Data content supervision system and method based on digital media

Abstract

The invention discloses a data content supervision system and method based on digital media, and belongs to the field of data content supervision, and the method comprises the steps: traversing propagation paths of digital contents among digital media platforms; generating a digital content propagation behavior label based on the digital content set and the propagation path; identifying feature data of the digital content, analyzing feature values of the feature data, and quantifying feature value differences; analyzing and evaluating the spreading degree based on the digital content spreading behavior; based on the characteristic value difference and the propagation degree, whether the digital content has a propagation risk during diffusion is judged; according to the method, the propagation path and the multi-dimensional feature data of the digital content are collected and analyzed, so that the propagation path of the content is tracked, and the features of the content are quantified; an early warning mechanism is designed based on the spreading degree and the characteristic value difference, potential abnormal spreading behaviors are effectively monitored and found in time, the content management efficiency is improved, the risk prevention capability is enhanced, and safe operation of a digital media platform is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of data content supervision, and specifically provides a data content supervision system and method based on digital media. Background Art

[0002] With the rapid development of the Internet and social media, digital media platforms have become the main channels for users to create, share, and disseminate various types of content; these contents cover various forms such as text, pictures, audio, and video, greatly enriching the ways of information dissemination; however, the wide spread of digital content has also brought a large number of supervision problems, such as false information, infringing content, and bad information, which seriously affect the ecological health of digital media and the user experience; traditional content supervision methods mainly rely on manual review and keyword filtering, but in the face of massive data and complex dissemination paths, these methods are inefficient and difficult to meet the needs of modern digital media platforms; therefore, the research on digital content supervision technology has gradually become an important topic in the academic and industrial fields.

[0003] Existing content feature extraction methods are usually too simple, relying only on basic text or multimedia features, and unable to comprehensively reflect the complexity and multi-dimensional attributes of the content; this method is prone to misjudgment and missed judgment, especially when dealing with digital media containing various forms of content; secondly, the analysis of the content dissemination path in the existing technology is relatively rough, often ignoring the feature changes and the differences in dissemination behaviors of digital content when it spreads between different platforms; in addition, the setting of the early warning mechanism is usually relatively static, lacking real-time monitoring and feedback on the dynamic dissemination process, and it is difficult to effectively respond to potential risks in content dissemination. Summary of the Invention

[0004] The purpose of the present invention is to provide a data content supervision system and method based on digital media to solve the problems raised in the above background art.

[0005] To solve the above technical problems, the present invention provides the following technical solutions:

[0006] A method for supervising data content based on digital media, the method comprising the following steps: retrieving digital content presented on digital media platforms, taking the digital media platforms as data collection centers, sorting out the digital content to generate a digital content set; traversing the dissemination paths of the digital content among the digital media platforms and marking the dissemination paths; generating digital content dissemination behavior tags based on the digital content set and the dissemination paths; identifying the characteristic data of the digital content, including text characteristic data, picture characteristic data, audio characteristic data and video characteristic data, analyzing the characteristic values of each characteristic data, and quantifying the difference in characteristic values of the same digital content among different digital media platforms; analyzing and evaluating the dissemination degree of the same digital content diffused among different digital media platforms based on the digital content dissemination behavior; judging whether there is a dissemination risk when the same digital content diffuses among different digital media platforms based on the difference in characteristic values and the dissemination degree, and if there is a dissemination risk, issuing a warning prompt, otherwise not issuing a warning prompt.

[0007] As a preferred solution of the method for supervising data content based on digital media according to the present invention, a digital media platform database is constructed, denoted as DMP = {DMP a |a ∈ [1, A]}, where DMP a represents the a-th digital media platform, and A represents the total number of digital media platforms; the digital media platform allows users to create, share and disseminate digital content, and the digital content includes text, pictures, audio and video.

[0008] Uniformly number the digital content created, shared and disseminated by users in the digital media platform to construct a digital content set NR = {NR i |, i ∈ [1, I]}, where NR i represents the i-th digital content, and I represents the total number of digital content.

[0009] As a preferred solution of the method for supervising data content based on digital media according to the present invention, collect the digital content dissemination paths of the digital content NR i among different digital media platforms, denoted as The digital content dissemination path represents the path formed during the process of the digital content NR i from the initial release to being shared and forwarded at least once.

[0010] Based on the digital content set NR = {NR i |, i ∈ [1, I]} and the digital content dissemination path construct a digital content dissemination behavior tag, denoted as

[0011] As a preferred solution of the data content supervision method based on digital media described in the present invention, digital content NR is collected i in the digital media platform DMP a The digital content feature data are successively expressed as text feature data TFV, picture feature data IFV, audio feature data AFV, and video feature data VFV.

[0012] Calculate the eigenvalue TFV of the text feature data i = α1·Length i + α2·TFIDF i , where Length i represents the text length data of the text feature data, and TFIDF i represents the word frequency data of the text feature data. α1 and α2 are preset weights of the text feature data.

[0013] Calculate the eigenvalue IFV of the picture feature data i = β1·TResolution i + β2·EdgeDensity i + β3·Texture i , where TResolution i represents the picture resolution data of the picture feature data, EdgeDensity i represents the edge density data of the picture feature data, and Texture i represents the texture data of the picture feature data. β1, β2, and β3 are preset weights of the picture feature data.

[0014] Calculate the eigenvalue AFV of the audio feature data i = γ1·YDuration i + γ2·SpectralCentroid i + γ3·Pitch i , where YDuration i represents the audio duration data of the audio feature data, and Pitch i represents the pitch data of the audio feature data, and SpectralCentroid i represents the spectrum data of the audio feature data. γ1, γ2, and γ3 are preset weights of the audio feature data.

[0015] Calculate the eigenvalue VFV of the video feature data i = δ1·FrameRate i + δ2·SResolution i + δ3·SDuration i, where FrameRate i represents the video frame rate data of the video feature data, and SResolution i represents the video resolution data of the video feature data, and SDuration i represents the video duration data of the video feature data, and δ1, δ2, and δ3 are preset weights of the video feature data.

[0016] Calculate the digital content NR i The overall feature value of the digital content in the digital media platform DMP a is calculated by the formula: where and are preset adjustment parameters.

[0017] Calculate the digital content NR i in the digital media platform DMP a to the digital media platform DMP b The difference in the overall feature value of the digital content in is calculated by the formula: ΔF[NR i , (DMP a →DMP b )] = |F(NR i , DMP a ) - F(NR i , DMP b )|, where F(NR i , DMP b ) represents the overall feature value of the digital content NR i in the digital media platform DMP b .

[0018] In the present invention, by calculating the difference in the overall feature value of the digital content NR i in the digital media platform DMP a to the digital media platform DMP b , the change degree of the digital content NR i on different digital media platforms can be quantified, reflecting the consistency of the digital content NR i , helping to identify possible content tampering or illegal modification, and providing a basis for subsequent calculation of the dissemination degree and early warning value.

[0019] Based on the digital content dissemination behavior label G, calculate the dissemination degree of the digital content NR i in the digital media platform DMP a to the digital media platform DMP b by the following formula:

[0020]

[0021] Among them, PW[NR i ,(DMP a →DMP b )] represents the spread, and Power(DMP b ) represents the influence of the digital media platform DMP (based on the number of platform users, activity, and content publication volume) after statistics b of, and Power(DMP a ) represents the influence of the digital media platform DMP (based on the number of platform users, activity, and content publication volume) after statistics a of, and F(NR i ,DMP b ) represents the overall characteristic value of the digital content NR i in the digital media platform DMP b .

[0022] In the present invention, by calculating the spread of the digital content NR i between different digital media platforms, the importance of the content during the dissemination process is quantified; the formula combines the characteristic value of the digital content NR i with the influence of the digital media platform, reflecting the importance and potential influence of the digital content NR i when moving from the digital media platform DMP a to the digital media platform DMP b ; this formula helps to evaluate the degree of change of the digital content NR i when spreading between different platforms and provides a reference basis for content supervision. If the spread is very high and the content characteristics have changed significantly, it may be necessary to conduct more stringent reviews or take corresponding supervision measures for the content.

[0023] As a preferred solution of a data content supervision method based on digital media according to the present invention, based on the spread PW[NR i ,(DMP a →DMP b )] and the difference ΔF[NR i ,(DMP a →DMP b )] of the overall characteristic value of the digital content, an early warning value is designed, and the calculation formula is as follows:

[0024] PV[NR i ,(DMP a →DMP b )] = ΔF[NR i ,(DMP a →DMP b)]·PW[NR i ,(DMP a →DMP b )];

[0025] Among them, PV[NR i ,(DMP a →DMP b )] represents the warning value.

[0026] The preset warning threshold τ.

[0027] If PV[NR i ,(DMP a →DMP b )]>τ, then it is determined that the digital content NR i has changed during the propagation from the digital media platform DMP a to the digital media platform DMP b .

[0028] A data content supervision system based on digital media, the system includes: a digital platform construction module, a propagation path collection module, a content feature analysis module, and a propagation warning monitoring module;

[0029] The digital platform construction module retrieves the digital content displayed on the digital media platform, takes the digital media platform as the data storage center, and organizes the digital content to generate a digital content set.

[0030] The propagation path collection module traverses the propagation paths of the digital content between the digital media platforms and marks the propagation paths; based on the digital content set and the propagation paths, digital content propagation behavior tags are generated.

[0031] The content feature analysis module identifies the feature data of the digital content, including text feature data, picture feature data, audio feature data, and video feature data, analyzes the feature values of each feature data, and quantifies the difference in feature values of the same digital content between different digital media platforms; based on the digital content propagation behavior, analyzes and evaluates the propagation degree of the same digital content spreading between different digital media platforms.

[0032] The propagation warning monitoring module determines whether there is a propagation risk when the same digital content spreads between different digital media platforms based on the feature value difference and the propagation degree. If there is a propagation risk, a warning prompt is issued, otherwise no warning prompt is issued.

[0033] Furthermore, the digital platform construction module further includes a platform database unit and a content set construction unit.

[0034] The platform database unit constructs a digital media platform database, denoted as DMP = {DMP a |a ∈ [1, A]}, where DMP a represents the a-th digital media platform, and A represents the total number of digital media platforms; the digital media platform allows users to create, share, and disseminate digital content, and the digital content includes text, pictures, audio, and video.

[0035] The content set construction unit uniformly numbers the digital content created, shared, and disseminated by users in the digital media platform, and constructs a digital content set NR = {NR i |, i ∈ [1, I]}, where NR i represents the i-th digital content, and I represents the total number of digital content.

[0036] Furthermore, the dissemination path collection module further includes a path data collection unit and a dissemination behavior tagging unit.

[0037] The path data collection unit collects the digital content dissemination paths of the digital content NR i between different digital media platforms, denoted as The digital content dissemination path represents the path formed by the digital content NR i from the initial release to being shared and forwarded at least once.

[0038] The dissemination behavior tagging unit constructs digital content dissemination behavior tags based on the digital content set NR = {NR i |, i ∈ [1, I]} and the digital content dissemination path denoted as

[0039] Furthermore, the content feature analysis module further includes a feature data collection unit and a feature value calculation unit.

[0040] The feature data collection unit collects the digital content feature data of the digital content NR i in the digital media platform DMP a , which are successively represented as text feature data TFV, picture feature data IFV, audio feature data AFV, and video feature data VFV.

[0041] The feature value calculation unit calculates the feature value of the text feature data TFV i = α1·Length i + α2·TFIDF i , where Length i represents the text length data of the text feature data, and TFIDF iThe word frequency data representing the text feature data, where α1 and α2 are the preset weights of the text feature data.

[0042] Calculate the eigenvalue IFV of the image feature data i = β1·TResolution i + β2·EdgeDensity i + β3·Texture i , where TResolution i represents the image resolution data of the image feature data, EdgeDensity i represents the edge density data of the image feature data, Texture i represents the texture data of the image feature data, and β1, β2, and β3 are the preset weights of the image feature data.

[0043] Calculate the eigenvalue AFV of the audio feature data i = γ1·YDuration i + γ2·SpectralCentroid i + γ3·Pitch i , where YDuration i represents the audio duration data of the audio feature data, Pitch i represents the pitch data of the audio feature data, SpectralCentroid i represents the spectral data of the audio feature data, and γ1, γ2, and γ3 are the preset weights of the audio feature data.

[0044] Calculate the eigenvalue VFV of the video feature data i = δ1·FrameRate i + δ2·SResolution i + δ3·SDuration i , where FrameRate i represents the video frame rate data of the video feature data, SResolution i represents the video resolution data of the video feature data, SDuration i represents the video duration data of the video feature data, and δ1, δ2, and δ3 are the preset weights of the video feature data.

[0045] Calculate the digital content NR i The overall eigenvalue of the digital content in the digital media platform DMP a ; Calculate the digital content NR i in the digital media platform DMP a to the digital media platform DMPb The difference in the overall characteristic values of the digital content

[0046] Based on the digital content dissemination behavior label G, calculate the digital content NR i On the digital media platform DMP a To the digital media platform DMP b Degree of dissemination.

[0047] Furthermore, the dissemination early warning monitoring module further includes an early warning value calculation unit and a change judgment unit.

[0048] The early warning value calculation unit designs an early warning value based on the dissemination degree PW[NR i , (DMP a →DMP b )] and the difference ΔF[NR in the overall characteristic values of the digital content i , (DMP a →DMP b )].

[0049] The change judgment unit presets an early warning threshold τ.

[0050] If PV[NR i , (DMP a →DMP b )] > τ, then it is determined that the digital content NR i On the digital media platform DMP a To the digital media platform DMP b Has changed during dissemination.

[0051] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: By constructing a digital media platform database, uniformly numbering and managing the digital content created, shared, and disseminated by users, the efficient supervision of content is systematically realized. By collecting and analyzing the dissemination paths of digital content and their multi-dimensional characteristic data, each piece of content is labeled with a dissemination behavior label, thereby accurately tracking the dissemination trajectory of the content and quantifying its characteristics. The early warning mechanism designed based on the dissemination degree and the difference in content characteristic values effectively monitors and timely discovers potential abnormal dissemination behaviors, and finally achieves the beneficial effects of improving content management efficiency, enhancing risk prevention capabilities, and ensuring the safe operation of the digital media platform. Description of the Drawings

[0052] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0053] Wherein:

[0054] Figure 1 is a schematic flow chart of a data content supervision method based on digital media according to an embodiment of the present invention;

[0055] Figure 2 is a schematic structural diagram of a data content supervision system based on digital media according to the present invention. Specific embodiments

[0056] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0057] Please refer to Figure 1 - Figure 2 , the present invention provides a technical solution:

[0058] Please refer to Figure 1 , in the first embodiment: A data content supervision method based on digital media is provided, and the method includes the following steps:

[0059] Step S1: Retrieve the digital content displayed on the digital media platform, and take the digital media platform as the data collection center to organize the digital content to generate a digital content set.

[0060] Specifically, a digital media platform database is constructed, denoted as DMP = {DMP a |a ∈ [1, A]}, where DMP a represents the a-th digital media platform, and A represents the total number of digital media platforms; the digital media platform allows users to create, share, and disseminate digital content, and the digital content includes text, pictures, audio, and video.

[0061] The digital content created, shared, and disseminated by users in the digital media platform is uniformly numbered to construct a digital content set NR = {NR i |, i ∈ [1, I]}, where NR i represents the i-th digital content, and I represents the total number of digital content.

[0062] Step S2: Traverse the dissemination paths of the digital content among the digital media platforms and mark the dissemination paths; based on the digital content set and the dissemination paths, generate digital content dissemination behavior tags.

[0063] Specifically, collect the digital content NR iThe digital content dissemination path between different digital media platforms is denoted as The digital content dissemination path represents the digital content NR i The path formed during the process from the initial release to being shared and forwarded at least once.

[0064] Based on the digital content set NR = {NR i |, i ∈ [1, I]} and the digital content dissemination path Construct digital content dissemination behavior tags, denoted as

[0065] Step S3: Identify the feature data of digital content, including text feature data, picture feature data, audio feature data, and video feature data, analyze the feature values of each feature data, and quantify the difference in feature values of the same digital content between different digital media platforms; Based on the digital content dissemination behavior, analyze and evaluate the dissemination degree of the same digital content spreading between different digital media platforms.

[0066] Specifically, collect the digital content NR i The digital content feature data on the digital media platform DMP a Are successively represented as text feature data TFV, picture feature data IFV, audio feature data AFV, and video feature data VFV.

[0067] Calculate the feature value TFV of the text feature data i = α1·Length i + α2·TFIDF i , where, Length i Represents the text length data of the text feature data, TFIDF i Represents the word frequency data of the text feature data, and α1 and α2 are preset weights of the text feature data.

[0068] Calculate the feature value IFV of the picture feature data i = β1·TResolution i + β2·EdgeDensity i + β3·Texture i , where, TResolution i Represents the picture resolution data of the picture feature data, EdgeDensity i Represents the edge density data of the picture feature data, Texture i Represents the texture data of the picture feature data, and β1, β2, and β3 are preset weights of the picture feature data.

[0069] Calculate the feature value AFV of the audio feature datai = γ1·YDuration i + γ2·SpectralCentroid i + γ3·Pitch i , where YDuration i represents the audio duration data of the audio feature data, and Pitch i represents the pitch data of the audio feature data, and SpectralCentroid i represents the spectral data of the audio feature data. γ1, γ2, and γ3 are preset weights of the audio feature data.

[0070] Calculate the eigenvalue VFV of the video feature data i = δ1·FrameRate i + δ2·SResolution i + δ3·SDuration i , where FrameRate i represents the video frame rate data of the video feature data, and SResolution i represents the video resolution data of the video feature data, and SDuration i represents the video duration data of the video feature data. δ1, δ2, and δ3 are preset weights of the video feature data.

[0071] Furthermore, calculate the digital content NR i The overall eigenvalue of the digital content in the digital media platform DMP a is calculated by the formula: where and are preset adjustment parameters.

[0072] For example, assume that the eigenvalue of the text feature data of the digital content NR i in the digital media platform DMP a is 10, the eigenvalue of the picture feature data is 12, the eigenvalue of the audio feature data is 11, and the eigenvalue of the video feature data is 14, and are 0.1, 0.1, 0.3, and 0.5 respectively. Substituting into the formula for calculation, the overall eigenvalue of the digital content is obtained as 0.1 * 10 + 0.1 * 12 + 0.3 * 11 + 0.5 * 14 = 12.5; the digital content NR i in the digital media platform DMP bThe eigenvalue of the text feature data is 11, the eigenvalue of the picture feature data is 13, the eigenvalue of the audio feature data is 10, and the eigenvalue of the video feature data is 14. and are 0.1, 0.1, 0.3, and 0.5 respectively. Substituting into the formula for calculation, the overall eigenvalue of the digital content is 0.1 * 11 + 0.1 * 13 + 0.3 * 10 + 0.5 * 14 = 12.4.

[0073] Calculate the digital content NR i On the digital media platform DMP a to the digital media platform DMP b The difference in the overall eigenvalue of the digital content in, the calculation formula is: ΔF[NR i , (DMP a →DMP b )] = |F(NR i , DMP a ) - F(NR i , DMP b )|, where F(NR i , DMP b ) represents the overall eigenvalue of the digital content NR i on the digital media platform DMP b .

[0074] For example, based on the above calculation content, substituting into the formula, the difference in the overall eigenvalue of the digital content is 0.1.

[0075] Furthermore, based on the digital content dissemination behavior label G, calculate the dissemination degree of the digital content NR i from the digital media platform DMP a to the digital media platform DMP b , and the calculation formula is as follows:

[0076]

[0077] where PW[NR i , (DMP a →DMP b )] represents the dissemination degree, Power(DMP b ) represents the influence of the digital media platform DMP b after statistics, Power(DMP a ) represents the influence of the digital media platform DMP a , F(NR i , DMP b ) represents the overall eigenvalue of the digital content NR i on the digital media platform DMP bThe overall characteristic value of the digital content in it.

[0078] For example, assume that Power(DMP b ) is 10, and Power(DMP a ) is 12. Substituting into the formula, the spread degree is 0.83 * 12.4 = 10.3

[0079] Step S4: Based on the eigenvalue difference and the spread degree, determine whether there is a spread risk when the same digital content spreads between different digital media platforms. If there is a spread risk, issue a warning prompt; otherwise, do not issue a warning prompt.

[0080] Specifically, based on the spread degree PW[NR i ,(DMP a →DMP b )] and the difference ΔF[NR i ,(DMP a →DMP b )] of the overall characteristic value of the digital content, design a warning value. The calculation formula is as follows:

[0081] PV[NPi,(DMP a →DMP b )] = ΔF[NR i ,(DMP a →DMP b )]·PW[NR i ,(DMP a →DMP b )];

[0082] Among them, PVpNR i ,(DMP a →DMP b )] represents the warning value.

[0083] Preset a warning threshold τ.

[0084] Furthermore, if PV[NR i ,(DMP a →DMP b )] > τ, then it is judged that the digital content NR i has changed during the spread from the digital media platform DMP a to the digital media platform DMP b .

[0085] For example, assume that the preset warning threshold τ is 1. Based on the above calculation results, substituting into the formula, the warning value is 0.1 * 10.3 = 1.03. Then PV[NR i ,(DMP a →DMP b)]>τ, then judge the digital content NR i In the digital media platform DMP a To the digital media platform DMP b Has changed during the spread.

[0086] Please refer to Figure 2 In the second embodiment: Provide a data content supervision system based on digital media, the system includes: a digital platform construction module, a propagation path collection module, a content feature analysis module and a propagation early warning monitoring module;

[0087] The digital platform construction module retrieves the digital content displayed on the digital media platform, takes the digital media platform as the data storage center, and organizes the digital content to generate a digital content set.

[0088] The propagation path collection module traverses the propagation paths of digital content between digital media platforms and marks the propagation paths; Based on the digital content set and the propagation path, generate digital content propagation behavior tags.

[0089] The content feature analysis module identifies the feature data of digital content, including text feature data, picture feature data, audio feature data and video feature data, analyzes the feature values of each feature data, and quantifies the difference in feature values of the same digital content between different digital media platforms; Based on the digital content propagation behavior, analyze and evaluate the propagation degree of the same digital content spreading between different digital media platforms.

[0090] The propagation early warning monitoring module judges whether there is a propagation risk when the same digital content spreads between different digital media platforms based on the feature value difference and the propagation degree. If there is a propagation risk, an early warning prompt is issued, otherwise no early warning prompt is issued.

[0091] Further, the digital platform construction module further includes a platform database unit and a content set construction unit.

[0092] The platform database unit constructs a digital media platform database, denoted as DMP = {DMP a |a ∈ [1, A]}, where DMP a Represents the a-th digital media platform, and A represents the total number of digital media platforms; The digital media platform allows users to create, share and spread digital content, and the digital content includes text, pictures, audio and video.

[0093] The content set construction unit uniformly numbers the digital content created, shared and spread by users in the digital media platform to construct a digital content set NR = {NR i |, i ∈ [1, I]}, where NRi Represents the i-th digital content, and I represents the total number of digital contents.

[0094] Furthermore, the propagation path collection module further includes a path data collection unit and a propagation behavior tagging unit.

[0095] The path data collection unit collects the digital content NR i The digital content propagation path between different digital media platforms, denoted as The digital content propagation path represents the digital content NR i The path formed during the process from the initial release to being shared and forwarded at least once.

[0096] The propagation behavior tagging unit, based on the digital content set NR = {NR i |, i ∈ [1, I]} and the digital content propagation path Constructs a digital content propagation behavior tag, denoted as

[0097] Furthermore, the content feature analysis module further includes a feature data collection unit and a feature value calculation unit.

[0098] The feature data collection unit collects the digital content NR i The digital content feature data on the digital media platform DMP a Are successively represented as text feature data TFV, image feature data IFV, audio feature data AFV, and video feature data VFV.

[0099] The feature value calculation unit calculates the feature value TFV of the text feature data i = α1·Length i + α2·TFIDF i , where Length i Represents the text length data of the text feature data, and TFIDF i Represents the word frequency data of the text feature data, and α1 and α2 are preset weights of the text feature data.

[0100] Calculates the feature value IFV of the image feature data i = β1·TResolution i + β2·EdgeDensity i + β3·Texture i , where TResolution i Represents the image resolution data of the image feature data, EdgeDensity i Represents the edge density data of the image feature data, Texturei The texture data representing the picture feature data, where β1, β2, and β3 are preset weights of the picture feature data.

[0101] Calculate the eigenvalue AFV of the audio feature data i = γ1·YDuration i + γ2·SpectralCentroid i + γ3·Pitch i , where YDuration i represents the audio duration data of the audio feature data, Pitch i represents the pitch data of the audio feature data, and SpectralCentroid i represents the spectral data of the audio feature data, where γ1, γ2, and γ3 are preset weights of the audio feature data.

[0102] Calculate the eigenvalue VFV of the video feature data i = δ1·FrameRate i + δ2·SResolution i + δ3·SDuration i , where FrameRate i represents the video frame rate data of the video feature data, SResolution i represents the video resolution data of the video feature data, and SDuration i represents the video duration data of the video feature data, where δ1, δ2, and δ3 are preset weights of the video feature data.

[0103] Calculate the digital content NR i The overall eigenvalue of the digital content in the digital media platform DMP a ; Calculate the digital content NR i in the digital media platform DMP a to the digital media platform DMP b The difference in the overall eigenvalue of the digital content.

[0104] Based on the digital content dissemination behavior label G, calculate the digital content NR i in the digital media platform DMP a to the digital media platform DMP b Dissemination degree.

[0105] Furthermore, the dissemination early warning monitoring module further includes an early warning value calculation unit and a change judgment unit.

[0106] The early warning value calculation unit, based on the dissemination degree PW[NR i , (DMPa → DMP b )] and the difference ΔF[NR in the overall characteristic value of the digital content i , (DMP a → DMP b )], design the warning value.

[0107] The change judgment unit preset the warning threshold τ.

[0108] If PV[NR i , (DMP a → DMP b )] > τ, then it is judged that the digital content NR i has changed during the dissemination from the digital media platform DMP a to the digital media platform DMP b .

[0109] It should be noted that in this article, relational terms such as first and second are only used 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 term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0110] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A data content supervision method based on digital media, characterized in that: The method comprises the following steps: Step S1: Retrieving digital content displayed on a digital media platform, using the digital media platform as a data storage center, and organizing the digital content to generate a digital content collection; Step S2: traverse the propagation paths of digital content between various digital media platforms and mark the propagation paths; generate digital content propagation behavior tags based on the digital content set and the propagation paths; Step S3: Identify the characteristic data of the digital content, including text characteristic data, image characteristic data, audio characteristic data and video characteristic data, analyze the characteristic value of each characteristic data, and quantify the difference in characteristic values ​​of the same digital content between different digital media platforms; based on the digital content dissemination behavior, analyze and evaluate the degree of dissemination of the same digital content between different digital media platforms; Step S4: Based on the difference in characteristic values ​​and the degree of propagation, determine whether there is a propagation risk when the same digital content spreads between different digital media platforms. If there is a propagation risk, issue a warning prompt, otherwise do not issue a warning prompt.

2. A method for monitoring data content based on digital media according to claim 1, characterized in that: The step S1 comprises the following steps: Construct a digital media platform database, denoted as DMP = {DMP a |a∈[1,A]}, where DMP a represents the ath digital media platform, and A represents the total number of digital media platforms; the digital media platform allows users to create, share and disseminate digital content, and the digital content includes text, pictures, audio and video; The digital content created, shared and disseminated by users in the digital media platform is uniformly numbered to construct a digital content set NR = {NR i |, i∈[1,I]}, where NR i represents the i-th digital content, and I represents the total number of digital contents.

3. A method for monitoring data content based on digital media according to claim 2, characterized in that: The step S2 comprises the following steps: Collect the digital content NR i The digital content dissemination path between different digital media platforms is denoted as The digital content propagation path represents the digital content NR i The path from the initial post to being shared or forwarded at least once; Based on the digital content set NR={NR i |,i∈[1,I]} and the digital content dissemination path Construct digital content dissemination behavior labels, denoted as 4. A method for monitoring data content based on digital media according to claim 3, characterized in that: The step S3 comprises the following steps: Collect digital content NR i In the digital media platform DMP a The digital content feature data is represented in sequence as text feature data TFV, picture feature data IFV, audio feature data AFV and video feature data VFV; Calculate the eigenvalue TFV of text feature data i =α1·Length i +α2·TFIDF i , where Length i Text length data representing text feature data, TFIDF i The word frequency data representing the text feature data, α1 and α2 are the preset text feature data weights; Calculate the eigenvalue IFV of the image feature data i =β1·TResolution i +β2·EdgeDebsity i +β3·Texture i , where TResolution i Image resolution data representing image feature data, EdgeDensity i Represents edge density data of image feature data, Texture i The texture data representing the image feature data, β1, β2 and β3 are preset weights of the image feature data; Calculate the eigenvalue AFV of audio feature data i =γ1·YDuration i +γ2·SpectralCentroid i +γ3·Pitch i , where YDuration i Audio duration data representing audio feature data, Pitch i Tonal data representing audio feature data, SpectralCentroid i The frequency spectrum data representing the audio feature data, γ1, γ2 and γ3 are preset audio feature data weights; Calculate the feature value VFV of the video feature data i =δ1·FrameRate i +δ2·SResolution i +δ3·SDuration i , where FrameRate i Video frame rate data representing video feature data, SResolution i Video resolution data representing video feature data, SDuration i The video duration data representing the video feature data, δ1, δ2 and δ3 are preset video feature data weights; Calculate digital content NR i In the digital media platform DMP a The overall characteristic value of the digital content in is calculated as follows: in, as well as is the preset adjustment parameter; Calculate digital content NR i In the digital media platform DMP a To the digital media platform DMP b The overall characteristic value difference of digital content in is calculated as: ΔF[NR i , (DMP a →DMP b )]=|F(NR i , DMP a )-F(NR i , DMP b )|, where F(NR i , DMP b ) indicates digital content NR i In the digital media platform DMP b The overall characteristic value of the digital content in ; Based on the digital content dissemination behavior label G, the digital content NR is calculated. i In the digital media platform DMP a To the digital media platform DMP b The propagation degree is calculated as follows: Among them, PW[NR i , DMP a →DMP b )] represents the propagation degree, Power (DMP b ) represents the digital media platform DMP after statistics b Influence, Power(DMP a ) represents the digital media platform DMP after statistics a The influence of F(NR i , DMP b ) indicates digital content NR i In the digital media platform DMP b The overall characteristic value of the digital content in .

5. A method for monitoring data content based on digital media according to claim 4, characterized in that: The step S4 comprises the following steps: Based on the propagation degree PW[NR i , (DMP a →DMP b )] and the overall characteristic value difference ΔF[NR i , (DMP a →DMP b )], design the warning value, and the calculation formula is as follows: PV[NR i ,(DMP a →DMP b )]=ΔF[NR i ,(DMP a →DMP b )]·PW[NR i ,(DMP a →DMP b )]; Among them, PV[NR i , (DMP a →DMP b )] indicates the warning value; Preset warning threshold τ; If PV[NR i , (DMP a →DMP b )]>τ, then the digital content NR i In the digital media platform DMP a To the digital media platform DMP b changes have occurred in the spread of 6. A digital media-based data content supervision system, executing a digital media-based data content supervision method as claimed in any one of claims 1 to 5, characterized in that: The system includes: a digital platform construction module, a communication path collection module, a content feature analysis module, and a communication early warning monitoring module; The digital platform construction module retrieves the digital content displayed on the digital media platform, organizes the digital content using the digital media platform as a data storage center to generate a digital content collection; The propagation path collection module traverses the propagation paths of digital content between various digital media platforms and marks the propagation paths; based on the digital content collection and the propagation paths, generates digital content propagation behavior tags; The content feature analysis module identifies feature data of digital content, including text feature data, image feature data, audio feature data and video feature data, analyzes feature values ​​of each feature data, and quantifies the difference in feature values ​​of the same digital content between different digital media platforms; based on the digital content dissemination behavior, analyzes and evaluates the degree of dissemination of the same digital content between different digital media platforms; The propagation warning monitoring module determines whether there is a propagation risk when the same digital content spreads between different digital media platforms based on the characteristic value difference and the degree of propagation. If there is a propagation risk, a warning prompt is issued, otherwise no warning prompt is issued.

7. A digital media-based data content monitoring system according to claim 6, characterized in that: The digital platform construction module also includes a platform database unit and a content collection construction unit; The platform database unit constructs a digital media platform database, denoted as DMP={DMP a |a∈[1,A]}, where DMP a represents the ath digital media platform, and A represents the total number of digital media platforms; the digital media platform allows users to create, share and disseminate digital content, and the digital content includes text, pictures, audio and video; The content set construction unit uniformly numbers the digital contents created, shared and disseminated by users in the digital media platform, and constructs a digital content set NR={NR i |,i∈[1,I]}, where NR i represents the i-th digital content, and I represents the total number of digital contents.

8. A digital media-based data content monitoring system according to claim 7, characterized in that: The propagation path collection module also includes a path data collection unit and a propagation behavior label unit; The path data collection unit collects the digital content NR i The digital content dissemination path between different digital media platforms is denoted as The digital content propagation path represents the digital content NR i The path from the initial post to being shared or forwarded at least once; The propagation behavior label unit is based on the digital content set NR={NR i |,i∈[1,I]} and the digital content dissemination path Construct digital content dissemination behavior labels, denoted as 9. A digital media-based data content monitoring system according to claim 8, characterized in that: The content feature analysis module also includes a feature data collection unit and a feature value calculation unit; The feature data collection unit collects the digital content NR i In the digital media platform DMP a The digital content feature data is represented in sequence as text feature data TFV, picture feature data IFV, audio feature data AFV and video feature data VFV; The feature value calculation unit calculates the feature value TFV of the text feature data i =α1·Length i +α2·TFIDF i , where Length i Text length data representing text feature data, TFIDF i The word frequency data representing the text feature data, αx and α2 are the preset text feature data weights; Calculate the eigenvalue IFV of the image feature data i =β1·TResolution i +β2·EdgeDensity i +β3·Texture i , where TResolution i Image resolution data representing image feature data, EdgeDensity i Represents edge density data of image feature data, Texture i The texture data representing the image feature data, β1, β2 and β3 are preset weights of the image feature data; Calculate the eigenvalue AFV of audio feature data i =γ1·YDuration i +γ2·SpectralCentroid i +γ3·Pitch i , where YDuration i Audio duration data representing audio feature data, Pitch i Tonal data representing audio feature data, SpectralCentroid i The frequency spectrum data representing the audio feature data, γ1, γ2 and γ3 are preset audio feature data weights; Calculate the feature value VFV of the video feature data i =δ1·FrameRate i +δ2·SResolution i +δ3·SDuration i , where FrameRate i Video frame rate data representing video feature data, SResolution i Video resolution data representing video feature data, SDuration i The video duration data representing the video feature data, δ1, δ2 and δ3 are preset video feature data weights; Calculate digital content NR i In the digital media platform DMP a The overall characteristic value of the digital content in; Calculate the digital content NR i In the digital media platform DMP a To the digital media platform DMP b The overall characteristic value difference of digital content in ; Based on the digital content dissemination behavior label G, the digital content NR is calculated. i In the digital media platform DMP a To the digital media platform DMP b degree of spread.

10. A digital media-based data content monitoring system according to claim 9, characterized in that: The propagation warning monitoring module also includes a warning value calculation unit and a change judgment unit; The warning value calculation unit is based on the propagation degree PW[NR i , (DMP a →DMP b )] and the overall characteristic value difference ΔF[NR i , (DMP a →DMP b )], design warning value; The change judgment unit presets a warning threshold τ; If PV[NR i , (DMP a →DMP b )]>τ, then the digital content NR i In the digital media platform DMP a To the digital media platform DMP b changes have occurred in the spread of

Images