Internet new media data feedback generation method and system

Abstract

The invention relates to the field of data feedback analysis, in particular to an internet new media data feedback generation method and system. The method comprises the following steps: obtaining a new media video in production and big data fed back by Internet users; performing multi-period decomposition and video segment-by-segment deep semantic analysis on the new media video, and constructing a plurality of content label video segments; according to the internet user feedback big data, social communication effect mining and video communication effect conversion rate calculation are carried out to generate a social communication conversion rate of each video segment; and performing positive and negative propagation effect inference according to the social propagation conversion rate, and performing propagation conversion value evaluation on the plurality of content tag video segments to generate the propagation conversion value of each video segment. According to the invention, dynamic content optimization is carried out on new media data based on user demand and interest feedback, and the quality and propagation effect of video content are extracted.

Description

Technical Field

[0001] The present invention relates to the field of data feedback analysis, and in particular to a method and system for generating Internet new media data feedback. Background Art

[0002] With the rapid development of the internet and the continuous advancement of information technology, new media has become a vital vehicle for information dissemination in modern society. Whether on social platforms, video sharing sites, or mobile applications, the generation and dissemination of new media content has profoundly impacted people's daily lives, work, and entertainment. Driven by technologies such as big data, artificial intelligence, and machine learning, the generation and optimization of new media content has become increasingly intelligent and precise. The key to optimizing new media content generation lies in effectively capturing audience needs and analyzing their behavior to create high-quality content that better aligns with their interests and needs, thereby enhancing content dissemination effectiveness and audience engagement.

[0003] However, with the surge in new media content and the increasing diversity of audience demands, content creators face a significant challenge: how to stand out from the overwhelming amount of information and ensure their content is accepted and enjoyed by audiences. Traditional new media content generation methods often rely on the creator's experience and intuition, lacking real-time analysis and feedback on audience behavior. This approach is not only inefficient but also prone to mismatches between content and audience needs, making it impossible to achieve refined personalized recommendations and optimization.

[0004] Currently, new media content optimization methods primarily rely on conventional data analysis tools, such as basic metrics like viewing time, likes, and comments. However, these traditional metrics often fail to deeply tap into users' emotional needs and personalized preferences. Furthermore, existing content generation and optimization methods often rely on manual intervention or periodic analysis and inspection, making it difficult to capture and respond to dynamic changes in audience needs in real time, nor can they predict and optimize the dissemination effect of content in real time. Consequently, traditional methods are unable to cope with complex user behaviors, personalized needs, and changing audience preferences. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention proposes a method and system for generating Internet new media data feedback to solve at least one of the above technical problems.

[0006] To achieve the above object, the present invention provides a method for generating Internet new media data feedback, comprising the following steps: Step S1: Obtaining new media videos in production and internet user feedback big data; performing multi-time segment decomposition and deep semantic analysis on each video segment of the new media videos to construct multiple content-labeled video segments; Step S2: mining the social communication effect and calculating the video communication effect conversion rate based on the internet user feedback big data to generate the social communication conversion rate of each video segment; Step S3: Inferring positive and negative communication effects based on the social communication conversion rate, and evaluating the communication conversion value of multiple content-tagged video segments to generate the communication conversion value of each video segment; Step S4: Quantitative analysis of user interest and emotional color is performed based on the big data of Internet user feedback to generate personalized user emotional feedback information for each video segment; Step S5: conducting in-depth content demand mining based on the communication conversion value and the emotional feedback information, and identifying elements to be optimized to generate video elements to be optimized for each video segment; Step S6: Dynamic content optimization is performed based on the video elements to be optimized, and global coordinated fitting is performed to construct a feedback-optimized new media video.

[0007] By breaking down videos into multiple time segments and performing semantic analysis on each segment, this method ensures in-depth analysis of the key content elements of each video segment. Emotional turning points, key dialogues, and plot climaxes can all be accurately identified and extracted. This process not only annotates video content with basic tags (such as theme, characters, and locations), but also includes more detailed tags such as emotional tone and key plot points. This multi-tag system provides a foundation for subsequent personalized recommendations and content optimization. Through deep semantic analysis and tagging, customized data support is provided for each video segment, facilitating subsequent audience behavior analysis and communication effectiveness evaluation, ensuring that video content precisely meets the needs of different audiences. By analyzing user feedback big data, the social communication effect of each video segment can be quantified, including reach, sharing frequency, and discussion volume. This provides video producers with specific communication effectiveness data, helping them assess which content has the greatest potential for virality. By calculating the viral conversion rate of each video segment, it is possible to identify which segments have a high virality effect among viewers and which have failed to resonate effectively. This helps content creators understand audience preferences and provides a reference for subsequent content creation. By calculating the conversion rate of virality, content creators can adjust their content distribution and promotion strategies in real time, focusing on promoting high-performing video segments and improving overall content effectiveness. Analyzing social virality conversion rates can infer which video segments have generated positive social impact (e.g., increased engagement and brand recognition) and which have potentially generated negative effects (e.g., sparking controversy or dissatisfaction). This identification of positive and negative effects facilitates content optimization and adjustment. Evaluating the virality conversion value of each video segment assesses the virality and potential influence of different segments. This helps content creators or advertisers precisely select which segments to prioritize for promotion or optimization, thereby increasing overall virality. Analyzing virality conversion value provides content creators with data support, helping them understand which content elements have the highest audience appeal and viral potential, providing a basis for decision-making when optimizing video content. Quantifying user interests and sentiment analysis can translate audience emotional needs and viewing preferences into actionable data, enabling content creators to create personalized content for different audience groups, enhancing viewing experience and engagement. Emotional analysis of each video segment identifies the varying emotional responses viewers experience while watching. One segment might elicit joy, while another might cause anxiety. Using this emotional feedback, creators can adjust the emotional tone of their video content to better align with viewers' emotional needs. By analyzing user interests and emotions, the platform can provide personalized recommendations. If a viewer has a preference for a certain emotional tone or topic, relevant content can be pushed through precise data analysis, enhancing user engagement and satisfaction.By combining conversion value and emotional feedback, we can deeply explore users' underlying needs. For example, some users may prefer certain plot developments or specific emotional dynamics. Leveraging these insights, creators can incorporate more elements that align with user needs into their content. By analyzing the needs of each video segment, we can clearly identify elements requiring optimization (such as pacing, emotional tone, and storyline). This process helps refine optimization targets and ensure that the optimized content precisely meets the audience's emotional and interest needs. By deeply exploring content needs, we can ensure that each video segment is more refined and precise in terms of emotion and plot, thereby increasing audience engagement and retention. Based on the elements to be optimized, content can be dynamically adjusted to better suit user needs and interests. This dynamic adjustment responds to viewer feedback in real time, ensuring continuous optimization of video content. When adjusting video elements, we must also consider overall coordination and logic to avoid any inconsistencies or incoherence in the content details. This global optimization ensures the coherence and integrity of the video content, enhancing the user viewing experience. Through global optimization and feedback adjustment, a feedback-optimized video that has been optimized and adjusted multiple times is finally generated. It can improve the overall content quality and dissemination effect while meeting user needs. This optimization method enhances the market competitiveness and dissemination effect of video content.

[0008] In this specification, a system for generating feedback of new Internet media data is provided, which is used to execute the above-mentioned method for generating feedback of new Internet media data, including: A semantic parsing module is used to obtain big data on new media videos in production and internet user feedback; decompose the new media videos into multiple time periods and conduct in-depth semantic analysis of each video segment to construct multiple content-labeled video segments; A communication conversion rate module is used to mine the social communication effect and calculate the video communication effect conversion rate based on the Internet user feedback big data to generate the social communication conversion rate of each video segment; The communication value assessment module is used to infer positive and negative communication effects based on social communication conversion rates, and to evaluate the communication conversion value of multiple content-tagged video segments to generate the communication conversion value of each video segment; The emotional feedback analysis module is used to quantify user interest and analyze emotional color based on big data of Internet user feedback to generate personalized user emotional feedback information for each video segment; A content demand mining module is used to conduct in-depth content demand mining based on the communication conversion value and the emotional feedback information, and identify elements to be optimized to generate video elements to be optimized for each video segment; The dynamic content tuning module is used to perform dynamic content tuning according to the video elements to be optimized, and perform global coordinated fitting to construct feedback optimized new media videos.

[0009] By breaking down videos into multiple segments and performing in-depth semantic analysis, this invention enables detailed annotation of the content of each segment, ensuring that subsequent analysis and optimization are based on accurate data. This labeled, multi-dimensional data helps platforms better deliver personalized recommendations and precise content push, improving user viewing satisfaction. It also provides reliable foundational data (such as the sentiment, theme, and key plot points of the video content) for subsequent modules, ensuring data accuracy and consistency throughout the optimization process. By calculating the communication effect conversion rate of each video segment, creators are provided with quantitative data on the communication effect of each segment. This allows content creators to accurately assess which segments have high communication potential and which require further optimization. Based on the communication conversion rate, creators can selectively optimize segments with poor communication effects, thereby improving the overall communication effect of the content. This provides data support for creators and marketers, enabling them to determine advertising investment, content promotion, and other strategies based on communication effects, further enhancing the social communication influence of content. By evaluating the communication conversion value of each video segment, creators can determine which segments are popular and have a positive impact, and which segments fail to attract audiences. This provides guidance for subsequent content optimization. The high communication value of positive communication effects can provide feedback to creators, indicating which plots or elements are successful. Conversely, negative communication effects can prompt creators which content may have problems and need to be adjusted or optimized.

[0010] By assessing the value of content conversion, creators can more precisely adjust content structure to maximize its reach and audience engagement. Through sentiment analysis and interest quantification, creators can more clearly understand audience responses to different segments of their videos. This helps them identify which content resonates strongly with viewers and which fails to resonate as expected. Based on the emotional feedback from different audience groups, creators can adjust video content to better meet users' emotional needs. Some users may prefer more lighthearted and enjoyable content, while others may prefer more challenging or in-depth content. By better understanding user emotional responses, creators can create more engaging content, thereby increasing audience engagement, interaction frequency, and content sharing rates. Through in-depth demand analysis, creators can identify the content elements that are most important to their audiences and focus on optimizing these elements. By identifying the elements to be optimized, creators can apply improvement measures to specific video segments, ensuring that the optimizations directly address the core needs of viewers, improving content enjoyment and satisfaction. In-depth demand analysis helps content creators understand market trends and audience preferences, thereby improving the market adaptability and competitiveness of their content. Dynamic content optimization responds to user feedback and data analysis in real time, ensuring creators can make necessary optimizations before and after video release, maintaining high-quality content across multiple stages. Through global coordination and fitting, the video's emotional tone, rhythm, and other aspects are balanced, enhancing the audience's viewing experience. Precise content adjustments not only enhance the video's viewing experience but also increase user emotional resonance, thereby promoting word-of-mouth and social media dissemination. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 This is a schematic flow chart of the steps of a method for generating Internet new media data feedback according to the present invention; Figure 2 Detailed implementation flow chart of step S1; Figure 3 Detailed implementation flow chart of step S2; Figure 4 Schematic diagram of the detailed implementation steps of step S3. DETAILED DESCRIPTION

[0012] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0013] This application provides a method and system for generating feedback on new media data on the Internet. The execution entities of the method and system include, but are not limited to, mechanical equipment, data processing platforms, cloud server nodes, network upload devices, etc. that are equipped with the system, which can be regarded as general computing nodes of this application. The data processing platform includes, but is not limited to, at least one of an audio and image management system, an information management system, and a cloud data management system.

[0014] See also Figures 1 to 4 The present invention provides a method for generating feedback of new Internet media data, which comprises the following steps: Step S1: Obtaining new media videos in production and internet user feedback big data; performing multi-time segment decomposition and deep semantic analysis on each video segment of the new media videos to construct multiple content-labeled video segments; Step S2: mining the social communication effect and calculating the video communication effect conversion rate based on the internet user feedback big data to generate the social communication conversion rate of each video segment; Step S3: Inferring positive and negative communication effects based on the social communication conversion rate, and evaluating the communication conversion value of multiple content-tagged video segments to generate the communication conversion value of each video segment; Step S4: Quantitative analysis of user interest and emotional color is performed based on the big data of Internet user feedback to generate personalized user emotional feedback information for each video segment; Step S5: conducting in-depth content demand mining based on the communication conversion value and the emotional feedback information, and identifying elements to be optimized to generate video elements to be optimized for each video segment; Step S6: Dynamic content optimization is performed based on the video elements to be optimized, and global coordinated fitting is performed to construct a feedback-optimized new media video.

[0015] By breaking down videos into multiple time segments and performing semantic analysis on each segment, this method ensures in-depth analysis of the key content elements of each video segment. Emotional turning points, key dialogues, and plot climaxes can all be accurately identified and extracted. This process not only annotates video content with basic tags (such as theme, characters, and locations), but also includes more detailed tags such as emotional tone and key plot points. This multi-tag system provides a foundation for subsequent personalized recommendations and content optimization. Through deep semantic analysis and tagging, customized data support is provided for each video segment, facilitating subsequent audience behavior analysis and communication effectiveness evaluation, ensuring that video content precisely meets the needs of different audiences. By analyzing user feedback big data, the social communication effect of each video segment can be quantified, including reach, sharing frequency, and discussion volume. This provides video producers with specific communication effectiveness data, helping them assess which content has the greatest potential for virality. By calculating the viral conversion rate of each video segment, it is possible to identify which segments have a high virality effect among viewers and which have failed to resonate effectively. This helps content creators understand audience preferences and provides a reference for subsequent content creation. By calculating the conversion rate of virality, content creators can adjust their content distribution and promotion strategies in real time, focusing on promoting high-performing video segments and improving overall content effectiveness. Analyzing social virality conversion rates can infer which video segments have generated positive social impact (e.g., increased engagement and brand recognition) and which have potentially generated negative effects (e.g., sparking controversy or dissatisfaction). This identification of positive and negative effects facilitates content optimization and adjustment. Evaluating the virality conversion value of each video segment assesses the virality and potential influence of different segments. This helps content creators or advertisers precisely select which segments to prioritize for promotion or optimization, thereby increasing overall virality. Analyzing virality conversion value provides content creators with data support, helping them understand which content elements have the highest audience appeal and viral potential, providing a basis for decision-making when optimizing video content. Quantifying user interests and sentiment analysis can translate audience emotional needs and viewing preferences into actionable data, enabling content creators to create personalized content for different audience groups, enhancing viewing experience and engagement. Emotional analysis of each video segment identifies the varying emotional responses viewers experience while watching. One segment might elicit joy, while another might cause anxiety. Using this emotional feedback, creators can adjust the emotional tone of their video content to better align with viewers' emotional needs. By analyzing user interests and emotions, the platform can provide personalized recommendations. If a viewer has a preference for a certain emotional tone or topic, relevant content can be pushed through precise data analysis, enhancing user engagement and satisfaction.By combining conversion value and emotional feedback, we can deeply explore users' underlying needs. For example, some users may prefer certain plot developments or specific emotional dynamics. Leveraging these insights, creators can incorporate more elements that align with user needs into their content. By analyzing the needs of each video segment, we can clearly identify elements requiring optimization (such as pacing, emotional tone, and storyline). This process helps refine optimization targets and ensure that the optimized content precisely meets the audience's emotional and interest needs. By deeply exploring content needs, we can ensure that each video segment is more refined and precise in terms of emotion and plot, thereby increasing audience engagement and retention. Based on the elements to be optimized, content can be dynamically adjusted to better suit user needs and interests. This dynamic adjustment responds to viewer feedback in real time, ensuring continuous optimization of video content. When adjusting video elements, we must also consider overall coordination and logic to avoid any inconsistencies or incoherence in the content details. This global optimization ensures the coherence and integrity of the video content, enhancing the user viewing experience. Through global optimization and feedback adjustment, a feedback-optimized video that has been optimized and adjusted multiple times is finally generated. It can improve the overall content quality and dissemination effect while meeting user needs. This optimization method enhances the market competitiveness and dissemination effect of video content.

[0016] In the embodiment of the present invention, see Figure 1 , is a schematic flow chart of the steps of a method for generating Internet new media data feedback according to the present invention. In this example, the steps of the method for generating Internet new media data feedback include: Step S1: Obtaining new media videos in production and internet user feedback big data; performing multi-time segment decomposition and deep semantic analysis on each video segment of the new media videos to construct multiple content-labeled video segments; In this embodiment, after obtaining authorization from the user and the relevant platform, the source of the new media videos to be analyzed is identified. These can include proprietary platforms, social media, video sharing websites, and so on. The video's theme and target audience are determined for subsequent analysis. If the goal is to analyze introductory videos about technology products, relevant videos can be selected from platforms such as YouTube and Bilibili. Appropriate tools are used to capture videos, ensuring high quality and completeness. This can be accomplished through APIs, crawler technology, or direct download tools. Assuming 10 videos are selected, each between 5 and 10 minutes long, ensure that the captured content is representative and covers different time periods and themes. The captured videos are organized, ensuring that basic information (such as title, duration, and uploader) is recorded for each video. Furthermore, the video files need to be stored in a designated folder for subsequent processing. An Excel spreadsheet can be created to record each video's title, duration, and download link for easy retrieval. Determine which channels to collect user feedback data from. Feedback data can come from comment sections, social media sharing and discussion, and likes and dislikes data. This feedback helps understand audience perceptions of the video content. You can analyze comments under YouTube videos or access relevant user discussions through social media platform APIs. Use data collection tools (such as crawlers or APIs) to extract user feedback data. Ensure that each user feedback entry includes relevant information, such as user ID, feedback content, timestamp, number of likes, and number of replies. Assuming 200 user comments are extracted from each video, record this data and store it in a database or Excel spreadsheet, ensuring it is structured for later analysis. Cleanse the collected user feedback data to remove invalid comments (such as advertisements, spam, or comments without content) to ensure the quality and reliability of the remaining data. If 400 invalid comments are removed from 2,000 comments, the remaining 1,600 valid comments will be used for subsequent analysis. Break each video into multiple time periods and determine appropriate segmentation points. Segments can be based on natural changes in the video content, such as scene changes, topic shifts, or emotional fluctuations. For an 8-minute video, you might divide it into four 2-minute segments, ensuring that each segment conveys information independently. Use video analysis tools (such as Adobe Premiere Pro or FFmpeg) to segment the video. Use timeline positioning to ensure the start and end times of each segment are accurate. For example, consider a video segment divided into the following sections: Segment 1 (0:00-2:00), Segment 2 (2:00-4:00), Segment 3 (4:00-6:00), and Segment 4 (6:00-8:00). Ensure that these segments cover different key points. Record information for each segment, ensuring it includes the duration, theme, and content summary. This will provide a foundation for subsequent in-depth semantic analysis.Segment 1 is titled "Product Introduction," Segment 2 is titled "User Feedback," Segment 3 is titled "Market Comparison," and Segment 4 is titled "Summary and Outlook." Appropriate deep semantic analysis tools and technologies can be used, such as natural language processing (NLP) models (such as BERT or the GPT series) to analyze text content and extract key themes, sentiment, and related semantic features from the video segments. NLP models can be used to analyze user comments and extract key sentiment and keywords. Deep semantic analysis is performed on each video segment to extract relevant keywords, sentiment, and potential user needs. This can be achieved through methods such as keyword extraction and sentiment analysis. Analysis of the text content in Segment 1 identifies keywords such as "innovation," "user experience," and "market demand," and assesses the sentiment as "positive." The deep semantic analysis results for each video segment are organized into structured data for subsequent analysis. This dataset will contain the theme, sentiment score, and key semantic features of each segment. The semantic features of Segment 1 are: "Theme: Innovation, Sentiment Score: 8 / 10, Keywords: User Experience, Market Demand."

[0017] Step S2: mining the social communication effect and calculating the video communication effect conversion rate based on the internet user feedback big data to generate the social communication conversion rate of each video segment; In this example, data collected from user feedback, such as comments, likes, shares, and reposts, is integrated. This ensures that the relevant social communication data for each video segment matches the corresponding user feedback information, providing a comprehensive perspective for subsequent analysis. Assume that feedback from 1,000 users is collected for each video segment, including 500 comments, 300 likes, and 200 shares. This data is recorded in a structured database to facilitate subsequent processing. Relevant indicators of social communication effects are defined and calculated, including the number of comments, likes, and shares. These indicators can be used to evaluate the dissemination effect of each video segment on social platforms. If segment 1 has 150 comments, 200 likes, and 100 shares, the social communication effect indicators for this segment can be calculated to assess its popularity among users. Social network analysis (SNA) methods are used to evaluate user interaction patterns and dissemination paths. By analyzing the interactive relationships between users, key nodes of content dissemination and influential users can be identified. Use a network visualization tool (such as Gephi or Cytoscape) to visualize the network of user comments and shares, making it easier to identify the users and content with the most significant virality. Organize the results of the social virality analysis into a report, ensuring that the virality metrics for each video segment are recorded and interpreted. This will provide important evidence for the subsequent calculation of virality conversion rates. Record the virality of segment 1 as "Number of comments: 150, Number of likes: 200, Number of shares: 100," and analyze the positive audience response. Determine the method for calculating the virality conversion rate of the video. Generally, the conversion rate can be defined as the ratio of valid views generated through social sharing and user interaction to the total number of views. If a video segment has 2,000 total views and 500 views generated through sharing and interaction, the conversion rate formula is: Conversion rate = (Number of views generated / Total views) × 100%. Calculate the conversion rate for each video segment based on the integrated social virality data and number of views. Ensure accurate statistical data is used in the calculation and account for potential errors. If segment 1 has a total of 2,000 views, 300 views due to sharing, and 200 views due to likes, the conversion rate for segment 1 is: Conversion rate = (500 / 2,000) × 100% = 25%. Record the calculated conversion rates for each video segment in the database and visualize them. Use bar charts or pie charts to intuitively understand the conversion effect of each segment. Use data visualization tools to plot the conversion rates for each segment, demonstrating the differences in the effectiveness of different segments and facilitating subsequent optimization decisions.

[0018] Step S3: Inferring positive and negative communication effects based on the social communication conversion rate, and evaluating the communication conversion value of multiple content-tagged video segments to generate the communication conversion value of each video segment;In this embodiment, criteria for evaluating positive and negative communication effects are determined. Positive communication effects are generally associated with positive user feedback (such as likes, positive comments, and shares), while negative communication effects are associated with negative user feedback (such as negative reviews and negative comments). A comprehensive evaluation model is established by analyzing user feedback data. If a video segment receives significantly more likes than dislikes and the comments contain a large number of positive sentiment words, it can be determined that the segment has a positive communication effect. Sentiment analysis is performed on user comments for each video segment, using natural language processing techniques to identify emotional tendencies. Sentiment analysis tools such as VADER or TextBlob can be used to score comments and assess their emotional tone. For example, if 80% of the comments on a video segment are positive, 10% are neutral, and 10% are negative, it can be inferred that the segment has a significant positive communication effect. A comprehensive evaluation of each video segment is performed by combining the social communication conversion rate with the results of the sentiment analysis. A weighted model is used to combine the conversion rate with the positive and negative communication effects to derive a comprehensive effect score. If the conversion rate for segment 1 is 25% and the proportion of positive sentiment is 80%, the overall communication effect score can be calculated as: Overall score = Conversion rate × Positive sentiment ratio = 25% × 0.8 = 20%. Organize the inferred positive and negative communication effects for each video segment into a report to ensure a clear description of the communication effects of each segment. Record the positive and negative effects of each segment and their corresponding overall scores. Define a calculation method for communication conversion value. Positive communication effects, negative communication effects, and social communication conversion rates can typically be combined to form a comprehensive communication conversion value assessment model. Communication conversion value can be defined as: Communication conversion value = (Positive communication effect - Negative communication effect) × Social communication conversion rate. Based on the inferred positive and negative communication effects and calculated social communication conversion rates in the previous steps, calculate the communication conversion value for each video segment. Ensure accurate metrics and data are used in the calculation process. Assuming that segment 1 has a positive effect score of 15%, a negative effect score of 5%, and a social conversion rate of 25%, the communication conversion value = (15% - 5%) × 25% = 2.5%. The calculated conversion value of each video segment is recorded in the database and visualized. Bar charts or pie charts can be used to intuitively understand the conversion value of different video segments. Using data visualization tools, the conversion value of each segment is plotted into a chart, showing the differences in the dissemination effect of different video segments, facilitating subsequent optimization decisions. The conversion value of each video segment is analyzed to identify high- and low-performing segments and explore possible reasons. This analysis will provide guidance for content optimization and help creators understand which content is more likely to trigger user interaction and sharing.If the communication conversion value of segment 1 is 2.5, while the conversion value of segment 3 is only 0.5, you can analyze the content and communication strategy of segment 1, extract the success factors, and make corresponding adjustments in segment 3.

[0019] Step S4: Quantitative analysis of user interest and emotional color is performed based on the big data of Internet user feedback to generate personalized user emotional feedback information for each video segment; In this embodiment, relevant data collected from user feedback is integrated, including comment content, number of likes, number of shares, and viewing time. This data provides a basis for quantifying user interest. Assuming a video segment has 300 user comments, 150 likes, 50 shares, and a total viewing time of 2000 seconds, this data will help us understand the user's level of interest in that segment. The various indicators used to quantify user interest are determined. The number of comments, likes, shares, and viewing time can be combined to form a comprehensive interest scoring model. Each indicator can be assigned a different weight based on its importance. Assume that the weights are defined as: number of comments accounts for 30%, number of likes accounts for 40%, number of shares accounts for 20%, and viewing time accounts for 10%. Based on these weights, the contribution of each indicator can be calculated. Based on the defined weights, the user interest score for each video segment is calculated individually. Each indicator is multiplied by its weight and the sum is calculated to obtain the comprehensive interest score. If a video has 300 comments, 150 likes, 50 shares, and a viewing time of 2000 seconds, the interest score can be calculated as: Comment contribution = 300 × 0.3 = 90 Like contribution = 150 × 0.4 = 60 Share contribution = 50 × 0.2 = 10 Viewing time contribution = (2000 / 2000) × 10 = 10 Overall interest score = 90 + 60 + 10 + 10 = 170.

[0020] Select appropriate sentiment analysis tools and techniques. Natural language processing (NLP) models are typically used to analyze the sentiment of user comments. Common tools include VADER and TextBlob, which are effective at identifying the emotional tone of comments. If positive words such as "love," "like," and "wonderful" are used in comments on a particular video, sentiment analysis tools can identify these sentiments. Each user comment is scored for sentiment, analyzing the proportion of positive, negative, and neutral sentiment within the comments. The number of comments in each sentiment category can be recorded to calculate the emotional distribution. Suppose that among the 300 comments on a video, 60% are positive, 10% are negative, and 30% are neutral. The emotional profile of this segment can be calculated as "positive: 60%, negative: 10%, neutral: 30%." Based on the sentiment analysis results, an emotional profile is generated for each video segment. This can be accomplished by calculating the proportion of different emotion types for subsequent analysis and personalized feedback. For example, the emotional profile of segment 1 is recorded as "positive: 60%, negative: 10%, neutral: 30%." This provides support for subsequent personalized emotional feedback. The emotional analysis results for each video segment are compiled into a report to ensure clarity and understanding. This report will serve as the basis for generating subsequent personalized user emotional feedback. The emotional characteristics of segment 1 are recorded and stored along with the user's interest score, providing a basis for content optimization decisions. The quantitative user interest score is combined with the emotional characteristics to generate personalized user emotional feedback for each video segment. This feedback provides content creators with specific user preferences and emotional needs. For example, if segment 1 has an interest score of 170 and an emotional characteristic of "60% positive emotion," feedback can be generated: "Users in segment 1 are highly interested in the content, with a high proportion of positive emotions. It is recommended that the current content style be maintained." This personalized user emotional feedback is recorded in a database to ensure structured and traceable information, providing a basis for subsequent content optimization and strategy adjustments. The personalized feedback for segment 1 is recorded for reference in subsequent content production, ensuring that creators can adjust according to user needs. Finally, the personalized user emotional feedback is integrated into a report to ensure that content creators can intuitively understand user needs, which will promote content optimization and improvement, and increase user satisfaction and engagement.

[0021] Step S5: conducting in-depth content demand mining based on the communication conversion value and the emotional feedback information, and identifying elements to be optimized to generate video elements to be optimized for each video segment; In this embodiment, the communication conversion value and emotional feedback information are integrated. This ensures that the communication conversion value and user emotional feedback for each video segment accurately correspond to each other, facilitating in-depth analysis. This data integration provides a solid foundation for identifying elements for optimization. Assume that segment 1 has a communication conversion value of 2.5, and the emotional feedback indicates that 60% of users have positive emotions toward the content. This information is recorded in a structured data table for subsequent analysis. Using the communication conversion value and emotional feedback information, the content demand characteristics of each video segment are analyzed. Statistical methods such as cluster analysis or factor analysis can be used to identify user preferences and demand for different content types. Users may exhibit higher positive emotions and engagement in certain segments, indicating that these content types are more popular. Assuming that the analysis reveals that users generally have higher positive sentiment scores for the content type "Product Use Cases," and that its communication conversion value is also prominent, this content type can be identified as a high-demand feature. The identified deep-level content demand characteristics are mapped to ensure that the demand characteristics of each video segment are clearly recorded. This process helps content creators understand users' true needs and provides guidance for subsequent optimization. The demand characteristic for segment 1 is "Users have a high demand for product use cases. We recommend adding relevant examples in subsequent content." Define criteria for identifying elements for optimization. This may include issues raised in user feedback, the proportion of negative sentiment in emotional feedback, and content segments with low conversion value. Clarifying these criteria will help more systematically identify elements for optimization. If a video segment has a negative sentiment ratio exceeding 20% and a conversion value below 1, it can be identified as an element for optimization. Based on the defined criteria, identify elements for optimization in each video segment. Text analysis of user feedback can identify specific content issues, such as "Information is not clear enough" or "The pacing is too slow." Assuming that user comments for segment 1 contain multiple instances of "Information is not clear enough," these can be marked as elements for optimization. Organize the identified elements for optimization, ensuring that the areas for optimization in each video segment are clearly documented. These elements may include clarity of narration, pacing of the visuals, and logical coherence of the content. The element for optimization in segment 1 is "Information is not clearly conveyed. We recommend adding examples."

[0022] Step S6: Dynamic content optimization is performed based on the video elements to be optimized, and global coordinated fitting is performed to construct a feedback-optimized new media video.

[0023] In this embodiment, a detailed analysis is performed on each video segment to identify elements to be optimized. These elements may include unclear information, inappropriate pacing, and inconsistent emotional expression. The analysis should take into account audience feedback and emotional overtones to ensure that optimization decisions effectively address user concerns. If user feedback for a particular video segment commonly mentions "information unclear," the segment's narration, accompanying images, and information presentation should be focused on. A corresponding optimization strategy is developed based on the specific elements to be optimized. For issues with unclear information, examples can be added, language can be simplified, or diagrams can be used to aid explanation. For issues with inappropriate pacing, editing points or the tempo of the background music can be adjusted. For example, if the narration in segment 1 is too complex, the creator may decide to simplify it, use more direct language, and add relevant examples to help viewers better understand the content. The optimization strategy is implemented in video editing software. Based on the developed plan, the video segment is edited and adjusted accordingly. Frame-by-frame analysis can be used to ensure that each adjustment achieves the desired effect and takes the audience's viewing experience into consideration. For segment 1, after modifying the narration text, multiple replays are performed to ensure clearer information is conveyed, and adjustments are made immediately based on feedback. If a particular example explanation is found to be too long, it can be shortened to maintain audience attention. After optimizing each video segment, conduct an overall content assessment to ensure logical coherence and emotional consistency across segments. Review the video as a whole to observe its fluidity and effectiveness in conveying information. Play the entire video to observe the natural transitions between segments and the consistency of emotional expression. If the emotional expression of a particular segment is inconsistent with the overall style, further adjustments are necessary. Based on the overall evaluation results, develop a coordinated adjustment strategy. You may need to fine-tune the emotional tone, rhythm, and narration style of certain segments to ensure the overall content is coherent and engaging. For example, if the emotional expression in segment 3 is too bland, while the preceding and following segments are full of energy, consider adding more engaging language or background music to enhance its emotional expression. During the coordination and adjustment process, promptly collect audience feedback and make necessary corrections. Conduct small-scale audience testing to understand their reactions to the optimized content and make further improvements based on their feedback. Invite a small group of target audiences to watch the optimized video and gather their feedback. If the audience generally reports an improvement in the emotional expression of the segment, this adjustment strategy can be maintained; if feedback still points to issues, further revisions will be required. After all adjustments and coordination are completed, the final version of the optimized new media video will be compiled to ensure that all changes have been implemented and the video quality meets the expected goals. Video editing software can be used to make detailed adjustments to the final version, such as optimizing the editing, sound effects, and subtitles. Ensure that the volume of each segment is consistent and that the subtitles are displayed in sync with the narration to enhance the audience's viewing experience.

[0024] In this embodiment, refer to Figure 2 , is a flowchart of the detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: Obtaining big data on new media videos in production and internet user feedback; Decomposing the new media video into multiple time periods and performing deep semantic analysis on each video segment to generate deep semantic features for each video segment; Classify content labels based on the deep semantic features to generate accurate content labels for each video segment; The new media video is subjected to temporal positioning tag fitting according to the precise content tag to construct a plurality of content tag video segments.

[0025] In this embodiment, after obtaining authorization from the user and the relevant platform, the storage location and access channels of the new media video are determined, including social media platforms, video sharing websites, etc. User feedback data such as the number of visits, likes, and comments of the relevant videos are obtained for comprehensive analysis. If a video is obtained from YouTube, the record format is "video ID, title, release date, number of visits, likes, and comments", such as "video_001,'New Media Content Analysis', 2023-01-01, 10000, 500, 200". Relevant user comments and feedback are extracted, and the user's emotional tendencies and opinions are analyzed. These feedbacks will be used as the basis for subsequent content optimization. The format for extracting user comments is "user ID, comment content, timestamp", such as "user_001,'The content is very interesting!', 2023-01-0208:00". The collected data is cleaned to remove invalid or duplicate information to ensure the accuracy and completeness of the data. The processed data is stored in the database for subsequent analysis. If a comment is found to be an invalid link, it will be marked and deleted, with a record of "Data cleaning completed, invalid comment deleted." New media videos should be broken down into time segments, typically 30 seconds or minutes, to facilitate subsequent analysis. Each video segment should be labeled with its start and end time. For a 6-minute video, the segment should be broken down into "Segment 1: 0:00-0:30, Segment 2: 0:30-1:00," and so on. Deep semantic analysis is performed on each video segment, using natural language processing (NLP) techniques to analyze the dialogue, narration, and subtitles within the video. The core themes, sentiment, and keywords within the video segment are extracted. Sentiment analysis tools are used to analyze the dialogue content of Segment 1, recording the segment as "Segment 1 deep semantic features: Theme: 'User Experience', Sentiment: Positive." The deep semantic features of each video segment are recorded in a database to facilitate subsequent label classification and analysis. The record format is "Segment number, theme, sentiment, keyword," such as "Segment 1: User Experience, Positive, 'Fun, Interactive'." A content label classification model is constructed based on the deep semantic features. Machine learning algorithms (such as support vector machines or random forests) can be used to classify video segments to ensure that the model can accurately identify different types of content. Set the label categories to "education, entertainment, technology, emotion" and train according to the features. Use the constructed label classification model to classify each video segment and output the precise content label of each segment. Ensure the accuracy and effectiveness of the classification results. If segment 1 is classified as "education", it will be recorded as "Segment 1 precise content label: education". Verify the classification results and use cross-validation and other methods to ensure the accuracy of the model. If there are inaccurate labels, adjust the model parameters. After verification, it was found that the accuracy of the education label was 90%, which was recorded as "Label classification accuracy verification completed". According to the precise content label of each video segment, the video is time-series positioned and marked to ensure that each label matches the corresponding time period.Create a precise time-series labeling framework. The recording format is "segment number, start time, end time, label", such as "segment 1, 0:00, 0:30, education". Integrate each video segment with a label to generate multiple content-labeled video segments, which can be used for subsequent content recommendations or reprocessing. The generated records are "segment 1: education, segment 2: entertainment, segment 3: technology", ensuring that each segment has a clear theme. Record all content-labeled video segments and time-series markers in the database, and generate reports for subsequent analysis and optimization of content strategies. The recording format is "video ID, segment information", such as "video_001, segment 1: education, segment 2: entertainment".

[0026] In this embodiment, the specific steps of performing multi-period decomposition of the new media video and deep semantic analysis of each video segment to generate deep semantic features of each video segment are as follows: Identifying the full video content rhythm and scene change turning points of the new media video; Calculating the optimal segmentation time point based on the content rhythm and scene change turning points, and extracting multiple video segmentation time points; Perform multi-time period decomposition based on the multiple video segmentation time points to obtain video segments of multiple time windows; Extract key frames from video segments of multiple time windows to obtain key frames for each video segment; Extracting key semantic fields and key content elements of the key frames; Performing potential association analysis on the key semantic fields and key content elements to obtain a latent semantic subject feature vector; Based on the latent semantic subject feature vector, deep semantic analysis is performed on each video segment to generate deep semantic features for each video segment.

[0027] In this embodiment, the audio and visual content in the video must first be analyzed to identify its rhythm. Audio signal processing techniques are used to extract the audio signal from the video, focusing primarily on the beat of the music and the speed of the dialogue. By analyzing the audio waveform and spectrum, changes in the audio rhythm can be determined and synchronized with changes in the video image. Using tools such as audio analysis software, the audio can be segmented and marked to identify different beats. For example, faster beats may indicate more frequent scene cuts. This method allows the identification of the video's tempo and provides a basis for subsequent scene change analysis. For scene changes in the video, visual content analysis techniques are used to detect changes in the image. By comparing the color histograms between adjacent frames or utilizing edge detection algorithms, significant change points in the image can be identified. A noticeable change in image hue, brightness, or content is considered a turning point in a scene change. By analyzing each frame of the video and marking these turning points, the reasons for their occurrence can be understood, such as whether they are due to plot development, emotional changes, or a shift in theme. After identifying beats and turning points, this information is organized into a dataset for subsequent analysis. This dataset includes the timestamp, type, and potential impact of each beat and scene transition, ensuring that subsequent analysis accurately reflects the overall structure of the video. Based on the identified content beats and scene transitions, a reasonable segmentation strategy is designed. The goal is to maintain content coherence within each video segment while avoiding cuts at key information or emotional highs. The optimal segmentation locations are determined by analyzing the content density and emotional variation of different segments. The distribution of beats and scene transitions is analyzed to calculate multiple appropriate time intervals for segmentation. Considering the length of each segment and the importance of its content, these intervals are ensured to reflect both natural content changes and audience viewing habits. If the pace accelerates and scene transitions are frequent during a particular period, a segment may be added there. The video content is then broken down based on the previously calculated segmentation times. Each segment should maintain visual and narrative consistency to ensure that the audience can understand the segment's theme and emotional content. If a segment discusses product features, it should include relevant visual information and explanation. Keyframes are extracted from each decomposed paragraph, typically using uniform sampling or content-based selection methods to ensure that representative frames are selected for each paragraph. These keyframes should reflect the core information and sentiment of the paragraph. A frame is extracted every few frames from a paragraph to form a visual summary of the paragraph. Each keyframe is deeply analyzed, primarily using image recognition and optical character recognition (OCR) technology to extract content. For keyframes containing textual information, OCR technology is used to obtain the text content and identify objects in the image to extract key semantic fields and important content elements. Information such as "product display" or "user feedback" is identified in the keyframes; these fields form the basis for subsequent analysis.In-depth association analysis is performed on the extracted key semantic fields and content elements to identify potential relationships between them. Cluster analysis or association rule learning methods can be used to identify which elements frequently appear together and which fields are semantically connected. A strong correlation was found between "user feedback" and "product features," indicating that they often appear in the same context. Based on the results of the latent association analysis, feature vectors are generated for each video segment. These feature vectors reflect the thematic and sentiment characteristics of the paragraph, facilitating subsequent deep semantic parsing. Each key semantic field is mapped to a vector and weighted according to its importance within the paragraph to form a comprehensive feature representation. Using the generated latent semantic subject feature vector, deep semantic parsing is performed on each video segment. By analyzing the key content and sentiment of each paragraph, rich semantic information is extracted, including the theme, sentiment, and important concepts. Sentiment analysis techniques are used to analyze the paragraph content, determine its sentiment as positive, negative, or neutral, and provide a comprehensive description based on the theme. The parsing results are integrated to generate deep semantic features for each video segment. These features include information such as the theme, sentiment, and keywords, providing a comprehensive understanding of the paragraph. The report identifies a paragraph's theme as "product introduction" and its sentiment as "positive," and extracts relevant keywords. The deep semantic features of each video segment are organized into a document to ensure the analysis results can be effectively utilized. These results provide important reference for subsequent content optimization and evaluation.

[0028] In this embodiment, refer to Figure 3 , is a flowchart of the detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Calculate the user viewing time, video clicks and playback volume of each video segment in the internet user feedback big data; Conducting user behavior analysis on the user's viewing time, video clicks, and playback volume to obtain user viewing behavior characteristics; Calculate the sharing volume and forwarding rate of each video segment of the internet user feedback big data; Performing user sharing behavior analysis on the sharing volume and forwarding rate to generate user sharing behavior characteristics; Mining social communication effects based on user sharing behavior characteristics and user viewing behavior characteristics to generate social communication effect characteristics; Based on the social communication effect characteristics, the video communication effect conversion rate of multiple content tag video segments is calculated one by one to generate the social communication conversion rate of each video segment.

[0029] In this example, the user viewing time, clicks, and play count for each video segment are obtained from internet user feedback data. This data is typically obtained through video platform analytics tools, which record user behavior while viewing the video, including video start and end timestamps. If a video segment has 2,000 plays and a total user viewing time of 3,000 seconds, the detailed viewing time and playback status of each segment can be recorded. For each video segment, the average user viewing time is calculated. This can be achieved by dividing the total user viewing time by the number of users who viewed the segment. This data can help us understand the audience's level of engagement with the video segment. For example, if a video segment has 1,000 viewers and a total viewing time of 3,000 seconds, the average viewing time is 3 seconds, which can reflect the appeal of the content. Clicks refer to the number of times users click to watch the video, while plays refer to the number of times the video is actually played. By comparing these two data points, the video's appeal and view-to-conversion rate can be assessed. If the clicks are 2,500 and the plays are 2,000, the view-to-conversion rate can be calculated as 80%. Based on calculated viewing time, clicks, and play counts, we extract features from user viewing behavior. We analyze user viewing habits across different video segments and identify segments with longer viewing times or higher viewing rates. For example, if the average viewing time for a particular segment is 5 seconds, while other segments are only 2 seconds, we can consider that segment to be popular.

[0030] Use statistical methods such as cluster analysis to identify patterns in user viewing behavior. Users can be categorized as high-engagement and low-engagement users, and their viewing habits for different video segments can be analyzed. By analyzing viewing frequency and interaction levels (such as comments and likes), we can further understand user preferences and habits. Obtain the number of shares and forwarding rates for each video segment from social media platforms or video sharing websites. This data typically includes the number of times a video is shared on social networks and the number of times the video is viewed via the shared link. If a video segment is shared 500 times and has 1000 views through sharing, the share-forwarding rate can be calculated. The forwarding rate is calculated by dividing the number of shares by the number of views, then multiplying by 100 to obtain a percentage. The forwarding rate is an important indicator of the effectiveness of content on social media. For example, if the number of shares is 500 and the number of views is 2000, the forwarding rate is 25%, which reflects the social dissemination effect of the video. Based on the number of shares and forwarding rate data, user sharing behavior characteristics can be extracted. Identify which video segments have high share rates and analyze the reasons. This may be related to the content's interest, information content, or emotional resonance. If a video's share rate is above 30%, it may be because the content has strong emotional resonance or is a social topic. Build user profiles based on sharing behavior. This can be achieved by analyzing users' social media behavior, sharing frequency, and preferred content types. If a user frequently shares technology videos, it can be assumed that they have a high interest in technology content. Compile the analysis results into a report that showcases the characteristics and patterns of user sharing behavior and provides content creators with actionable optimization recommendations. This report helps content creators understand which content is most likely to be shared and tailor its content creation accordingly. Analyze the social dissemination effect of video content based on user sharing and viewing behavior characteristics. This includes assessing the breadth, depth, and influence of the content. Use diffusion network analysis to understand which video segments have the best social dissemination effect and which content has sparked widespread discussion and sharing. Generate a social dissemination effect feature vector for each video segment. This vector includes features across multiple dimensions, such as share volume, forwarding rate, and viewing time. Using these vectors, the social dissemination effect of each video segment can be quantified. If a video segment has a high share count and a long view time, these dimensions are assigned a higher weight in the feature vector. Based on the characteristics of social communication effects, the communication effect conversion rate is calculated for each video segment with a content tag. The conversion rate can be calculated by comparing the effective interactions (such as comments and likes) generated by sharing and viewing with the total number of views. If a video segment has 1,000 views and 200 effective interactions through sharing, the communication conversion rate is 20%. By analyzing the conversion rates of video segments with multiple content tags, it is possible to identify which content is more likely to trigger user interaction and sharing. This analysis will directly influence the direction and strategy of future content production.If content creators find that certain topics have a high conversion rate, they can consider increasing the proportion of such topics in their future creations.

[0031] In this embodiment, refer to Figure 4 , is a flowchart of the detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include: Calculating the conversion rate difference based on the social communication conversion rate to obtain the conversion rate difference values of different video segments; Evaluate the user feedback value of social communication effect characteristics to obtain the user feedback value of each video segment; Inferring the positive and negative communication effects of the social communication conversion rate based on the user feedback value to obtain the positive and negative communication effects of each video segment; The communication conversion value is evaluated based on the positive and negative communication effects to generate the communication conversion value of each video segment.

[0032] In this embodiment, the conversion rate difference is typically calculated using basic statistical analysis methods. This can be obtained by calculating the difference between the conversion rate of each video segment and the average conversion rate of all video segments. If the average conversion rate of the three video segments is 21.67%, then the difference for segment 1 is 20% - 21.67% = -1.67%, the difference for segment 2 is 30% - 21.67% = 8.33%, and the difference for segment 3 is 15% - 21.67% = -6.67%. These difference values can reflect the relative strengths or weaknesses of each video segment in social communication relative to the overall performance. The calculated conversion rate difference values for each video segment are compiled into a report to facilitate subsequent analysis. In addition, these difference values can be visualized to provide a more intuitive understanding of the performance of each video segment. User feedback data related to each video segment is collected, including comments, likes, shares, and viewing time. This data directly reflects the user's response to the video segment content and can provide a basis for subsequent value assessment. If a video has 150 comments, 500 likes, and a viewing time of 2000 seconds, these data will serve as important indicators for assessing the value of user feedback. Based on this collected user feedback data, a feedback value assessment model can be developed. This model can consider multiple factors, such as the sentiment of the comments, the percentage of likes, and the viewing time. A weighted scoring system can be used to assign different weights to each metric to calculate the total feedback value. Assuming a weight of 0.5 for comments, 0.3 for likes, and 0.2 for viewing time, the total feedback value can be calculated as: Feedback Value = 0.5 × Number of Comments + 0.3 × Number of Likes + 0.2 × Viewing Time (adjusted to seconds). This yields a composite feedback value score. When inferring positive and negative communication effects, it's first necessary to define what constitutes positive and negative communication effects. Positive effects are typically associated with positive user feedback, sharing, and interaction, while negative effects may be associated with negative comments or low viewing time. If a video segment receives a large number of positive comments and a high share rate, it can be inferred to have a positive viral effect; conversely, it may have a negative viral effect. A positive and negative viral effect inference model is constructed based on user feedback value and social viral conversion rate. This model determines the viral effect of each video segment by analyzing the sentiment trends in user feedback (using sentiment analysis tools) and interaction data. If 80% of the feedback for a video segment is positive and its conversion rate is above average, it can be determined to have a significant positive viral effect. Based on the positive and negative viral effects, the viral conversion value of each video segment is assessed. The viral conversion value is calculated using a weighted model combining user feedback value and viral effect. If the weight of the positive viral effect is 0.7 and the weight of the negative effect is 0.3, the viral conversion value can be expressed as: viral conversion value = positive effect × 0.7 - negative effect × 0.3.By calculating the conversion value of each video segment, we can identify segments with the highest potential for conversion in social media. This analysis helps creators optimize their content strategies and focus on high-conversion-value content. If a video segment has a high conversion value and receives predominantly positive feedback, creators can consider developing follow-up content on the relevant topic.

[0033] In this embodiment, step S4 includes the following steps: Extract user comment content from multiple video segments based on big data of internet user feedback; Identify malicious comments on the user's comments and mark them as malicious; Calculate the high-frequency identical barrages of the user's barrage content; Perform abnormal barrage detection on the high-frequency identical barrages and extract abnormal high-frequency barrages; Perform barrage cleaning on the malicious barrage and abnormally high-frequency barrage to obtain valid barrage content; Analyze the user emotion color of each video segment of the effective barrage content and generate the user emotion color characteristics of each video segment; Performing a quantitative analysis of the user's interest level based on the effective barrage content to obtain a quantitative value of the user's interest level; The personalized emotion information is analyzed on the quantitative value of the user's interest level and the user's emotion color characteristics to generate the user's personalized emotion feedback information for each video segment.

[0034] In this embodiment, the bullet screen content of users is extracted from a video platform, usually obtained through an API interface or data crawler technology. Ensure that the data includes the bullet screen content of each video segment, as well as additional information such as the sending time and user ID. Suppose 1000 bullet screen contents are extracted from a certain video segment, and this data will provide a basis for subsequent analysis. Sort out the extracted bullet screen content and clean up invalid information (such as duplicate, empty or incorrectly formatted bullet screens). Associate each bullet screen with its corresponding timestamp and user ID to form structured data. Eliminate 50 invalid bullet screens from the 1000 bullet screens to ensure that the remaining 950 valid data can be used for subsequent analysis. Use natural language processing (NLP) technology to build a malicious bullet screen recognition model. Through supervised learning methods, use labeled malicious bullet screens to train the model so that it can identify bullet screens such as aggressive, insulting or spam information. Use a word vector model (such as Word2Vec) to convert the bullet screen content into vectors and train a classifier (such as SVM or random forest) for malicious bullet screen recognition. Mark the identified malicious bullet screens to ensure that valid bullet screens and malicious bullet screens can be distinguished during subsequent processing. For each bullet screen, mark it as "malicious" or "normal" and record its recognition result. If 20 bullet screens are identified as malicious, mark these bullet screens as "malicious" and annotate them in the dataset. Conduct a frequency statistics on the extracted bullet screen content to identify high-frequency identical bullet screens, which can be achieved by constructing a word frequency matrix or using a hash table to record the occurrence times of each bullet screen. Among the 950 bullet screens, if a certain bullet screen appears 150 times, it can be marked as a high-frequency bullet screen. Record the high-frequency identical bullet screens and provide data support for subsequent anomaly detection. Ensure that the recorded content includes the bullet screen text and its occurrence frequency. Record "Bullet screen content: 'It's so good-looking', occurrence times: 150". Conduct anomaly detection on high-frequency bullet screens, mainly identifying those bullet screens with abnormally high frequencies, which can be achieved by setting a threshold (such as the occurrence times exceeding twice the average of all bullet screens) to identify abnormal bullet screens. If the average occurrence times of bullet screens is 10, then consider bullet screens with occurrence times exceeding 20 as abnormal bullet screens. Extract the identified abnormal high-frequency bullet screens to ensure that these bullet screens can be further analyzed. Record the content and occurrence frequency of each abnormal bullet screen. If a certain bullet screen appears 300 times while the average is 10, record it as "Abnormal bullet screen: 'It's extremely good-looking', occurrence times: 300". Conduct cleaning on the identified malicious bullet screens and abnormal high-frequency bullet screens, eliminate these inappropriate contents, and ensure that the remaining bullet screens are valid bullet screens. Sort out the cleaning rules, such as deleting all bullet screens marked as "malicious" and bullet screens with abnormally high occurrence frequencies. After cleaning, sort out the valid bullet screen content and store it as the basic data for downstream analysis. Ensure that each valid bullet screen can reflect the true feedback of users. After cleaning, the remaining valid bullet screens are 800, and these bullet screens will be used for sentiment analysis.Sentiment analysis techniques (such as sentiment lexicon or deep learning models) are used to analyze the emotional tone of effective comments. The sentiment conveyed by each comment is identified, for example, as positive, negative, or neutral. If the analysis results show 500 positive comments, 200 negative comments, and 100 neutral comments, the emotional distribution can be clearly recorded. An emotional tone feature is generated for each video segment, typically expressed as the proportion of each emotion. This feature provides data support for subsequent personalized emotional feedback. For example, the emotional characteristics of a video segment are "positive: 62.5%, negative: 25%, neutral: 12.5%." Based on the content of effective comments and the level of user interaction, user interest in each video segment is quantified. This can be achieved by comprehensively evaluating the number of comments, their emotional orientation, and user interactions (such as likes and shares). If a video segment has a large number of comments and a majority of them are positive, the quantified user interest value for that segment can be set high. A quantitative interest level value is calculated for each video segment, typically expressed on a scale of 0 to 100, with higher values indicating greater user interest. For example, a value of 80 for a particular video segment indicates a strong user interest in the content. The user interest level value is combined with the user's emotional color characteristics to generate personalized emotional feedback information for each video segment. By associating emotional characteristics with user interest levels, a more in-depth user feedback perspective can be provided. If a video segment's emotional characteristics are positive and its interest level value is 80, then "user emotional feedback on this segment is positive and the interest level value is 80" can be generated. The personalized emotional feedback information for each video segment is organized into a report to ensure clarity and readability. This report will provide content creators with practical insights for targeted adjustments in future content production. The report lists user emotional feedback for each video segment, helping creators understand the audience's true feelings and preferences.

[0035] In this embodiment, step S5 includes the following steps: Based on the user's personalized emotional feedback information and the dissemination conversion value of each video segment, deep content demand mining is carried out to obtain the user's deep content demand characteristics for different video segments; Performing emotional tone and rhythm optimization analysis based on the user's deep-level content demand characteristics to obtain emotional tone and rhythm optimization data for each video segment; Analyze the adaptive expression optimization method based on the user's deep content demand characteristics; The elements to be optimized are identified for the adaptive expression optimization mode, emotional tone and rhythm optimization data to generate the video elements to be optimized for each video segment.

[0036] In this embodiment, personalized user emotional feedback is integrated with the communication conversion value of each video segment. User feedback includes emotional tendencies, interest levels, and interaction data, while the communication conversion value reflects the dissemination effect of the content. Assume that user feedback for a video segment indicates that viewers give it an emotional rating of 8 / 10 and a communication conversion value of 0.3. This integration forms a dataset for subsequent analysis. Deep-level content demand features are extracted from this integrated data. By analyzing the relationship between user emotional feedback and communication conversion value, the audience's deep-level demand for content is identified. Certain segments may exhibit high emotional ratings and high communication conversion values, indicating a strong user demand for these contents. Using cluster analysis, the video segments are divided into different categories, and common demand features within each category are identified. These features may include preference for specific themes, intensity of emotional resonance, and so on. Based on these deep-level user content demand features, an emotional tone and rhythm optimization analysis model is constructed. Regression analysis or machine learning algorithms can be used to analyze the relationship between user feedback and video segment content to identify the optimal emotional tone and rhythm patterns. By analyzing user emotional feedback data, we can identify users who respond more positively to specific tempos (such as fast or slow), providing data support for tempo optimization. We also analyze the emotional tone of user feedback to identify user preferences for different video segments. Using sentiment analysis tools, we can quantify user feedback and extract emotional tones (such as warmth, coolness, positivity, and negativity). If users' emotional tone preference for a particular video segment is "positive," and its emotional rating is generally higher than other segments, we can determine whether the emotional tone of that segment needs to be maintained or enhanced. Based on the results of the emotional tone analysis and user needs, we generate tempo optimization data for each video segment. This includes recommended tempo changes and potential editing points to ensure a better viewing experience. For user-preferred "fast-paced" segments, we recommend quick cuts at key points to attract viewers' attention. Based on users' deeper content needs, we analyze appropriate adaptive expression methods. This may include adjusting the voiceover tone, visual style, and background music to better align with user emotional needs. If the emotional tone of user feedback in a certain video is generally "positive", it can be recommended to use a more cheerful tone and bright images when expressing it. By evaluating the feedback on different expressions, the most popular expression method can be identified. The difference in user feedback under different expressions can be analyzed through A / B testing. If two different narration styles are used in the same paragraph, analyze the user's viewing time and emotional feedback to determine which method is more popular. Based on the adaptive expression optimization method, emotional tone and rhythm optimization data, the elements to be optimized in the video segment are identified. This can be done by analyzing the specific problems mentioned in the user feedback, such as "the rhythm is too slow" or "the emotions are not strong enough", etc., to determine the content that needs to be optimized.If users generally report that a certain paragraph has a slow tempo and the emotional tone is not positive enough, then that paragraph is an element to be optimized. The identified elements to be optimized will be systematically organized to ensure that the parts to be optimized for each video segment are clearly recorded. These elements may include narration, editing rhythm, visual style, etc. The elements to be optimized found in a certain video are recorded as "the narration tone needs to be more lively, and the editing rhythm needs to be faster." The elements to be optimized will be organized into a report to facilitate subsequent content production and optimization. Ensure that this information can provide content creators with effective improvement suggestions to improve the overall quality of the video and audience satisfaction. The report will list the video segments to be optimized and their specific problems to help creators make targeted adjustments in subsequent production.

[0037] In this embodiment, step S6 includes the following steps: Performing video position positioning on the video elements to be optimized, and extracting the video position of each video element to be optimized; Dynamically optimizing content based on the video position to generate optimized content for multiple video segments; Perform global content detail rationality analysis on the optimized content of multiple video segments and extract the rationality of the details; Identify abnormal logic based on the rationality of details and mark abnormal logic content; Locally optimize the content details of abnormal logical content, perform global coordinated fitting, and build feedback optimization new media video.

[0038] In this embodiment, a frame-by-frame analysis of the video content is performed for the video segment to be optimized. Computer vision technology is used to identify the specific locations of the elements to be optimized within the video. These elements may include narration clips, specific frames, or subtitles. If a video segment contains narration that requires optimization, the start and end timestamps of each narration are recorded for subsequent location. Video frames are extracted using video editing and analysis tools (such as FFmpeg or OpenCV). By extracting each frame and combining the timestamp information, the specific location of the element to be optimized within the timeline and frame is determined. For example, 100 frames are extracted from a 10-minute video. If the narration to be optimized starts at the 2nd minute and ends at 2 minutes and 30 seconds, its location is recorded as "2:00-2:30." The location of each element to be optimized is organized into structured data to ensure that the time and location of each element can be accurately referenced. This record provides essential information for subsequent dynamic content optimization. The narration to be optimized is recorded as "Narration location: 2:00-2:30" to facilitate quick location during subsequent adjustments. Design a dynamic content optimization strategy based on the location of the elements to be optimized. The goal of optimization is to make real-time adjustments to key elements in the video based on user feedback and content needs. For sections where the narration rhythm is slow, consider shortening the narration or speeding it up. In video editing software, dynamically adjust the elements to be optimized based on the recorded location. This can be achieved by scaling the timeline, adjusting the speed, or altering the narration's tone. For example, between 2:00 and 2:30, increase the narration speed by 20% and appropriately increase the rhythm of the background music to make the overall content more cohesive and engaging. Record the content changes of each video segment after optimization for subsequent performance analysis. Ensure that each adjustment reflects changes in user needs. Feedback on the adjusted videos shows an increase in average viewing time from 3 minutes to 4 minutes, indicating a successful optimization. Conduct a global content detail analysis of the optimized segments. Compare the original and optimized videos to assess their logical coherence, emotional consistency, and rhythmic flow. A combination of expert review and user feedback can be used to gather opinions from all sides and ensure a comprehensive analysis. During the analysis, focus on the transition effects between video segments, emotional expression, and the effectiveness of information transmission. Record the performance of each paragraph, especially its correlation with the audience's emotional feedback. If the emotional expression of a paragraph is found to be inconsistent with the preceding and following paragraphs, mark it as an unreasonable detail. Organize the results of the detail rationality analysis into a report to ensure that the content is clear and easy to understand. This will provide data support for the subsequent identification of abnormal logic. The report states that "paragraphs 3 and 4 are emotionally inconsistent and need to be adjusted." Based on the results of the organized detail rationality analysis, use logic analysis tools to identify abnormal logic in the video content. Natural language processing technology can be used to analyze the content of narration and subtitles to check whether they conform to logical reasoning.If an event mentioned in a paragraph contradicts the description in a previous paragraph, mark it as a logic anomaly. Mark identified logic anomalies to ensure quick identification of the issue during subsequent optimization. These marks will help content creators identify the areas that need to be addressed. For example, if "The character mentioned in paragraph 2 is inconsistent with that in paragraph 3," mark it as "Logic anomaly: character inconsistency." Document the results of logic anomaly identification to ensure each logic issue is effectively tracked. This documentation will provide a reference for subsequent optimization of content details. Note that "the character inconsistency issue occurs between paragraphs 2 and 3." Based on the identified logic anomalies, develop a local optimization strategy to ensure the logic is properly presented in the video. Optimization may include modifying the narration, adjusting video editing, or re-producing certain scenes. For the "character inconsistency" issue, re-recording the narration in paragraph 3 may be necessary to align it with the description in paragraph 2. Perform local optimization in video editing software to ensure the adjusted content is coherent and logical. Frame-by-frame analysis can be used to ensure that each adjustment achieves the desired effect. Incorporate the character characteristics mentioned in the previous paragraph into the narration in paragraph 3 to enhance logical coherence. After completing local optimization, verify the results. User feedback and expert review can be used to ensure that the optimized content effectively solves the problems identified previously. Collect user feedback again. If the user's understanding of the paragraph improves and the viewing time increases, it indicates that the optimization is effective. Coordinate local optimization with global content to ensure the consistency of the overall structure and emotional expression of the video. Through overall playback and review, ensure that the transition between different paragraphs is natural and smooth. Adjust the transition music and visual effects between paragraphs to make the overall viewing experience smoother. Conduct a comprehensive evaluation of the optimized new media video, collect audience feedback and ratings, and ensure that the new video can meet user expectations and needs.

[0039] In this embodiment, a system for generating feedback of new Internet media data is provided, which is used to execute the above-mentioned method for generating feedback of new Internet media data, including: A semantic parsing module is used to obtain big data on new media videos in production and internet user feedback; decompose the new media videos into multiple time periods and conduct in-depth semantic analysis of each video segment to construct multiple content-labeled video segments; A communication conversion rate module is used to mine the social communication effect and calculate the video communication effect conversion rate based on the Internet user feedback big data to generate the social communication conversion rate of each video segment; The communication value assessment module is used to infer positive and negative communication effects based on social communication conversion rates, and to evaluate the communication conversion value of multiple content-tagged video segments to generate the communication conversion value of each video segment; The emotional feedback analysis module is used to quantify user interest and analyze emotional color based on big data of Internet user feedback to generate personalized user emotional feedback information for each video segment; A content demand mining module is used to conduct in-depth content demand mining based on the communication conversion value and the emotional feedback information, and identify elements to be optimized to generate video elements to be optimized for each video segment; The dynamic content tuning module is used to perform dynamic content tuning according to the video elements to be optimized, and perform global coordinated fitting to construct feedback optimized new media videos.

[0040] By breaking down videos into multiple segments and performing in-depth semantic analysis, this invention enables detailed annotation of the content of each segment, ensuring that subsequent analysis and optimization are based on accurate data. This labeled, multi-dimensional data helps platforms better deliver personalized recommendations and precise content push, improving user viewing satisfaction. It also provides reliable foundational data (such as the sentiment, theme, and key plot points of the video content) for subsequent modules, ensuring data accuracy and consistency throughout the optimization process. By calculating the communication effect conversion rate of each video segment, creators are provided with quantitative data on the communication effect of each segment. This allows content creators to accurately assess which segments have high communication potential and which require further optimization. Based on the communication conversion rate, creators can selectively optimize segments with poor communication effects, thereby improving the overall communication effect of the content. This provides data support for creators and marketers, enabling them to determine advertising investment, content promotion, and other strategies based on communication effects, further enhancing the social communication influence of content. By evaluating the communication conversion value of each video segment, creators can determine which segments are popular and have a positive impact, and which segments fail to attract audiences. This provides guidance for subsequent content optimization. The high communication value of positive communication effects can provide feedback to creators, indicating which plots or elements are successful. Conversely, negative communication effects can prompt creators which content may have problems and need to be adjusted or optimized.

[0041] By assessing the value of content conversion, creators can more precisely adjust content structure to maximize its reach and audience engagement. Through sentiment analysis and interest quantification, creators can more clearly understand audience responses to different segments of their videos. This helps them identify which content resonates strongly with viewers and which fails to resonate as expected. Based on the emotional feedback from different audience groups, creators can adjust video content to better meet users' emotional needs. Some users may prefer more lighthearted and enjoyable content, while others may prefer more challenging or in-depth content. By better understanding user emotional responses, creators can create more engaging content, thereby increasing audience engagement, interaction frequency, and content sharing rates. Through in-depth demand analysis, creators can identify the content elements that are most important to their audiences and focus on optimizing these elements. By identifying the elements to be optimized, creators can apply improvement measures to specific video segments, ensuring that the optimizations directly address the core needs of viewers, improving content enjoyment and satisfaction. In-depth demand analysis helps content creators understand market trends and audience preferences, thereby improving the market adaptability and competitiveness of their content. Dynamic content optimization responds to user feedback and data analysis in real time, ensuring creators can make necessary optimizations before and after video release, maintaining high-quality content across multiple stages. Through global coordination and fitting, the video's emotional tone, rhythm, and other aspects are balanced, enhancing the audience's viewing experience. Precise content adjustments not only enhance the video's viewing experience but also increase user emotional resonance, thereby promoting word-of-mouth and social media dissemination.

[0042] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0043] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for generating Internet new media data feedback, characterized in that: The following steps are involved: Step S1: Obtaining new media videos in production and internet user feedback big data; performing multi-time segment decomposition and deep semantic analysis on each video segment of the new media videos to construct multiple content-labeled video segments; Step S2: mining the social communication effect and calculating the video communication effect conversion rate based on the internet user feedback big data to generate the social communication conversion rate of each video segment; Step S3: Inferring positive and negative communication effects based on the social communication conversion rate, and evaluating the communication conversion value of multiple content-tagged video segments to generate the communication conversion value of each video segment; Step S4: Quantitative analysis of user interest and emotional color is performed based on the big data of Internet user feedback to generate personalized user emotional feedback information for each video segment; Step S5: conducting in-depth content demand mining based on the communication conversion value and the emotional feedback information, and identifying elements to be optimized to generate video elements to be optimized for each video segment; Step S6: Dynamic content optimization is performed based on the video elements to be optimized, and global coordinated fitting is performed to construct a feedback-optimized new media video.

2. The method for generating Internet new media data feedback according to claim 1, characterized in that: The specific steps of step S1 are: Obtaining big data on new media videos in production and internet user feedback; Decomposing the new media video into multiple time periods and performing deep semantic analysis on each video segment to generate deep semantic features for each video segment; Classify content labels based on the deep semantic features to generate accurate content labels for each video segment; The new media video is subjected to temporal positioning tag fitting according to the precise content tag to construct a plurality of content tag video segments.

3. The method for generating Internet new media data feedback according to claim 1, characterized in that: The specific steps of performing multi-period decomposition of the new media video and deep semantic analysis of each video segment to generate deep semantic features of each video segment are as follows: Identifying the full video content rhythm and scene change turning points of the new media video; Calculating the optimal segmentation time point based on the content rhythm and scene change turning points, and extracting multiple video segmentation time points; Perform multi-time period decomposition based on the multiple video segmentation time points to obtain video segments of multiple time windows; Extract key frames from video segments of multiple time windows to obtain key frames for each video segment; Extracting key semantic fields and key content elements of the key frames; Performing potential association analysis on the key semantic fields and key content elements to obtain a latent semantic subject feature vector; Based on the latent semantic subject feature vector, deep semantic analysis is performed on each video segment to generate deep semantic features for each video segment.

4. The method for generating Internet new media data feedback according to claim 1, characterized in that: The specific steps of step S2 are: Calculate the user viewing time, video clicks and playback volume of each video segment in the internet user feedback big data; Conducting user behavior analysis on the user's viewing time, video clicks, and playback volume to obtain user viewing behavior characteristics; Calculate the sharing volume and forwarding rate of each video segment of the internet user feedback big data; Performing user sharing behavior analysis on the sharing volume and forwarding rate to generate user sharing behavior characteristics; Mining social communication effects based on user sharing behavior characteristics and user viewing behavior characteristics to generate social communication effect characteristics; Based on the social communication effect characteristics, the video communication effect conversion rate of multiple content tag video segments is calculated one by one to generate the social communication conversion rate of each video segment.

5. The method for generating Internet new media data feedback according to claim 1, characterized in that: The specific steps of step S3 are: Calculating the conversion rate difference based on the social communication conversion rate to obtain the conversion rate difference values of different video segments; Evaluate the user feedback value of social communication effect characteristics to obtain the user feedback value of each video segment; Inferring the positive and negative communication effects of the social communication conversion rate based on the user feedback value to obtain the positive and negative communication effects of each video segment; The communication conversion value is evaluated based on the positive and negative communication effects to generate the communication conversion value of each video segment.

6. The method for generating Internet new media data feedback according to claim 1, characterized in that: The specific steps of step S4 are: Extract user comment content from multiple video segments based on big data of internet user feedback; Identify malicious comments on the user's comments and mark them as malicious; Calculate the high-frequency identical barrages of the user's barrage content; Perform abnormal barrage detection on the high-frequency identical barrages and extract abnormal high-frequency barrages; Perform barrage cleaning on the malicious barrage and abnormally high-frequency barrage to obtain valid barrage content; Analyze the user emotion color of each video segment of the effective barrage content and generate the user emotion color characteristics of each video segment; Performing a quantitative analysis of the user's interest level based on the effective barrage content to obtain a quantitative value of the user's interest level; The personalized emotion information is analyzed on the quantitative value of the user's interest level and the user's emotion color characteristics to generate the user's personalized emotion feedback information for each video segment.

7. The method for generating Internet new media data feedback according to claim 1, characterized in that: The specific steps of step S5 are: Based on the user's personalized emotional feedback information and the dissemination conversion value of each video segment, deep content demand mining is carried out to obtain the user's deep content demand characteristics for different video segments; Performing emotional tone and rhythm optimization analysis based on the user's deep-level content demand characteristics to obtain emotional tone and rhythm optimization data for each video segment; Analyze the adaptive expression optimization method based on the user's deep content demand characteristics; The elements to be optimized are identified for the adaptive expression optimization mode, emotional tone and rhythm optimization data to generate the video elements to be optimized for each video segment.

8. The method for generating Internet new media data feedback according to claim 1, characterized in that: The specific steps of step S6 are: Performing video position positioning on the video elements to be optimized, and extracting the video position of each video element to be optimized; Dynamically optimizing content based on the video position to generate optimized content for multiple video segments; Perform global content detail rationality analysis on the optimized content of multiple video segments and extract the rationality of the details; Identify abnormal logic based on the rationality of details and mark abnormal logic content; Locally optimize the content details of abnormal logical content, perform global coordinated fitting, and build feedback optimization new media video.

9. An Internet new media data feedback generation system, characterized in that: The method for generating Internet new media data feedback according to claim 1 comprises: A semantic parsing module is used to obtain big data on new media videos in production and internet user feedback; decompose the new media videos into multiple time periods and conduct in-depth semantic analysis of each video segment to construct multiple content-labeled video segments; A communication conversion rate module is used to mine the social communication effect and calculate the video communication effect conversion rate based on the Internet user feedback big data to generate the social communication conversion rate of each video segment; The communication value assessment module is used to infer positive and negative communication effects based on social communication conversion rates, and to evaluate the communication conversion value of multiple content-tagged video segments to generate the communication conversion value of each video segment; The emotional feedback analysis module is used to quantify user interest and analyze emotional color based on big data of Internet user feedback to generate personalized user emotional feedback information for each video segment; A content demand mining module is used to conduct in-depth content demand mining based on the communication conversion value and the emotional feedback information, and identify elements to be optimized to generate video elements to be optimized for each video segment; The dynamic content tuning module is used to perform dynamic content tuning according to the video elements to be optimized, and perform global coordinated fitting to construct feedback optimized new media videos.

Images