Work infringement intelligent monitoring system

Abstract

The invention relates to the technical field of network infringement monitoring, in particular to an intelligent monitoring system for works infringement, which comprises a monitoring system and is used for monitoring works infringement, and the monitoring system comprises an acquisition unit for acquiring target works and related contents from the Internet, a database and a social media platform; the acquisition unit is used for receiving and collecting work data uploaded by a plurality of users and storing the work data as a work database as basic data for infringement monitoring; the preprocessing unit is used for carrying out cleaning, duplicate removal and format conversion operation on the collected data; the feature extraction unit is used for extracting key features from the preprocessed data of the specific type of works; the comparison unit is used for comparing the monitored work data uploaded by the user with the works in the basic database one by one to generate comparison result data of similarity and similar points; monitoring of multiple types of works is supported, infringement is efficiently and accurately recognized, the efficiency is improved through the automatic process, and copyright disputes can be quickly coped with.

Description

Technical Field

[0001] The present invention relates to the technical field of network infringement monitoring, and in particular to an intelligent monitoring system for work infringement. Background Art

[0002] With the rapid development of the Internet, the creation and dissemination of works have become more convenient, but it has also brought about infringement issues.

[0003] After checking the publication number: CN112667972B, a method for monitoring infringement, a first platform and a device for monitoring infringement are disclosed. This technology discloses "a method for monitoring infringement, a first platform and a device for monitoring infringement. The method includes: obtaining ciphertext data of first digital content, the first digital content including: digital content corresponding to a subject with copyright protection; searching for second digital content that matches key information of the first digital content on various network platforms; extracting a first feature of the ciphertext data, and extracting a second feature of the second digital content; comparing the first feature and the second feature to obtain the similarity between the first feature and the second feature; and when it is determined that the similarity meets a preset condition and the platform corresponding to the second digital content is not authorized, determining that the corresponding platform is infringing" and other technical solutions, which have technical effects such as "preventing digital content from being leaked and improving the security of digital content";

[0004] Existing work infringement monitoring systems have two core flaws: First, they use a single feature extraction strategy for cross-domain complex works (such as software documents containing code and text), only extracting shallow text features or code infrastructure, resulting in the systematic omission of deep infringement evidence such as code logical similarity and text semantic relevance; second, they lack cross-modal comparison technology capabilities and are unable to detect multimedia mixed infringement (such as stealing video soundtracks and tampering with subtitle content). The traditional single-type comparison framework makes complex scenarios involving multi-factor collaborative infringement a monitoring blind spot, causing complex infringement behaviors to evade technical identification. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention provides an intelligent monitoring system for work infringement, which supports monitoring of multiple types of works, efficiently and accurately identifies infringements, automates processes to improve efficiency, and helps to quickly respond to copyright disputes.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: an intelligent monitoring system for work infringement, including a monitoring system for monitoring work infringement, the monitoring system comprising:

[0007] The collection unit collects target works and related content from the Internet, databases, and social media platforms;

[0008] The acquisition unit receives and aggregates the work data uploaded by multiple users and stores it in a work database, which serves as the basic data for infringement monitoring;

[0009] The pre-processing unit cleans, removes duplicates, and converts the format of the collected data;

[0010] A feature extraction unit extracts key features from the pre-processed data of a specific type of work;

[0011] The comparison unit compares the monitored work data uploaded by the user with the works in the basic database one by one, and generates comparison result data of similarity and similarity points.

[0012] Preferably, the acquisition unit includes:

[0013] Data receiving module, receiving work data uploaded by users;

[0014] Data verification module, which performs format verification and content check on uploaded data;

[0015] The data storage module stores the verified works data in the database;

[0016] Data classification module, classifies the work data;

[0017] Data index module, creating indexes for work data;

[0018] Data labeling module, adding labels to work data;

[0019] Data security module encrypts and stores data.

[0020] Preferably, the pre-processing unit includes:

[0021] Data cleaning module to remove noise data, invalid data and redundant data;

[0022] Data deduplication module, identifying and deleting duplicate data;

[0023] Format conversion module, converting data into a unified format;

[0024] Standardization module, which normalizes the data.

[0025] Preferably, the feature extraction unit includes:

[0026] Music feature extraction module, which extracts melody, rhythm, and harmony features from music score data;

[0027] Artwork feature extraction module, extracting color, texture, and shape features from image data;

[0028] Film and television feature extraction module, which extracts plot, scene, and character features from text scripts and video materials;

[0029] The Chinese character feature extraction module extracts the features of glyph shape, semantics, and structure from text data;

[0030] The computer software work feature extraction module extracts the features of grammatical structure, function call, and logical flow from computer language code data.

[0031] Preferably, the comparison unit includes:

[0032] The writing style comparison module compares the similarity of text works and identifies similarities;

[0033] Image comparison module, which compares the similarity of image works and identifies similar areas;

[0034] An audio comparison module compares the similarity of audio works and identifies similar segments;

[0035] Video comparison module, which compares the similarity of video works and identifies similar segments;

[0036] The code comparison module compares the similarity of computer software works and identifies similar code segments.

[0037] Preferably, the monitoring system further comprises an infringement detection unit for determining whether infringement occurs by comparing features.

[0038] Preferably, the monitoring system further comprises a result analysis unit for analyzing the detection results and generating a report.

[0039] Preferably, the monitoring system further comprises a feedback unit for feeding back the parts of the comparison result data whose similarity exceeds the user preset value to the preset users one by one for further monitoring and processing by the users.

[0040] The present invention provides an intelligent work infringement monitoring system. Compared with the existing technology, it has the following advantages:

[0041] 1. Through the feature extraction unit, dedicated modules for different types of works (such as music, art, film and television, text, and code) can be used to achieve efficient and accurate feature extraction. The music work module extracts features such as melody, rhythm, and harmony; the art work module extracts features such as color, texture, and shape; the film and television work module extracts features such as plot, scene, and character; the Chinese character work module extracts features such as glyph, semantics, and structure; and the computer software work module extracts features such as grammar, function, and logic. Through technologies such as pitch sequence analysis, convolutional neural networks (CNN), word embedding (such as BERT), and syntax tree construction, the system can extract key features from complex data and convert them into vector representations, providing high-quality feature input for subsequent infringement detection and similarity calculations.

[0042] 2. The effectiveness of infringement detection is significantly improved through the use of comparison unit multimodal processing technology, supporting the comparison of all types of works including text, images, audio, video and code. Specialized algorithms such as cosine similarity and image hashing are used to achieve efficient feature matching, and technologies such as SSIM and spectrum analysis are combined to ensure the accuracy of scoring. Advanced algorithms such as Needleman-Wunsch alignment and SIFT region matching can accurately locate similar paragraphs, regions or fragments, generating structured data containing scoring and location information, providing a dual basis for infringement determination, improving detection efficiency and strengthening the integrity of the evidence chain, effectively dealing with complex copyright dispute scenarios.

[0043] 3. The acquisition unit enables efficient reception, verification, storage, and management of user-uploaded work data. Users can upload work data through the API interface or file upload function. The system verifies the format and content of the data to ensure the legality and validity of the data. Data that passes the verification is stored in the database and classified and subdivided according to type (such as text, image, audio, video) and content theme (such as literature, art, music, film and television). By creating indexes and automatic labeling, the system significantly improves data retrieval efficiency and facilitates subsequent comparison and analysis. In addition, the data security module encrypts and stores the data to ensure data security and privacy. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a system block diagram of the present invention;

[0045] Figure 2 A block diagram of an acquisition unit in the present invention;

[0046] Figure 3 is a block diagram of the pre-processing unit in the present invention;

[0047] Figure 4 is a block diagram of a feature extraction unit in the present invention;

[0048] Figure 5FIG. 4 is a block diagram of a comparison unit in the present invention.

[0049] In the figure: 1. Monitoring system; 11. Acquisition unit; 12. Acquisition unit; 121. Data receiving module; 122. Data verification module; 123. Data storage module; 124. Data classification module; 125. Data indexing module; 126. Data labeling module; 127. Data security module; 13. Preprocessing unit; 131. Data cleaning module; 132. Data deduplication module; 133. Format conversion module; 134. Standardization module; 14. Feature extraction unit; 141. Music feature extraction module; 142. Art feature extraction module; 143. Film and television feature extraction module; 144. Chinese character feature extraction module; 145. Computer software feature extraction module; 15. Comparison unit; 151. Writing comparison module; 152. Image comparison module; 153. Audio comparison module; 154. Video comparison module; 155. Code comparison module; 16. Infringement detection unit; 17. Result analysis unit; 18. Feedback unit. DETAILED DESCRIPTION

[0050] The following will provide a clear and complete description of 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. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0051] See also Figure 1 - Figure 5 The present invention provides a technical solution: an intelligent monitoring system for work infringement, including a monitoring system 1 for monitoring work infringement, wherein the monitoring system 1 includes:

[0052] Collection unit 11, collecting target works and related content from the Internet, databases, and social media platforms;

[0053] The acquisition unit 12 receives and aggregates the work data uploaded by multiple users and stores it in a work database as basic data for infringement monitoring;

[0054] The pre-processing unit 13 performs operations of cleaning, deduplication, and format conversion on the collected data;

[0055] A feature extraction unit 14 extracts key features from the pre-processed data of a specific type of work;

[0056] The comparison unit 15 compares the monitored work data uploaded by the user with the works in the basic database one by one, and generates comparison result data of similarity and similarity points.

[0057] In this implementation plan, first, the system supports infringement monitoring of various types of works (such as text, images, audio, video, and code), which can meet the needs of different users and cover a wide range of copyright protection scenarios; secondly, the system significantly improves data quality and processing efficiency through data cleaning, deduplication, format conversion, and standardization; at the same time, it uses indexing technology and approximate nearest neighbor search (ANN) algorithm to accelerate large-scale data comparison and ensure efficient operation; finally, the system combines similarity calculation, pattern recognition, and classification algorithms to achieve accurate infringement judgment; through feature extraction and similarity analysis, it provides detailed infringement evidence to help users quickly locate infringement behaviors.

[0058] Specifically, the acquisition unit 12 includes:

[0059] The data receiving module 121 receives the work data uploaded by the user;

[0060] The data verification module 122 performs format verification and content checking on the uploaded data;

[0061] The data storage module 123 stores the verified work data in a database;

[0062] A data classification module 124 classifies the work data;

[0063] The data index module 125 creates an index for the work data;

[0064] Data labeling module 126, adding labels to work data;

[0065] The data security module 127 encrypts and stores the data.

[0066] In this embodiment, the user uploads the work data through an API interface (such as RESTful API, GraphQL) or a file upload function (such as HTTP file upload, FTP upload); the format of the uploaded original work data is then verified (such as file header verification, MIME type check), and the content is checked (such as text content filtering, image resolution check); the verified work data is then stored in a database, and the stored work data is classified according to type (such as text, image, audio, video), and then subdivided according to content theme (such as literature, art, music, film and television), and an index is created for the classified work data to improve retrieval efficiency, and then NLP keyword extraction and image tag generation technology are used to achieve automatic tagging, add tags to the indexed work data, facilitate subsequent retrieval and comparison, and encrypt and store the labeled work data to ensure data security.

[0067] Specifically, the pre-processing unit 13 includes:

[0068] Data cleaning module 131, removing noise data, invalid data and redundant data;

[0069] a data deduplication module 132, which identifies and deletes duplicate data;

[0070] The format conversion module 133 converts the data into a unified format;

[0071] The standardization module 134 performs normalization processing on the data.

[0072] In this embodiment, NLP tools, image processing libraries, audio processing libraries, and video processing libraries are used for cleaning, and then hash algorithms or similarity algorithms are used for precise or fuzzy deduplication. Then, encoding conversion tools, image processing libraries, audio processing libraries, and video processing libraries are used for format conversion. Then, NLP tools, image processing libraries, audio processing libraries, and video processing libraries are used for standardization. The processed data is stored in a database, and metadata and log files are generated.

[0073] Specifically, the feature extraction unit 14 includes:

[0074] The music feature extraction module 141 extracts the features of melody, rhythm and harmony from the music score data;

[0075] Artwork feature extraction module 142 extracts color, texture, and shape features from image data;

[0076] The film and television work feature extraction module 143 extracts the features of the plot, scene, and character from the text script and image data;

[0077] The Chinese character work feature extraction module 144 extracts the features of the character shape, semantics, and structure from the text data;

[0078] The computer software work feature extraction module 145 extracts features of grammatical structure, function call, and logic flow from computer language code data.

[0079] In this embodiment, first, the corresponding feature extraction module is called according to the type of work; musical works use pitch sequence analysis, beat tracking, chord recognition and other technologies; art works use color histogram, gray-level co-occurrence matrix (GLCM), convolutional neural network (CNN) and other technologies; film and television works use plot analysis, key frame extraction, 3D convolutional neural network (3D CNN) and other technologies; Chinese character works use stroke count calculation, word embedding (such as BERT), sentence structure analysis and other technologies; computer software works use syntax tree construction, function call graph, control flow analysis and other technologies; finally, the extracted features are converted into vector representation for subsequent infringement detection and similarity calculation.

[0080] Specifically, the comparison unit 15 includes:

[0081] a writing style comparison module 151 that compares the similarity of text works and identifies similarities;

[0082] An image comparison module 152 compares the similarity of the image works and identifies similar areas;

[0083] an audio comparison module 153 that compares the similarity of audio works and identifies similar segments;

[0084] a video comparison module 154 that compares the similarity of video works and identifies similar segments;

[0085] The code comparison module 155 compares the similarity of computer software works and identifies similar code segments.

[0086] In this embodiment, first, the corresponding comparison module is called according to the type of work, and a similarity algorithm (such as cosine similarity, image hash, audio fingerprint, etc.) is used to perform feature matching, and the features of the monitored work are compared with the features of the works in the basic database; then, a dedicated algorithm (such as cosine similarity of text, SSIM of image, spectrum comparison of audio, etc.) is used according to the type of work to calculate the similarity score between the monitored work and the works in the basic database; then, alignment algorithms (such as Needleman-Wunsch), region matching (such as SIFT), fragment matching (such as DTW) and other technologies are used to identify specific similarities (such as text paragraphs, image regions, audio clips, etc.); finally, comparison result data including similarity scores and similarity point analysis is generated, and the similarity scores and similarity point information are integrated into structured data for subsequent infringement detection.

[0087] Specifically, the monitoring system 1 further includes an infringement detection unit 16 for determining whether an infringement occurs by comparing features.

[0088] In this embodiment, the features of the work uploaded by the user and the features of the work in the basic database are received, and similarity algorithms (such as cosine similarity, image hash, audio fingerprint, etc.) are used to perform feature matching, and the features of the work uploaded by the user are compared with the features of the work in the basic database; then pattern recognition (such as cluster analysis) and classification algorithms (such as SVM, random forest) are used to make a judgment, and based on the comparison results, it is determined whether there is infringement; finally, the similarity score and similarity information are integrated to generate an infringement detection report, which includes the infringement judgment results and evidence.

[0089] Specifically, the monitoring system 1 further includes a result analysis unit 17 for analyzing the detection results and generating a report.

[0090] In this embodiment, the output results of the infringement detection module (such as infringement determination results, similarity scores, and infringement evidence) are received, and statistical analysis methods such as descriptive statistics and hypothesis testing are used to perform statistical analysis on the detection results and extract key information (such as mean, variance, distribution, etc.); visualization tools (such as Matplotlib and D3.js) can be used to generate bar charts, line charts, pie charts, etc. to intuitively display the analysis results; and report generation tools (such as Jupyter Notebook and LaTeX) and automated scripts are used to generate a detailed report containing infringement determination results, similarity scores, and infringement evidence.

[0091] Specifically, the monitoring system 1 further includes a feedback unit 18 for feeding back the parts of the comparison result data whose similarity exceeds the user-preset value to the preset users one by one for further monitoring and processing by the users.

[0092] In this embodiment, a conditional filtering algorithm is used to filter out comparison results whose similarity exceeds the user's preset value, a template engine (such as Jinja2) is used to generate feedback text, and an image processing library (such as OpenCV, PIL) is used to generate infringement evidence screenshots. The feedback content includes detailed information on the infringing work, similarity score, similarity analysis and infringement evidence screenshots.

[0093] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0094] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent system for monitoring copyright infringement, characterized by: The monitoring system (1) is used to monitor the infringement of the works, and the monitoring system (1) includes: The collection unit (11) collects target works and related content from the Internet, databases, and social media platforms; An acquisition unit (12) receives and aggregates work data uploaded by multiple users and stores the data as a work database, which serves as basic data for infringement monitoring; A pre-processing unit (13) performs operations of cleaning, deduplication, and format conversion on the collected data; A feature extraction unit (14) extracts key features from the pre-processed data of a specific type of work; The comparison unit (15) compares the monitored work data uploaded by the user with the works in the basic database one by one, and generates comparison result data of similarity and similarity points.

2. The intelligent work infringement monitoring system according to claim 1, characterized in that: The acquisition unit (12) comprises: A data receiving module (121) receives work data uploaded by users; A data verification module (122) performs format verification and content checking on the uploaded data; A data storage module (123) stores the verified work data in a database; A data classification module (124) is used to classify the work data; A data indexing module (125) creates an index for the work data; Data labeling module (126), adding labels to work data; The data security module (127) encrypts and stores the data.

3. The intelligent work infringement monitoring system according to claim 1, characterized in that: The pre-processing unit (13) comprises: A data cleaning module (131) removes noise data, invalid data and redundant data; a data deduplication module (132) for identifying and deleting duplicate data; A format conversion module (133) converts data into a unified format; The standardization module (134) performs normalization processing on the data.

4. The intelligent work infringement monitoring system according to claim 1, characterized in that: The feature extraction unit (14) comprises: A music feature extraction module (141) extracts melody, rhythm, and harmony features from music score data; Artwork feature extraction module (142), extracting color, texture, and shape features from image data; Film and TV feature extraction module (143) extracts plot, scene, and character features from text scripts and video materials; A Chinese character work feature extraction module (144) extracts features of character shape, semantics, and structure from text data; The computer software work feature extraction module (145) extracts features of grammatical structure, function call and logic flow from computer language code data.

5. The intelligent work infringement monitoring system according to claim 1, characterized in that: The comparison unit (15) comprises: The writing style comparison module (151) compares the similarity of text works and identifies similarities; An image comparison module (152) compares the similarity of the image works and identifies similar areas; an audio comparison module (153) for comparing similarities between audio works and identifying similar segments; A video comparison module (154) compares the similarity of video works and identifies similar segments; The code comparison module (155) compares the similarity of computer software works and identifies similar code segments.

6. The intelligent work infringement monitoring system according to claim 1, characterized in that: The monitoring system (1) further comprises an infringement detection unit (16) for determining whether an infringement occurs by comparing features.

7. The intelligent work infringement monitoring system according to claim 1, characterized in that: The monitoring system (1) further comprises a result analysis unit (17) for analyzing the detection results and generating a report.

8. The intelligent work infringement monitoring system according to claim 1, characterized in that: The monitoring system (1) further comprises a feedback unit (18) for feeding back the parts of the comparison result data whose similarity exceeds a user preset value to a preset user one by one for further monitoring and processing by the user.

Images