AI creation real-time monitoring method and system based on big data analysis

Abstract

The invention discloses an AI creation real-time monitoring method and system based on big data analysis, and belongs to the technical field of data analysis. A semantic feature word and part-of-speech feature word library is established, and AI creation texts are stored and marked in real time; semantic feature words in the text are recognized, a word relation group is generated based on part-of-speech feature word separation, and a creation style portrait matrix is constructed; stripping the portrait block matrix through a creation time sliding window, and calculating style similarity of different creation subjects in the same time window; calculating the similarity stability of the creation behaviors based on the similarity to generate an early warning relation pair; according to the method, real-time monitoring and style analysis of AI creation are realized, similar behaviors among creation subjects can be continuously and effectively recognized and early warned, the originality and safety monitoring efficiency of AI creation content is improved, particularly, the combination relation of semantic feature words and part-of-speech feature words and a deep creation style mode can be captured, and the method is suitable for popularization and application. The method can effectively distinguish the accidental similar style from the persistent style.

Description

Technical Field

[0001] The present invention relates to the field of data analysis technology, and specifically to a method and system for real-time monitoring of AI creation based on big data analysis. Background Art

[0002] With the rapid development of artificial intelligence technology, AI creation is increasingly used in literature, art, coding and other fields; however, the convenience of AI creation also brings about problems such as content homogeneity and difficulty in plagiarism detection.

[0003] Existing AI creation monitoring methods are mostly based on keyword matching or simple text similarity calculations, which make it difficult to deeply analyze the intrinsic characteristics of creative style (such as semantic structure, part-of-speech usage habits, etc.), and lack the ability to track creative behavior in real time. For example, the feature analysis dimension is single: it only relies on surface text features (such as keyword repetition rate), and cannot capture the combination relationship between semantic feature words and part-of-speech feature words and deep creative style patterns; time dimension monitoring is missing: the dynamic changes of creative behavior in the time series are not taken into account, and it is difficult to distinguish between accidental similarities and continuous style convergence. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for real-time monitoring of AI creation based on big data analysis to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] A real-time monitoring system for AI creation based on big data analysis. The system includes: a text storage management module, a feature recognition and matrix construction module, a sliding window analysis and similarity assessment module, and a stability calculation and early warning module.

[0007] The text storage management module is used to establish a vocabulary and encode words, build a text library to store AI-created texts, add creation time to the texts and use the creator ID code as a label to generate a creation text sub-library;

[0008] The feature recognition and matrix construction module is used to identify semantic feature words in the creative text, separate semantic feature words based on part-of-speech feature words, form part-of-speech relationship groups and attach part-of-speech tags, count the number of part-of-speech relationship groups, and construct a creative style portrait matrix;

[0009] The sliding window analysis and similarity assessment module is used to set a creation time sliding window, separate the portrait block matrix from the creation style portrait matrix, and evaluate the similarity of the creation styles of different creators in the same time period;

[0010] The stability calculation and warning module is used to calculate the similarity stability of creative behaviors between creative subjects based on the similarity of creative styles, preset a similarity stability threshold, and form and output a warning relationship pair when the similarity stability reaches the standard.

[0011] Furthermore, the text storage management module includes a vocabulary configuration unit and a text real-time recording unit;

[0012] The word library configuration unit is used to establish a word library of semantic features and part-of-speech features, and to uniformly encode the semantic feature words and part-of-speech feature words respectively;

[0013] The real-time text recording unit is used to build a text library to store AI-created texts in real time, and to add the creation time to the AI-created texts, and to encode the creation subject ID as a creation text sub-library label to generate the creation subject's creation text sub-library.

[0014] Furthermore, the feature recognition and matrix construction module includes a part-of-speech relationship group recognition unit and a creative style portrait processing unit;

[0015] The part-of-speech relationship group identification unit is used to identify each semantic feature word in the creative text, and based on the part-of-speech feature words, separate each semantic feature word to form a part-of-speech relationship group and add a part-of-speech tag;

[0016] The creative style portrait processing unit counts the number of part-of-speech relationship groups under the same part-of-speech feature word in the creative text as matrix elements based on the part-of-speech tags to form a creative style portrait matrix.

[0017] Furthermore, the sliding window analysis and similarity assessment module includes a portrait segmentation unit and a creative style assessment unit;

[0018] The portrait block unit is used to set a creation time sliding window to separate the portrait block matrix from the creation style portrait matrix;

[0019] The creative style evaluation unit evaluates the similarity of the creative styles of different creative subjects in the same time period based on the portrait block matrix.

[0020] Furthermore, the stability calculation and warning module includes a long-term behavior analysis unit and a monitoring and warning unit;

[0021] The long-term behavior analysis unit calculates the similarity and stability of the creative behaviors between the creative subjects based on the similarity of the creative styles;

[0022] The monitoring and early warning unit is used to preset a similarity stability threshold. If the similarity stability is greater than or equal to the similarity stability threshold, an early warning relationship pair is formed between the creative subjects and the early warning relationship pair is output.

[0023] A method for real-time monitoring of AI creation based on big data analysis, comprising the following steps:

[0024] Step S1: Establish a vocabulary containing semantic feature words and part-of-speech feature words, and build a creative text library to store AI-created texts in real time. Add the creation time node to the creative text and use the creator ID code as a label to generate a creative text sub-library;

[0025] Step S2: Identify semantic feature words in the creative text, separate the semantic feature words based on the part-of-speech feature words, form part-of-speech relationship groups and attach part-of-speech tags, count the number of part-of-speech relationship groups under the same part-of-speech feature word, and construct a creative style portrait matrix of the creative subject;

[0026] Step S3: Setting a creation time sliding window, separating the portrait block matrix from the creation style portrait matrix based on the sliding window, and evaluating the similarity of the creation styles of different creators in the same time period;

[0027] Step S4: Calculate the similarity stability of creative behaviors between creative subjects based on the similarity of creative styles, and preset a similarity stability threshold to form and output a warning relationship pair between different creative subjects.

[0028] Furthermore, the specific implementation process of step S1 includes:

[0029] Establish a vocabulary of semantic features and part-of-speech features, and uniformly encode the semantic feature words and part-of-speech feature words respectively, wherein each semantic feature word and each part-of-speech feature word has a unique encoding attribute, and the rth semantic feature word is recorded as SW r , record the f-th part-of-speech feature word as RW f ;

[0030] Build a text library to store AI-created texts in real time, add creation time to the AI-created texts, use the creator ID code as the creation text sub-library tag, and generate the creation text sub-library of the creator, recorded as CD v ={cd i (t x )|i∈[1, I], x∈[1, y]}, where, CD v Indicates the creative text sub-library generated by the creative subject coded as v, cd i (t x ) represents the addition of the xth creation time node t x The i-th creative text, I represents the total number of creative texts, and y represents the total number of creation time nodes.

[0031] Furthermore, the specific implementation process of step S2 includes:

[0032] Identify creative text cd i (t x ) and separate the semantic feature words based on the part-of-speech feature words. If the semantic feature word SW r and semantic feature words SW m The part-of-speech relationship between them is the part-of-speech feature word RW f , then form a part-of-speech relationship group [SW r , SW m ], and the part-of-speech relationship group [SW r , SW m ]Add part-of-speech tag RW f :[SW r , SW m ], among which SW m represents the mth semantic feature word, and m≠r;

[0033] Based on part-of-speech tag RW f :[SW r , SW m ], statistical creation text cd i (t x ) in the same part of speech feature word RW f The number of part-of-speech relationship groups under i (t x )|RW f ], and construct the creative style portrait matrix of the creative subject v with the creative text coding number as the row number and the part-of-speech feature word coding as the column number, recorded as L IF (v) and the number of part-of-speech relationship groups NUM[cd i (t x )|RW f ] is mapped to the creative style portrait matrix L IF (v) The i-th row and the f-th column, where F is the total number of part-of-speech feature words.

[0034] Furthermore, the specific implementation process of step S3 includes:

[0035] Set a creation time sliding window, the scale value of the creation time sliding window is g, if the starting time of the qth creation time sliding window is t x , then the end time of the qth creation time sliding window is t x+g , based on the creation time sliding window T g (q), in the creation style portrait matrix L IF (v) The image block matrix is ​​stripped out and recorded as L IF (v, q), and the image block matrix L |q|F (v), where |q| represents the image block matrix L |q|F(v) The number of rows, the starting row number of the row number |q| is the starting time t x The corresponding creative text code, the end line number of line number |q| is the end time t x+g The corresponding creative text code;

[0036] Based on the portrait block matrix, evaluate the similarity of creative styles of different creative subjects at the same time stage Where, L |q|F (k) represents the image block matrix corresponding to the creative subject coded as k, Indicates that there is a block matrix L in the image |q|F (v)Number of part-of-speech relationship groups NUM[cd i (t x )|RW f ], Indicates that there is a block matrix L in the image |q|F (k)Number of part-of-speech relationship groups NUM[cd i (t x )|RW f ];

[0037] In the above method, the similarity of creative styles is calculated based on the matrix norm principle. Specifically, it is obtained by taking the square root of the sum of the squares of the differences of the matrix element values ​​(the number of part-of-speech relationship groups) corresponding to each row and column in the block matrix of the same-type portrait. The greater the similarity of creative styles, the more similar the creative styles of the subjects at different stages are.

[0038] Furthermore, the specific implementation process of step S4 includes:

[0039] Based on the similarity of creative styles, calculate the similarity stability of creative behaviors between creative subjects Where Q is the total number of creation time sliding windows, β is the mean value of creation style similarity, and σ 2 is the variance of creative style similarity, and

[0040] A similarity stability threshold is preset. If the similarity stability is greater than or equal to the similarity stability threshold, a warning relationship pair is formed between the creative subjects v and k, and the warning relationship pair is output.

[0041] Compared with the prior art, the beneficial effects achieved by the present invention are: establishing a semantic feature word and part-of-speech feature word library, storing and marking AI creation texts in real time; identifying semantic feature words in the text, generating part-of-speech relationship groups based on the separation of part-of-speech feature words, and constructing a creation style portrait matrix; peeling off the portrait block matrix through the creation time sliding window, and calculating the style similarity of different creation subjects in the same time window; calculating the similarity stability of the creation behavior based on the similarity to generate a warning relationship pair; the present invention realizes real-time monitoring and style analysis of AI creation, and can continuously and effectively identify similar behaviors between creation subjects and issue warnings, thereby improving the originality and security monitoring efficiency of AI creation content. In particular, the present invention can capture the combination relationship between semantic feature words and part-of-speech feature words and deep creation style patterns, and can effectively distinguish between accidental similarities and continuous style convergence. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0043] Figure 1 This is a schematic diagram of the steps of a real-time monitoring method for AI creation based on big data analysis of the present invention. DETAILED DESCRIPTION

[0044] 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.

[0045] In the first embodiment of the present invention, a real-time monitoring system for AI creation based on big data analysis is provided. The system includes: a text storage management module, a feature recognition and matrix construction module, a sliding window analysis and similarity assessment module, and a stability calculation and early warning module.

[0046] The text storage management module is used to establish a vocabulary and encode words, build a text library to store AI-created texts, add creation time to the texts, and generate a sub-library of created texts using the creator ID code as a label;

[0047] Among them, the text storage management module includes a vocabulary configuration unit and a text real-time recording unit;

[0048] The word library configuration unit is used to establish a word library of semantic features and part-of-speech features, and to uniformly encode the semantic feature words and part-of-speech feature words respectively;

[0049] The real-time text recording unit is used to build a text library to store AI-created texts in real time, add the creation time to the AI-created texts, encode the creator ID as the creation text sub-library tag, and generate the creation text sub-library of the creator;

[0050] The feature recognition and matrix construction module is used to identify semantic feature words in the creative text, separate semantic feature words based on part-of-speech feature words, form part-of-speech relationship groups and attach part-of-speech tags, count the number of part-of-speech relationship groups, and construct a creative style portrait matrix;

[0051] Among them, the feature recognition and matrix construction module includes a part-of-speech relationship group recognition unit and a creative style portrait processing unit;

[0052] A part-of-speech relationship group recognition unit is used to identify the semantic feature words in the creative text and separate the semantic feature words based on the part-of-speech feature words to form a part-of-speech relationship group and add a part-of-speech tag;

[0053] The creative style portrait processing unit counts the number of part-of-speech relationship groups under the same part-of-speech feature words in the creative text based on the part-of-speech tags as matrix elements to form a creative style portrait matrix;

[0054] The sliding window analysis and similarity assessment module is used to set a creation time sliding window, separate the portrait block matrix from the creation style portrait matrix, and evaluate the similarity of the creation styles of different creators in the same time period;

[0055] Among them, the sliding window analysis and similarity assessment module includes an image segmentation unit and a creative style assessment unit;

[0056] A portrait block unit is used to set a creation time sliding window to separate a portrait block matrix from a creation style portrait matrix;

[0057] The creative style evaluation unit evaluates the similarity of creative styles of different creative subjects at the same time period based on the portrait block matrix;

[0058] The stability calculation and early warning module is used to calculate the similarity stability of creative behaviors between creative subjects based on the similarity of creative styles, preset similarity stability thresholds, and form and output early warning relationship pairs when the similarity stability reaches the standard;

[0059] Among them, the stability calculation and early warning module includes a long-term behavior analysis unit and a monitoring and early warning unit;

[0060] The long-term behavior analysis unit calculates the similarity and stability of creative behaviors between creative subjects based on the similarity of creative styles;

[0061] The monitoring and early warning unit is used to preset a similarity stability threshold. If the similarity stability is greater than or equal to the similarity stability threshold, an early warning relationship pair is formed between the creative subjects and the early warning relationship pair is output.

[0062] See also Figure 1 In the second embodiment, a method for real-time monitoring of AI creation based on big data analysis is provided, which is applicable to the first embodiment. The method includes the following steps:

[0063] Step S1: Establish a vocabulary containing semantic feature words and part-of-speech feature words, and build a creative text library to store AI-created texts in real time. Add the creation time node to the creative text and use the creator ID code as a label to generate a creative text sub-library;

[0064] For example, a vocabulary of semantic features and part-of-speech features is established, and semantic feature words and part-of-speech feature words are uniformly coded, wherein each semantic feature word and each part-of-speech feature word has a unique coding attribute, and the rth semantic feature word is recorded as SW r , record the f-th part-of-speech feature word as RW f ;

[0065] Build a text library to store AI-created texts in real time, add creation time to AI-created texts, use the creator ID code as the creation text sub-library label, and generate the creation text sub-library of the creator, recorded as CD v ={cd i (t x )|i∈[1, I], x∈[1, y]}, where, CD v Indicates the creative text sub-library generated by the creative subject coded as v, cd i (t x ) represents the addition of the xth creation time node t x The i-th creative text, I represents the total number of creative texts, and y represents the total number of creation time nodes;

[0066] For example, semantic feature words can be nouns, verbs, and adjectives, such as "technology," "innovation," and "sustainability," while part-of-speech feature words can be synonyms and antonyms, etc.

[0067] Step S2: Identify semantic feature words in the creative text, separate the semantic feature words based on the part-of-speech feature words, form part-of-speech relationship groups and attach part-of-speech tags, count the number of part-of-speech relationship groups under the same part-of-speech feature word, and construct a creative style portrait matrix of the creative subject;

[0068] For example, identifying the creative text cd i (t x) and separate the semantic feature words based on the part-of-speech feature words. If the semantic feature word SW r and semantic feature words SW m The part-of-speech relationship between them is the part-of-speech feature word RW f , then form a part-of-speech relationship group [SW r , SW m ], and the part-of-speech relationship group [SW r , SW m ]Add part-of-speech tag RW f :[SW r , SW m ], among which SW m represents the mth semantic feature word, and m≠r;

[0069] Based on part-of-speech tag RW f :[SW r , SW m ], statistical creation text cd i (t x ) in the same part of speech feature word RW f The number of part-of-speech relationship groups under i (t x )|RW f ], and construct the creative style portrait matrix of the creative subject v with the creative text coding number as the row number and the part-of-speech feature word coding as the column number, recorded as L IF (v) and the number of part-of-speech relationship groups NUM[cd i (t x )|RW f ] is mapped to the creative style portrait matrix L IF (v) The i-th row and the f-th column, where F is the total number of part-of-speech feature words.

[0070] Step S3: Setting a creation time sliding window, separating the portrait block matrix from the creation style portrait matrix based on the sliding window, and evaluating the similarity of the creation styles of different creators in the same time period;

[0071] For example, a creation time sliding window is set, and the scale value of the creation time sliding window is g. If the starting time of the qth creation time sliding window is t x , then the end time of the qth creation time sliding window is t x+g , based on the creation time sliding window T g (q), in the creation style portrait matrix L IF (v) The image block matrix is ​​stripped out and recorded as L IF (v, q), and the image block matrix L |q|F (v), where |q| represents the image block matrix L|q|F (v) The number of rows, the starting row number of the row number |q| is the starting time t x The corresponding creative text code, the end line number of line number |q| is the end time t x+g The corresponding creative text code;

[0072] Based on the portrait block matrix, evaluate the similarity of creative styles of different creative subjects at the same time stage Where, L |q|F (k) represents the image block matrix corresponding to the creative subject coded as k, Indicates that there is a block matrix L in the image |q|F (v)Number of part-of-speech relationship groups NUM[cd i (t x )|RW f ], Indicates that there is a block matrix L in the image |q|F (k)Number of part-of-speech relationship groups NUM[cd i (t x )|RW f ].

[0073] Step S4: Calculate the similarity stability of creative behaviors between creative subjects based on the similarity of creative styles, and preset a similarity stability threshold to form and output warning relationship pairs between different creative subjects;

[0074] For example, based on the similarity of creative styles, the similarity stability of creative behaviors between creative subjects is calculated. Where Q is the total number of creation time sliding windows, β is the mean value of creation style similarity, and σ 2 is the variance of creative style similarity, and

[0075] A similarity stability threshold is preset. If the similarity stability is greater than or equal to the similarity stability threshold, a warning relationship pair is formed between the creative subjects v and k, and the warning relationship pair is output.

[0076] 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," "comprising," 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.

[0077] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A real-time monitoring method for AI creation based on big data analysis, characterized in that: The method comprises the following steps: Step S1: Establish a vocabulary containing semantic feature words and part-of-speech feature words, and build a creative text library to store AI-created texts in real time. Add the creation time node to the creative text and use the creator ID code as a label to generate a creative text sub-library; Step S2: Identify semantic feature words in the creative text, separate the semantic feature words based on the part-of-speech feature words, form part-of-speech relationship groups and attach part-of-speech tags, count the number of part-of-speech relationship groups under the same part-of-speech feature word, and construct a creative style portrait matrix of the creative subject; Step S3: Setting a creation time sliding window, separating the portrait block matrix from the creation style portrait matrix based on the sliding window, and evaluating the similarity of the creation styles of different creators in the same time period; Step S4: Calculate the similarity stability of creative behaviors between creative subjects based on the similarity of creative styles, and preset a similarity stability threshold to form and output a warning relationship pair between different creative subjects.

2. The method for real-time monitoring of AI creation based on big data analysis according to claim 1 is characterized in that: The specific implementation process of step S1 includes: Establish a vocabulary of semantic features and part-of-speech features, and uniformly encode the semantic feature words and part-of-speech feature words respectively, wherein each semantic feature word and each part-of-speech feature word has a unique encoding attribute, and the rth semantic feature word is recorded as SW r , record the f-th part-of-speech feature word as RW f ; Build a text library to store AI-created texts in real time, add creation time to the AI-created texts, encode the creation subject ID as the creation text sub-library tag, and generate the creation subject's creation text sub-library, recorded as CD v ={cd i (t x )|i∈[1, I], x∈[1, y]}, where, CD v Indicates the creative text sub-library generated by the creative subject coded as v, cd i (t x ) represents the addition of the xth creation time node t x The i-th creative text, I represents the total number of creative texts, and y represents the total number of creation time nodes.

3. The method for real-time monitoring of AI creation based on big data analysis according to claim 2 is characterized in that: The specific implementation process of step S2 includes: Identify creative text cd i (t x ) and separate the semantic feature words based on the part-of-speech feature words. If the semantic feature word SW r and semantic feature words SW m The part-of-speech relationship between them is the part-of-speech feature word RW f , then form a part-of-speech relationship group [SW r , SW m ], and the part-of-speech relationship group [SW r , SW m ]Add part-of-speech tag RW f :[SW r , SW m ], among which SW m represents the mth semantic feature word, and m≠r; Based on part-of-speech tag RW f :[SW r , SW m ], statistical creation text cd i (t x ) in the same part of speech feature word RW f The number of part-of-speech relationship groups under i (t x )|RW f ], and construct the creative style portrait matrix of the creative subject v with the creative text coding number as the row number and the part-of-speech feature word coding as the column number, recorded as L IF (v) and the number of part-of-speech relationship groups NUM[cd i (t x )|RW f ] is mapped to the creative style portrait matrix L IF (v) The i-th row and the f-th column, where F is the total number of part-of-speech feature words.

4. The method for real-time monitoring of AI creation based on big data analysis according to claim 3 is characterized in that: The specific implementation process of step S3 includes: Set a creation time sliding window, the scale value of the creation time sliding window is g, if the starting time of the qth creation time sliding window is t x , then the end time of the qth creation time sliding window is t x+g , based on the creation time sliding window T g (q), in the creation style portrait matrix L IF (v) The image block matrix is ​​stripped out and recorded as L IF (v, q), and the image block matrix L |q|F (v), where |q| represents the image block matrix L |q|F (v) The number of rows, the starting row number of the row number |q| is the starting time t x The corresponding creative text code, the end line number of line number |q| is the end time t x+g The corresponding creative text code; Based on the portrait block matrix, evaluate the similarity of creative styles of different creative subjects at the same time stage Where, L |q|F (k) represents the image block matrix corresponding to the creative subject coded as k, Indicates that there is a block matrix L in the image |q|F (v) The number of part-of-speech relationship groups NUM[cd i (t x )|RW f ], Indicates that there is a block matrix L in the image |q|F (k)Number of part-of-speech relationship groups NUM[cd i (t x )|RW f ].

5. The method for real-time monitoring of AI creation based on big data analysis according to claim 4 is characterized in that: The specific implementation process of step S4 includes: Based on the similarity of creative styles, calculate the similarity stability of creative behaviors between creative subjects Where Q is the total number of creation time sliding windows, β is the mean value of creation style similarity, and σ 2 is the variance of creative style similarity, and A similarity stability threshold is preset. If the similarity stability is greater than or equal to the similarity stability threshold, a warning relationship pair is formed between the creative subjects v and k, and the warning relationship pair is output.

6. A real-time monitoring system for AI creation based on big data analysis, which executes the real-time monitoring method for AI creation according to any one of claims 1 to 5, characterized in that: The system includes: a text storage management module, a feature recognition and matrix construction module, a sliding window analysis and similarity evaluation module, and a stability calculation and early warning module; The text storage management module is used to establish a vocabulary and encode words, build a text library to store AI-created texts, add creation time to the texts and use the creator ID code as a label to generate a creation text sub-library; The feature recognition and matrix construction module is used to identify semantic feature words in the creative text, separate semantic feature words based on part-of-speech feature words, form part-of-speech relationship groups and attach part-of-speech tags, count the number of part-of-speech relationship groups, and construct a creative style portrait matrix; The sliding window analysis and similarity assessment module is used to set a creation time sliding window, separate the portrait block matrix from the creation style portrait matrix, and evaluate the similarity of the creation styles of different creators in the same time period; The stability calculation and warning module is used to calculate the similarity stability of creative behaviors between creative subjects based on the similarity of creative styles, preset a similarity stability threshold, and form and output a warning relationship pair when the similarity stability reaches the standard.

7. The AI ​​creation real-time monitoring system based on big data analysis according to claim 6 is characterized in that: The text storage management module includes a vocabulary configuration unit and a text real-time recording unit; The word library configuration unit is used to establish a word library of semantic features and part-of-speech features, and to uniformly encode the semantic feature words and part-of-speech feature words respectively; The real-time text recording unit is used to build a text library to store AI-created texts in real time, and to add the creation time to the AI-created texts, and to encode the creation subject ID as a creation text sub-library label to generate the creation subject's creation text sub-library.

8. The AI ​​creation real-time monitoring system based on big data analysis according to claim 6 is characterized in that: The feature recognition and matrix construction module includes a part-of-speech relationship group recognition unit and a creative style portrait processing unit; The part-of-speech relationship group identification unit is used to identify each semantic feature word in the creative text, and based on the part-of-speech feature words, separate each semantic feature word to form a part-of-speech relationship group and add a part-of-speech tag; The creative style portrait processing unit counts the number of part-of-speech relationship groups under the same part-of-speech feature word in the creative text as matrix elements based on the part-of-speech tags to form a creative style portrait matrix.

9. The AI ​​creation real-time monitoring system based on big data analysis according to claim 6 is characterized in that: The sliding window analysis and similarity assessment module includes a portrait segmentation unit and a creative style assessment unit; The portrait block unit is used to set a creation time sliding window to separate the portrait block matrix from the creation style portrait matrix; The creative style evaluation unit evaluates the similarity of the creative styles of different creative subjects in the same time period based on the portrait block matrix.

10. The AI ​​creation real-time monitoring system based on big data analysis according to claim 6 is characterized in that: The stability calculation and early warning module includes a long-term behavior analysis unit and a monitoring and early warning unit; The long-term behavior analysis unit calculates the similarity and stability of the creative behaviors between the creative subjects based on the similarity of the creative styles; The monitoring and early warning unit is used to preset a similarity stability threshold. If the similarity stability is greater than or equal to the similarity stability threshold, an early warning relationship pair is formed between the creative subjects and the early warning relationship pair is output.

Images