

[0085] Beneficial effects of the present invention:

[0086] The present invention introduces an intelligent annotation recommendation and optimization module, which analyzes the book content and existing annotations through the BERT model and recommends relevant annotation content to users. The trained BERT model can understand the relationship between the book content and the annotations, and generate a recommended annotation list through similarity calculation, thereby improving the accuracy of annotation recommendations. At the same time, redundant similarity analysis is used to detect the redundancy and repetition between annotations, prompting users to merge or delete redundant annotations, thereby optimizing the annotation management process. Through the intelligent annotation recommendation and optimization functions, users can efficiently manage annotation information, avoid duplication of work, and improve the utilization rate and management efficiency of annotations.

[0087] The present invention realizes comprehensive tracing of annotation information by generating a unique traceability identifier through the inter-version annotation traceability and association module, recording the initial version and modification history of the annotation. Users can view the historical records and changes of annotations in different versions by clicking on the annotations, and intuitively understand the evolution of the annotations and the association between versions through the annotation version tree displayed in the graphical interface. In addition, the multi-version annotation comparison function enables users to view annotations in different versions at the same location at the same time, which is convenient for comparison and analysis. These innovations improve the transparency and traceability of annotation management, enable users to accurately understand the source and changes of annotations, and improve the reliability and use value of annotation information.

[0088] The present invention generates a version difference report by comparing multiple versions, and then uses a preliminary annotation association module to accurately associate the existing annotation information with the corresponding book content according to the version difference report, and displays it by version. By using the position mapping table in the version difference report, the annotation can be automatically located and associated to ensure the consistency and accuracy of the annotation information and the book content. When viewing the content of different versions of books, users can clearly see the annotations of the corresponding version, avoiding the problem of confusion of annotation positions and information loss. In addition, when the book content is updated, the system automatically adjusts the annotation position to maintain the effective association of the annotation, further improving the convenience and practicality of annotation management. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0090] Figure 1 A schematic diagram of a method for establishing an embodiment of the present invention;

[0091] Figure 2 It is a schematic diagram of the functional modules of the establishment system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0092] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0093] like Figure 1 As shown, a method for establishing a book library based on multi-version annotations includes the following steps:

[0094] S1, data collection: obtaining book data of different versions, including book content, version information and author notes;

[0095] S2, data preprocessing: preprocess the book data, including format conversion, deduplication and segmentation processing, to unify the data format and improve the data quality. Specifically:

[0096] Convert book data from different sources into standard e-book formats, including PDF, EPUB and TXT. Use format conversion tools to standardize book content, version information and author notes. Use hash algorithms to deduplicate duplicate book data to ensure that the content of each version of the book is unique and generate a unique version traceability identifier. Use natural language processing technology to segment book content, including chapter recognition, paragraph division and sentence segmentation, to improve the structured level and data quality of book data.

[0097] S3, multi-version comparison: Use text comparison algorithms to compare and analyze the contents of different versions of books and generate version difference reports, which include new content, deleted content, and modified content;

[0098] S4, preliminary annotation association: according to the version difference report, the existing annotation information is preliminarily associated with the corresponding book content and displayed by version;

[0099] S5, annotation traceability and association between versions: In the data preprocessing stage, a unique traceability identifier is generated for each annotation, recording the initial version and modification history of the annotation; the annotation traceability function is provided, and users click on the annotation to view its history and changes in different versions; the annotation version tree is displayed through a graphical interface to intuitively display the evolution process of the annotation and the association between different versions; a multi-version annotation comparison function is provided, and users can simultaneously view the annotations of the same location in different versions for easy comparison and analysis;

[0100] S6, data storage: storing the processed book data and its annotation information in a database, and establishing a version management and annotation management unit;

[0101] S7, User Interface Design: Design the user interface to provide book version selection, content browsing and annotation viewing functions, so that users can switch between different versions and view the corresponding annotations;

[0102] S8, Intelligent Annotation Recommendation and Optimization: Utilizes natural language processing technology and machine learning models to analyze book content and existing annotations, and recommend relevant annotation content to users; automatically analyzes the redundancy and repetition of annotations, and prompts users to merge or delete redundant annotations;

[0103] S9, permission control: set user permissions and control user access rights to different versions of book content and annotation information;

[0104] S10, Data Update: Update book data and annotation information regularly or irregularly to ensure the timeliness and accuracy of the library content.

[0105] S1 specifically includes:

[0106] S11, obtaining authorized book version data from authorized data sources, where the authorized data sources include publisher databases, academic resource platforms, and e-book suppliers;

[0107] S12, using optical character recognition (OCR) technology to scan paper books and convert them into electronic data, identify and extract book content, version information and author notes;

[0108] S13, user uploads book data. The user manually uploads the electronic book file through the system interface and supplements the version information and author comments.

[0109] The multi-version comparison in S3 specifically includes:

[0110] S31, normalize the text of different versions of books, including removing spaces, special symbols and formatting text to ensure the accuracy of comparative analysis;

[0111] S32, using the longest common subsequence (LCS) algorithm to compare and analyze the contents of different versions of books paragraph by paragraph, and determine the similarities and differences of the texts by identifying the longest common subsequence in the texts of two versions;

[0112] S33, identifying new content, deleted content, and modified content in different versions of the book content through the longest common subsequence algorithm, and generating a version difference report, which records the content changes between each version;

[0113] S34, marking the version differences in the book content, highlighting the new content, deleting the content with a strikethrough, and marking the modified content with different colors to facilitate users to intuitively view the version changes;

[0114] The identified version difference information is displayed in the form of a graphical report. The report includes a summary of changes, detailed change content and statistical information. Users can quickly understand the differences between versions through the report, including;

[0115] Summary of changes: Count the number of new, deleted, and modified content;

[0116] Detailed changes: List the version differences in detail by chapter and paragraph;

[0117] Statistics: Generate charts of version differences, such as bar charts and pie charts, showing the proportion and number of various changes.

[0118] The longest common subsequence algorithm determines the similarities and differences between two versions of a text by identifying the longest common subsequence between the two versions of the text. Specifically, it includes:

[0119] Define two strings and ;

[0120] Create a two-dimensional array ,in Represents a string Before Characters and strings Before The length of the longest common subsequence of characters;

[0121] Construct a recursive relationship:

[0122] = ,in, Indicates conditional judgment, that is, if the condition is met, the corresponding operation is performed, or means or, that is, when one of the conditions i = 0 or j = 0 is met, the corresponding operation can be performed;

[0123] at last, It's a string and The length of the longest common subsequence of ;

[0124] according to The array is backtracked to find the longest common subsequence and mark the added, deleted and modified content;

[0125] Difference identification: Use the longest common subsequence algorithm to identify new, deleted, and modified content in different versions of books, and generate a version difference report that records the content changes between each version in detail;

[0126] New content recognition: exists in The parts that do not exist in the are marked as new content;

[0127] Deleting Content Identification: exists in The parts that do not exist in the content are marked as deleted.

[0128] Modify content identification: and Parts that exist in both but have different contents are marked as modified content.

[0129] The version differences are marked in the book content, new content is highlighted, deleted content is indicated by strikethrough, and modified content is marked with different colors before and after the modification, so that users can intuitively view the version changes.

[0130] The preliminary annotation associations in S4 specifically include:

[0131] S41, version difference location: According to the generated version difference report, determine the location of the addition, deletion and modification of the book content, generate a location mapping table, and and version , define the location mapping table ,in Indicates version Middle Positions in version The corresponding position in the , generation method includes using the result of the longest common subsequence algorithm, traversing the content of the two versions of the book, and recording the corresponding position of each character or word, expressed as: ;

[0132] S42, annotation location: Use the location mapping table in the version difference report to locate the existing annotation information to the specific location of the book content, generate a preliminary annotation association table, and set the annotation table Contains the comment location and annotation content , generate a preliminary annotation association table :

[0133] , according to the location mapping table , the annotation table Each annotation position in Map to the location of the new version and generate a preliminary annotation association table , if in version The annotation position is , then according to Find its version The corresponding position in ;

[0134] S43, annotation classification: according to the version information of the annotation, the annotation is classified and stored by version for subsequent display; storage structure: the annotation information is stored in the database, classified and stored by version information, represented as: annotation table structure = {version number, annotation position, annotation content};

[0135] S44, annotation display: In the user interface, according to the book version selected by the user, the annotation information of the corresponding version is displayed. The annotation information includes the version identifier and the annotation content. The display logic is: according to the version selected by the user, the annotation information of the corresponding version is extracted from the database and marked in the book content for display, which is expressed as: display function ;

[0136] S45, annotation synchronization update: when the book content is updated, the annotation position mapping is automatically adjusted according to the new version difference report to ensure the relevance of the annotation information to the book content;

[0137] The update method is: regenerate the location mapping table , and update the preliminary annotation association table according to the new mapping table : = .

[0138] The annotation traceability and association between versions in S5 specifically include:

[0139] S51, traceability identification generation: In the data preprocessing stage, a unique traceability identification is generated for each comment. The traceability identification is generated by combining the hash value of the comment content and the comment position, and recording the initial version and modification history of the comment. Suppose the comment content is , the annotation position is , the initial version is , generate traceability identification , expressed as:

[0140] ;

[0141] S52, annotation traceability record: records the initial version and modification history of the annotation, including the timestamp and modification content of each modification. The format of each annotation record is ,in, is the timestamp, The modified annotation content;

[0142] S53, annotation traceability function: Provide annotation traceability function, users click on the annotation to view its history and changes in different versions, according to the traceability mark , retrieve the corresponding annotation history in the database, generate a history list and display it:

[0143] ;

[0144] S54, annotation version tree display: Based on the annotation traceability record, an annotation version tree is constructed, and a tree diagram is used to represent the evolution of annotations. Each node represents a version of the annotation, and the lines between nodes represent the inheritance relationship between versions.

[0145] S55, Multi-version annotation comparison: In the user interface, the multi-version switching and comparison functions are provided. After the user selects different versions, the annotation contents of multiple versions are displayed at the same time, and the annotations of different versions are distinguished by color and annotations: .

[0146] S6 data storage includes:

[0147] The database chooses a combination of MongoDB and MySQL, which can store structured data and flexibly manage annotation information.

[0148] Data table design: Design the database table structure, including the book content table, version information table, and annotation information table. The book content table stores the content of each version of the book, the version information table stores the basic information of the version, and the annotation information table stores the detailed information of the annotation and its traceability and modification history.

[0149] Data storage: Use the database insert operation to store the processed book data and annotation information in the corresponding data table to ensure data integrity and consistency.

[0150] Version management unit: Through database operations and background management system, a version management unit is established to realize the functions of creating, modifying and deleting book versions.

[0151] Annotation management unit: Through database operations and the background management system, an annotation management unit is established to realize the functions of adding, modifying, deleting and querying annotations, and maintain the update of the annotation information table.

[0152] The specific data table design is as follows: Design the database table structure, including the book content table, version information table and annotation information table, to ensure the logic of data storage and the efficiency of query;

[0153] Table of contents:

[0154] Table name: BookContent;

[0155] Fields: BookID, VersionID, Content;

[0156] BookID: unique identifier of the book;

[0157] VersionID: unique version identifier;

[0158] Content: book content;

[0159] Version information table:

[0160] Table name: VersionInfo;

[0161] Fields: VersionID, BookID, VersionNumber, ReleaseDate;

[0162] VersionID: unique version identifier;

[0163] BookID: unique identifier of the book;

[0164] VersionNumber: version number;

[0165] ReleaseDate: version release date;

[0166] Annotation information table:

[0167] Table name: AnnotationInfo;

[0168] Fields:

[0169] AnnotationID, BookID, VersionID, AnnotationContent, AnnotationPosition, SourceID, ModificationHistory;

[0170] AnnotationID: unique identifier of annotation;

[0171] BookID: unique identifier of the book;

[0172] VersionID: unique version identifier

[0173] AnnotationContent: annotation content;

[0174] AnnotationPosition: annotation position;

[0175] SourceID: traceability identification;

[0176] ModificationHistory: modification history.

[0177] In S8, a pre-trained language model based on the Transformer architecture, namely the BERT model, is trained to understand the relationship between book content and annotations, including:

[0178] Collecting datasets of book content and annotations ,in For the content of the book, For the corresponding annotations;

[0179] Use the pre-trained BERT model and fine-tune it on the dataset:

[0180] ;

[0181] For each pair of input , sentence embedding is obtained through the BERT model :

[0182] ;

[0183] Add a fully connected layer for classification to predict whether the annotation is relevant:

[0184] ;

[0185] Use the cross entropy loss function to train the model: ;

[0186] Based on the trained BERT model, recommend relevant annotations for new book content.

[0187] Based on the trained BERT model, the following annotations are recommended for new book content:

[0188] Calculating sentence embeddings through the BERT model and , and then calculate the cosine similarity;

[0189] New book content and comments in the database Calculate similarity:

[0190] ;

[0191] ;

[0192] Calculate the similarity score:

[0193] ,in, For book content With annotations similarity;

[0194] Sort by similarity score and generate a list of recommended annotations :

[0195] ,in, is a collection of annotations, is the recommended threshold.

[0196] Automatically analyze the redundancy and repetition of annotations, and prompt users to merge or delete redundant annotations based on the redundant similarity analysis algorithm (i.e. Jaccard similarity), expressed as: ;

[0197] Redundancy prompt: When the similarity exceeds the set threshold When , prompt the user to merge or delete redundant comments, prompt conditions: , then prompt the user to merge or delete.

[0198] like Figure 2 As shown, a system for establishing a book library based on multi-version annotations is used to implement the above-mentioned method for establishing a book library based on multi-version annotations, and includes the following modules:

[0199] Data collection module: used to obtain book data of different versions, including book content, version information and author notes, and the collection methods include OCR technology and user upload;

[0200] Data preprocessing module: preprocess book data, including format conversion, deduplication and segmentation;

[0201] Version comparison module: Use the longest common subsequence algorithm to compare and analyze the contents of different versions of books and generate a version difference report;

[0202] Preliminary annotation association module: Based on the version difference report, the existing annotation information is preliminarily associated with the corresponding book content;

[0203] Inter-version annotation traceability module: Generates a unique traceability identifier for each annotation, records the initial version and modification history of the annotation, provides annotation traceability function, displays the annotation version tree, displays the evolution process of the annotation and the relationship between different versions, and provides annotation comparison function to view the annotations in the same location in different versions;

[0204] Data storage module: stores the processed book data and its annotation information in the database, and establishes version management and annotation management units;

[0205] Interaction module: Design the user interface to provide book version selection, content browsing and annotation viewing functions;

[0206] Intelligent annotation recommendation and optimization module: uses natural language processing technology and the BERT model to analyze book content and existing annotations, recommend relevant annotation content to users, and prompt users to merge or delete redundant annotations;

[0207] Permission control module: set user permissions and control user access rights to different versions of book content and annotation information;

[0208] Data update module: updates book data and annotation information.

[0209] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present invention is limited to these examples. Under the concept of the present invention, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the present invention as described above, which are not provided in detail for the sake of simplicity.

[0210] The present invention is intended to cover all such substitutions, modifications and variations that fall within the broad scope of the claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for establishing a book library based on multi-version annotations, characterized in that: The following steps are involved: S1, data collection: obtaining book data of different versions, wherein the book data includes book content, version information and author notes; S2, data preprocessing: preprocessing the book data, including format conversion, deduplication and segmentation processing; S3, multi-version comparison: using a text comparison algorithm to compare and analyze the contents of different versions of books and generate a version difference report, which includes new content, deleted content, and modified content; S4, preliminary annotation association: according to the version difference report, the existing annotation information is preliminarily associated with the corresponding book content and displayed by version; S5, annotation traceability and association between versions: In the data preprocessing stage, a unique traceability identifier is generated for each annotation, recording the initial version and modification history of the annotation; the annotation traceability function is provided, and users can click on the annotation to view its history and changes in different versions; the annotation version tree is displayed through a graphical interface to intuitively display the evolution process of the annotation and the association between different versions; a multi-version annotation comparison function is provided, and users can simultaneously view the annotations of the same location in different versions; S6, data storage: storing the processed book data and its annotation information in a database, and establishing a version management and annotation management unit; S7, User Interface Design: Design the user interface to provide book version selection, content browsing and annotation viewing functions, so that users can switch between different versions and view the corresponding annotations; S8, Intelligent Annotation Recommendation and Optimization: Utilizes natural language processing technology and machine learning models to analyze book content and existing annotations, and recommend relevant annotation content to users; automatically analyzes the redundancy and repetition of annotations, and prompts users to merge or delete redundant annotations; S9, permission control: set user permissions and control user access rights to different versions of book content and annotation information; S10, data update: regularly or irregularly update book data and annotation information.

2. A method for establishing a book library based on multi-version annotations according to claim 1, characterized in that: The S1 specifically includes: S11, obtaining authorized book version data from an authorized data source, wherein the authorized data source includes a publisher database, an academic resource platform, and an e-book supplier; S12, using optical character recognition technology to scan paper books and convert them into electronic data, identify and extract book content, version information and author notes; S13, the user uploads book data. The user manually uploads the electronic book file through the system interface and supplements the version information and author notes.

3. A method for establishing a book library based on multi-version annotations according to claim 1, characterized in that: The multi-version comparison in S3 specifically includes: S31, normalize the text of different versions of books, including removing spaces, special symbols and formatting text; S32, using the longest common subsequence algorithm to compare and analyze the contents of different versions of the book paragraph by paragraph, and determine the similarities and differences of the text by identifying the longest common subsequence in the text of the two versions; S33, identifying new content, deleted content, and modified content in different versions of the book content by using the longest common subsequence algorithm, and generating a version difference report, wherein the report records the content changes between each version; S34, mark the version differences in the book content, highlight the new content, indicate the deleted content with a strikethrough, and mark the modified content with different colors to distinguish the content before and after the modification.

4. A method for establishing a book library based on multi-version annotations according to claim 3, characterized in that: The longest common subsequence algorithm determines the similarity and difference of texts by identifying the longest common subsequence in two versions of texts, specifically including: Define two strings and ; Create a two-dimensional array ,in Represents a string Before Characters and strings Before The length of the longest common subsequence of characters; Construct a recursive relationship: = ,in, Indicates conditional judgment, that is, if the condition is met, the corresponding operation is performed, or means or, that is, when one of the conditions i = 0 or j = 0 is met, the corresponding operation can be performed; at last, It's a string and The length of the longest common subsequence of ; according to The array is backtracked to find the longest common subsequence and mark the added, deleted and modified content; Difference identification: Use the longest common subsequence algorithm to identify new content, deleted content, and modified content in different versions of books, and generate a version difference report that records the content changes between each version in detail; New content recognition: exists in The parts that do not exist in the content are marked as new content; Deleting content identification: exists in The parts that do not exist in the content are marked as deleted. Modify content identification: For and Parts that exist in both but have different contents are marked as modified content.

5. The method for establishing a book library based on multi-version annotations according to claim 1, characterized in that: The preliminary annotation associations in S4 specifically include: S41, version difference location: According to the generated version difference report, determine the location of the addition, deletion and modification of the book content, generate a location mapping table, and and version , define the location mapping table ,in Indicates version Middle Positions in version The corresponding position in the , generation method includes using the result of the longest common subsequence algorithm, traversing the content of the two versions of the book, and recording the corresponding position of each character or word, expressed as: ; S42, annotation location: Use the location mapping table in the version difference report to locate the existing annotation information to the specific location of the book content, generate a preliminary annotation association table, and set the annotation table Contains the annotation location and annotation content , generate a preliminary annotation association table : , according to the location mapping table , the annotation table Each annotation position in Map to the location of the new version and generate a preliminary annotation association table ; S43, annotation classification: according to the version information of the annotation, the annotation is classified and stored by version; S44, annotation display: in the user interface, according to the book version selected by the user, the annotation information of the corresponding version is displayed, and the annotation information includes the version identifier and the annotation content; S45, synchronous update of annotations: when the book content is updated, the position mapping of the annotations is automatically adjusted according to the new version difference report to ensure the relevance of the annotation information to the book content.

6. The method for establishing a book library based on multi-version annotations according to claim 1, characterized in that: The annotation traceability and association between versions in S5 specifically include: S51, traceability identification generation: In the data preprocessing stage, a unique traceability identification is generated for each comment. The traceability identification is generated by combining the hash value of the comment content and the comment position, and recording the initial version and modification history of the comment. Suppose the comment content is , the annotation position is , the initial version is , generate traceability identification , expressed as: ； S52, annotation traceability record: records the initial version and modification history of the annotation, including the timestamp and modification content of each modification. The format of each annotation record is ,in, is the timestamp, The modified annotation content; S53, annotation traceability function: Provide annotation traceability function, users click on the annotation to view its history and changes in different versions, according to the traceability mark , retrieve the corresponding annotation history in the database, generate a history list and display it: ； S54, annotation version tree display: Based on the annotation traceability record, an annotation version tree is constructed, and a tree diagram is used to represent the evolution of annotations. Each node represents a version of the annotation, and the lines between nodes represent the inheritance relationship between versions. S55, Multi-version annotation comparison: In the user interface, the multi-version switching and comparison functions are provided. After the user selects different versions, the annotation contents of multiple versions are displayed at the same time, and the annotations of different versions are distinguished by color and annotations: .

7. The method for establishing a book library based on multi-version annotations according to claim 1, characterized in that: In S8, a pre-trained language model based on the Transformer architecture, namely the BERT model, is trained to understand the relationship between the book content and annotations, including: Collecting datasets of book content and annotations ,in For the content of the book, For the corresponding annotations; Use the pre-trained BERT model and fine-tune it on the dataset: ； For each pair of input , sentence embedding is obtained through the BERT model : ； Add a fully connected layer for classification to predict whether the annotation is relevant: ； Use the cross entropy loss function to train the model: ; Based on the trained BERT model, recommend relevant annotations for new book content.

8. The method for establishing a book library based on multi-version annotations according to claim 7, characterized in that: The method of recommending relevant annotations for new book content based on the trained BERT model specifically includes: New book content and comments in the database Calculate similarity: ； ； Calculate the similarity score: ,in, For book content With annotations similarity; Sort by similarity score and generate a list of recommended annotations : ,in, is a collection of annotations, is the recommended threshold.

9. A method for establishing a book library based on multi-version annotations according to claim 8, characterized in that: The automatic analysis of the redundancy and repetitiveness of annotations and prompting the user to merge or delete redundant annotations is based on a redundant similarity analysis algorithm, which is expressed as: ; Redundancy prompt: When the similarity exceeds the set threshold , prompts the user to merge or delete redundant comments.

10. A system for establishing a book library based on multi-version annotations, used to implement a method for establishing a book library based on multi-version annotations as claimed in any one of claims 1 to 9, characterized in that: Includes the following modules: Data collection module: used to obtain book data of different versions, including book content, version information and author notes, and the collection methods include OCR technology and user upload; Data preprocessing module: preprocessing the book data, including format conversion, deduplication and segmentation processing; Version comparison module: Use the longest common subsequence algorithm to compare and analyze the contents of different versions of books and generate a version difference report; Preliminary annotation association module: Based on the version difference report, the existing annotation information is preliminarily associated with the corresponding book content; Inter-version annotation traceability module: Generates a unique traceability identifier for each annotation, records the initial version and modification history of the annotation, provides annotation traceability function, displays the annotation version tree, displays the evolution process of the annotation and the relationship between different versions, and provides annotation comparison function to view the annotations in the same location in different versions; Data storage module: stores the processed book data and its annotation information in the database, and establishes version management and annotation management units; Interaction module: Design the user interface to provide book version selection, content browsing and annotation viewing functions; Intelligent annotation recommendation and optimization module: uses natural language processing technology and the BERT model to analyze book content and existing annotations, recommend relevant annotation content to users, and prompt users to merge or delete redundant annotations; Permission control module: set user permissions and control user access rights to different versions of book content and annotation information; Data update module: updates book data and annotation information.

Images