Multidimensional evaluation and optimization control system for news media communication effect

Abstract

The application relates to a news media communication effect multi-dimensional evaluation and optimization control system, which comprises a cross-platform heterogeneous data sensing layer, a three-dimensional nine-dimensional multi-dimensional evaluation layer, a cause-effect inference intelligent decision layer and a full-link closed-loop control layer. The cross-platform heterogeneous data sensing layer collects multi-dimensional communication data of mainstream platforms in the whole network; the three-dimensional nine-dimensional multi-dimensional evaluation layer generates dynamic quantitative score results based on a three-dimensional nine-dimensional evaluation model; the cause-effect inference intelligent decision layer identifies core factors influencing the communication effect and generates an individualized optimization scheme; and the full-link closed-loop control layer executes the optimization strategy and feeds back effect data, forming a self-optimization closed loop for continuous iteration. The application realizes multi-dimensional accurate evaluation and real-time dynamic control of the communication effect, effectively improves the communication effect and optimization efficiency, and is suitable for full-channel communication management of various news contents.

Description

Technical Field

[0001] This invention belongs to the field of news media technology, specifically involving a multi-dimensional evaluation and optimization control system for the effects of news media dissemination. Background Technology

[0002] In recent years, with the rapid development of internet technology, the news media industry has entered the era of multimedia communication. News content can be disseminated through multiple platforms, resulting in faster dissemination and wider coverage, but this also brings enormous challenges to the evaluation and optimization of dissemination effectiveness.

[0003] Existing news media dissemination effectiveness evaluation systems generally suffer from the following shortcomings: First, the evaluation dimensions are singular and distorted, relying solely on surface-level traffic data such as readership and likes, failing to distinguish between genuine user participation and machine-generated traffic, and struggling to quantify the depth of emotional resonance and long-term social value of the content, resulting in evaluation results that do not accurately reflect the actual dissemination effect of the news content; Second, the evaluation is severely lagging, mostly relying on post-event statistical analysis, failing to provide dynamic intervention during the golden window of dissemination, thus missing the best dissemination opportunities; Third, optimization and control rely entirely on human experience, lacking a data-driven scientific decision-making mechanism, leading to large fluctuations in optimization effects, low efficiency, and difficulty in adapting to the dissemination characteristics of different types of news content; Fourth, the problem of cross-platform data silos is prominent, making it impossible to achieve unified evaluation and collaborative optimization of dissemination effects across all channels, and hindering the formation of overall dissemination synergy.

[0004] Therefore, it is essential to provide a multi-dimensional evaluation and optimization control system for the dissemination effects of news media to address the aforementioned issues. Summary of the Invention

[0005] This invention provides a multi-dimensional evaluation and optimization control system for the dissemination effect of news media, which aims to solve the above-mentioned technical problems.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is: to provide a multi-dimensional evaluation and optimization control system for the dissemination effect of news media, including a cross-platform heterogeneous data perception layer, a three-dimensional nine-degree multi-dimensional evaluation layer, a causal inference intelligent decision-making layer, and a full-link closed-loop control layer; The cross-platform heterogeneous data perception layer is used to collect content data, user behavior data, propagation path tracing data and multi-dimensional public opinion data from mainstream platforms across the network in real time, and transmit the collected data to the three-dimensional nine-degree multi-dimensional evaluation layer. The three-dimensional nine-degree multi-dimensional evaluation layer is used to process the data transmitted by the cross-platform heterogeneous data perception layer based on the original three-dimensional nine-degree evaluation model, generate a dynamic quantitative score result of the news media dissemination effect, and transmit the dynamic quantitative score result to the causal inference intelligent decision-making layer. The causal inference intelligent decision-making layer is used to accurately identify the core factors affecting the dissemination effect of news media based on the dynamic quantitative scoring results, generate an executable personalized optimization plan, and transmit the personalized optimization plan to the full-link closed-loop control layer. The full-link closed-loop control layer is used to automatically execute the personalized optimization scheme and feed back the effect data after execution to the cross-platform heterogeneous data perception layer in real time, forming a continuously iterative self-optimizing closed loop.

[0007] Furthermore, the cross-platform heterogeneous data perception layer includes multiple platform data acquisition modules, which correspond to different mainstream news dissemination platforms and are used to collect content publishing data, user click data, user comment data, user forwarding data, user collection data, and user dwell time data on the corresponding platforms.

[0008] Furthermore, the three-dimensional nine-degree multi-dimensional evaluation layer adopts a three-dimensional nine-degree evaluation model that constructs nine quantifiable evaluation indicators from three core dimensions: the breadth of dissemination, the depth of dissemination, and the value of dissemination. The breadth of dissemination includes effective coverage, diffusion acceleration, and cross-platform penetration rate; The depth of dissemination includes average user dwell time, content re-creation rate, and emotional resonance. The dimensions of communication value include positive brand mention rate, conversion rate, and social influence index.

[0009] Furthermore, the three-dimensional nine-degree multi-dimensional evaluation layer also includes a dynamic weight adjustment module. The dynamic weight adjustment module is used to automatically identify the type of news content and adjust the weight of each evaluation indicator in the three-dimensional nine-degree evaluation model in real time according to different news content types, so as to generate a dynamic quantitative scoring result that matches the news content type.

[0010] Furthermore, the causal inference intelligent decision-making layer adopts a causal inference algorithm that combines the difference between two methods with propensity score matching to accurately identify the core factors that truly affect the dissemination effect of news media from multiple influencing factors such as title keywords, publication time, content format, and key dissemination nodes.

[0011] Furthermore, the causal inference intelligent decision-making layer also includes a counterfactual simulation framework. This framework is used to construct counterfactual scenarios corresponding to different optimization strategies based on the identified core factors, predict the potential propagation effects of different optimization strategies, and provide a scientific basis for generating personalized optimization solutions.

[0012] Furthermore, the full-link closed-loop control layer updates the dynamic quantitative scoring results of the news media dissemination effect at preset time intervals during the golden window period of news content dissemination after its release. When it finds that the evaluation index of a certain dimension is lower than the preset threshold, it automatically triggers the corresponding optimization strategy execution process.

[0013] Furthermore, the optimization strategies implemented by the full-link closed-loop control layer include adjusting the release time of content on different platforms, pushing content to high-influence, real-time dissemination nodes, optimizing content tags and recommended keywords, and adjusting the display format of content.

[0014] Furthermore, the full-link closed-loop control layer also includes an abnormal public opinion early warning module. This module is used to monitor the sentiment trend of comments on news content in real time. When it is found that the proportion of negative comments is rising rapidly, it automatically generates a graded response plan and pushes it to relevant staff.

[0015] Furthermore, the full-link closed-loop control layer feeds back the effect data after executing the optimization strategy to the cross-platform heterogeneous data perception layer in real time. The cross-platform heterogeneous data perception layer integrates the feedback effect data with the original collected data and then transmits it again to the three-dimensional nine-degree multi-dimensional evaluation layer for re-evaluation, forming a continuously iterative self-optimizing closed loop, and continuously improving the evaluation accuracy and optimization effect of the system.

[0016] The advantages of this invention compared to the prior art are: 1. Employing a unique three-dimensional nine-degree evaluation model, it conducts multi-dimensional evaluation from three core dimensions: breadth of dissemination, depth of dissemination, and value of dissemination. This effectively eliminates the interference of fake traffic and can truthfully and comprehensively reflect the actual dissemination effect of news content. Simultaneously, a dynamic weight adjustment mechanism is introduced, which can automatically adjust the weight of evaluation indicators according to the dissemination characteristics of different types of news content, further improving the accuracy of the evaluation. 2. This invention can update the dissemination effect evaluation results in real time at preset time intervals during the golden window period of news content dissemination and automatically trigger corresponding optimization strategies, realizing the transformation from "post-event statistics" to "real-time dynamic intervention", effectively seizing the best dissemination opportunity and significantly improving the dissemination effect; 3. This invention uses a causal inference algorithm to replace traditional correlation analysis, which can accurately identify the core factors that truly affect the dissemination effect. It also predicts the potential effects of different optimization strategies through a counterfactual simulation framework, generating scientific and executable personalized optimization solutions. This completely eliminates the reliance on human experience and greatly improves optimization efficiency and the stability of the effect. Attached Figure Description

[0017] Figure 1This is an overall architecture diagram of the multi-dimensional evaluation and optimization control system for news media dissemination effects provided in this embodiment of the invention; Figure 2 A flowchart of the dynamic weight adjustment module of the multi-dimensional evaluation and optimization control system for news media dissemination effects provided in this embodiment of the invention; Figure 3 The flowchart for identifying core factors in the multi-dimensional evaluation and optimization control system for news media dissemination effects provided in this embodiment of the invention; Figure 4 A flowchart illustrating the counterfactual simulation framework of the multi-dimensional evaluation and optimization control system for news media dissemination effects provided in this embodiment of the invention; Figure 5 A flowchart of an abnormal public opinion early warning system for a multi-dimensional evaluation and optimization control system for news media dissemination effects provided in this embodiment of the invention; Figure 6 The flowchart of the self-optimizing closed-loop iterative process of the multi-dimensional evaluation and optimization control system for news media dissemination effects provided in the embodiments of the present invention is shown. Detailed Implementation

[0018] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention. Example

[0019] Based on the above figure, this invention discloses a multi-dimensional evaluation and optimization control system for the dissemination effect of news media, including a cross-platform heterogeneous data perception layer, a three-dimensional nine-degree multi-dimensional evaluation layer, a causal inference intelligent decision-making layer, and a full-link closed-loop control layer. The cross-platform heterogeneous data perception layer is used to collect content data, user behavior data, propagation path tracing data and multi-dimensional public opinion data from mainstream platforms across the network in real time, and transmit the collected data to the three-dimensional nine-degree multi-dimensional evaluation layer. The three-dimensional nine-degree multi-dimensional evaluation layer is used to process the data transmitted by the cross-platform heterogeneous data perception layer based on the original three-dimensional nine-degree evaluation model, generate a dynamic quantitative score result of the news media dissemination effect, and transmit the dynamic quantitative score result to the causal inference intelligent decision-making layer. The causal inference intelligent decision-making layer is used to accurately identify the core factors affecting the dissemination effect of news media based on the dynamic quantitative scoring results, generate an executable personalized optimization plan, and transmit the personalized optimization plan to the full-link closed-loop control layer. The full-link closed-loop control layer is used to automatically execute the personalized optimization scheme and feed back the effect data after execution to the cross-platform heterogeneous data perception layer in real time, forming a continuously iterative self-optimizing closed loop.

[0020] In some embodiments, the cross-platform heterogeneous data perception layer includes multiple platform data acquisition modules, which correspond to different mainstream news dissemination platforms and are used to collect content publishing data, user click data, user comment data, user forwarding data, user collection data, and user dwell time data on the corresponding platforms.

[0021] In some embodiments, the three-dimensional nine-degree multi-dimensional evaluation layer uses a three-dimensional nine-degree evaluation model to construct nine quantifiable evaluation indicators from three core dimensions: the breadth of dissemination, the depth of dissemination, and the value of dissemination. The breadth of dissemination includes effective coverage, diffusion acceleration, and cross-platform penetration rate; The depth of dissemination includes average user dwell time, content re-creation rate, and emotional resonance. The dimensions of communication value include positive brand mention rate, conversion rate, and social influence index.

[0022] In some embodiments, the three-dimensional nine-degree multi-dimensional evaluation layer further includes a dynamic weight adjustment module, which is used to automatically identify the type of news content, adjust the weight of each evaluation indicator in the three-dimensional nine-degree evaluation model in real time according to different news content types, and generate a dynamic quantitative scoring result that matches the news content type.

[0023] In some embodiments, the causal inference intelligent decision-making layer employs a causal inference algorithm that combines the difference between two methods with propensity score matching to accurately identify the core factors that truly influence the dissemination effect of news media from multiple influencing factors such as title keywords, publication time, content format, and key dissemination nodes.

[0024] In some embodiments, the causal inference intelligent decision-making layer further includes a counterfactual simulation framework, which is used to construct counterfactual scenarios corresponding to different optimization strategies based on the identified core factors, predict the potential propagation effects of different optimization strategies, and provide a scientific basis for generating personalized optimization solutions.

[0025] In some embodiments, the end-to-end closed-loop control layer updates the dynamic quantitative scoring results of the news media dissemination effect at preset time intervals during the golden window period of news content dissemination. When it is found that the evaluation index of a certain dimension is lower than the preset threshold, the corresponding optimization strategy execution process is automatically triggered.

[0026] In some embodiments, the optimization strategies executed by the end-to-end closed-loop control layer include adjusting the release time of content on different platforms, pushing content to high-influence, real-time dissemination nodes, optimizing content tags and recommended keywords, and adjusting the display format of content.

[0027] In some embodiments, the end-to-end closed-loop control layer further includes an abnormal public opinion early warning module. The abnormal public opinion early warning module is used to monitor the sentiment trend of comments on news content in real time. When it is found that the proportion of negative comments is rising rapidly, it automatically generates a graded response plan and pushes it to relevant staff.

[0028] In some embodiments, the end-to-end closed-loop control layer feeds back the effect data after executing the optimization strategy to the cross-platform heterogeneous data perception layer in real time. The cross-platform heterogeneous data perception layer integrates the feedback effect data with the original collected data and then transmits it again to the three-dimensional nine-degree multi-dimensional evaluation layer for re-evaluation, forming a continuously iterative self-optimizing closed loop to continuously improve the evaluation accuracy and optimization effect of the system. Example

[0029] The multi-dimensional evaluation and optimization control system for news media dissemination effects disclosed in this invention includes a cross-platform heterogeneous data perception layer, a three-dimensional and nine-dimensional multi-dimensional evaluation layer, a causal inference intelligent decision-making layer, and a full-link closed-loop control layer.

[0030] The cross-platform heterogeneous data perception layer comprises multiple platform data acquisition modules, each corresponding to different mainstream news dissemination platforms, such as news apps, social media platforms, short video platforms, and portal websites. Each platform data acquisition module is responsible for collecting content publishing data, user click data, user comment data, user forwarding data, user favorites data, and user dwell time data on the corresponding platform. Simultaneously, the cross-platform heterogeneous data perception layer can also collect dissemination path tracing data and multi-dimensional public opinion data, comprehensively acquiring information on the dissemination of news content across the entire network. All collected data, after standardization processing, is transmitted to the three-dimensional nine-degree multi-dimensional evaluation layer.

[0031] The 3D Nine-Dimensional Evaluation Layer uses a unique 3D Nine-Dimensional Evaluation Model to process the transmitted data. This model constructs nine quantifiable evaluation indicators from three core dimensions: breadth of dissemination, depth of dissemination, and value of dissemination. The breadth of dissemination dimension includes effective coverage, diffusion acceleration, and cross-platform penetration rate, used to measure the scope and speed of news content dissemination; the depth of dissemination dimension includes average user dwell time, secondary content creation rate, and emotional resonance, used to measure user participation and emotional identification with the news content; and the value of dissemination dimension includes positive brand mention rate, conversion rate, and social influence index, used to measure the actual value and social impact of the news content.

[0032] The three-dimensional nine-degree multi-dimensional evaluation layer also includes a dynamic weight adjustment module. This module can automatically identify the type of news content, such as political news, business news, entertainment news, and social news. Different types of news content have different dissemination characteristics and value orientations, thus requiring different evaluation indicator weights. For example, for political news, the social influence index has a relatively high weight; for business news, the conversion rate has a relatively high weight. The dynamic weight adjustment module adjusts the weights of each evaluation indicator in the three-dimensional nine-degree evaluation model in real time according to the identified news content type, generating a dynamic quantitative score result that matches the news content type. The generated dynamic quantitative score result is transmitted to the causal inference intelligent decision-making layer.

[0033] The causal inference intelligent decision-making layer employs a causal inference algorithm combining the difference-in-differences method and propensity score matching. From hundreds of influencing factors, including headline keywords, publication time, content format, and key dissemination nodes, it accurately identifies the core factors truly affecting the dissemination effect of news media. Traditional correlation analysis is easily affected by confounding variables, producing spurious correlations and leading to decision-making errors. However, the causal inference algorithm used in this invention effectively eliminates the influence of confounding variables, accurately revealing the causal relationship between each influencing factor and the dissemination effect.

[0034] The causal inference intelligent decision-making layer also includes a counterfactual simulation framework. Based on identified core factors, this framework constructs counterfactual scenarios corresponding to different optimization strategies, simulating the dissemination effects of news content under different optimization strategies, thereby predicting the potential effects of different optimization strategies. By comparing the predictive effects of different optimization strategies, the causal inference intelligent decision-making layer can generate the optimal personalized optimization scheme and transmit this scheme to the end-to-end closed-loop control layer.

[0035] The end-to-end closed-loop control layer is responsible for automatically executing personalized optimization plans. During the golden window of dissemination after news content is released, the end-to-end closed-loop control layer will obtain the latest dynamic quantitative scoring results from the three-dimensional nine-dimensional multi-dimensional evaluation layer at preset time intervals. When it is found that the evaluation index of a certain dimension is lower than the preset threshold, the system will automatically trigger the corresponding optimization strategy execution process.

[0036] The end-to-end closed-loop control layer can execute various optimization strategies, including adjusting the release time of content on different platforms, targeted content push to high-influence, real-time dissemination nodes, optimizing content tags and recommended keywords, and adjusting the content display format. For example, when the system detects that the spread acceleration of a news item is below a preset threshold, it can automatically push the content to real users with high dissemination influence, thereby accelerating the spread of the content.

[0037] The end-to-end closed-loop control layer also includes an abnormal public opinion early warning module. This module monitors the sentiment trend of comments on news content in real time and performs sentiment analysis on user comments using natural language processing technology. When a rapid increase in the proportion of negative comments is detected, the system automatically generates a tiered response plan and pushes it to relevant staff so that timely measures can be taken to guide public opinion and prevent the further spread of negative public opinion.

[0038] After executing the optimization strategy, the end-to-end closed-loop control layer feeds back the resulting performance data to the cross-platform heterogeneous data perception layer in real time. This layer integrates the feedback data with the original collected data and then transmits it again to the three-dimensional, nine-degree multi-dimensional evaluation layer for re-evaluation. The evaluation results are then transmitted to the causal inference intelligent decision-making layer to optimize the causal inference algorithm and counterfactual simulation framework. This process repeats continuously, forming a self-optimizing closed loop that constantly improves the system's evaluation accuracy and optimization effectiveness.

[0039] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-dimensional evaluation and optimization control system for the dissemination effect of news media, characterized in that, It includes a cross-platform heterogeneous data perception layer, a three-dimensional and nine-degree multi-dimensional evaluation layer, a causal inference intelligent decision-making layer, and a full-link closed-loop control layer; The cross-platform heterogeneous data perception layer is used to collect content data, user behavior data, propagation path tracing data and multi-dimensional public opinion data from mainstream platforms across the network in real time, and transmit the collected data to the three-dimensional nine-degree multi-dimensional evaluation layer. The three-dimensional nine-degree multi-dimensional evaluation layer is used to process the data transmitted by the cross-platform heterogeneous data perception layer based on the original three-dimensional nine-degree evaluation model, generate a dynamic quantitative score result of the news media dissemination effect, and transmit the dynamic quantitative score result to the causal inference intelligent decision-making layer. The causal inference intelligent decision-making layer is used to accurately identify the core factors affecting the dissemination effect of news media based on the dynamic quantitative scoring results, generate an executable personalized optimization plan, and transmit the personalized optimization plan to the full-link closed-loop control layer. The full-link closed-loop control layer is used to automatically execute the personalized optimization scheme and feed back the effect data after execution to the cross-platform heterogeneous data perception layer in real time, forming a continuously iterative self-optimizing closed loop.

2. The multi-dimensional evaluation and optimization control system for news media dissemination effects according to claim 1, characterized in that, The cross-platform heterogeneous data perception layer includes multiple platform data acquisition modules, which correspond to different mainstream news dissemination platforms and are used to collect content publishing data, user click data, user comment data, user forwarding data, user collection data, and user dwell time data on the corresponding platforms.

3. The multi-dimensional evaluation and optimization control system for news media dissemination effects according to claim 1, characterized in that, The three-dimensional nine-degree multi-dimensional evaluation layer adopts a three-dimensional nine-degree evaluation model that constructs nine quantifiable evaluation indicators from three core dimensions: the breadth of dissemination, the depth of dissemination, and the value of dissemination. The breadth of dissemination includes effective coverage, diffusion acceleration, and cross-platform penetration rate; The depth of dissemination includes average user dwell time, content re-creation rate, and emotional resonance. The dimensions of communication value include positive brand mention rate, conversion rate, and social influence index.

4. The multi-dimensional evaluation and optimization control system for news media dissemination effects according to claim 3, characterized in that, The three-dimensional nine-degree multi-dimensional evaluation layer also includes a dynamic weight adjustment module. The dynamic weight adjustment module is used to automatically identify the type of news content and adjust the weight of each evaluation indicator in the three-dimensional nine-degree evaluation model in real time according to different news content types, so as to generate a dynamic quantitative score result that matches the news content type.

5. The multi-dimensional evaluation and optimization control system for news media dissemination effects according to claim 1, characterized in that, The causal inference intelligent decision-making layer adopts a causal inference algorithm that combines the difference between two methods with propensity score matching. It accurately identifies the core factors that truly affect the dissemination effect of news media from multiple influencing factors such as title keywords, publication time, content format, and key dissemination nodes.

6. The multi-dimensional evaluation and optimization control system for news media dissemination effects according to claim 5, characterized in that, The causal inference intelligent decision-making layer also includes a counterfactual simulation framework. This framework is used to construct counterfactual scenarios corresponding to different optimization strategies based on the identified core factors, predict the potential propagation effects of different optimization strategies, and provide a scientific basis for generating personalized optimization solutions.

7. The multi-dimensional evaluation and optimization control system for news media dissemination effects according to claim 1, characterized in that, During the golden window of dissemination after the release of news content, the full-link closed-loop control layer updates the dynamic quantitative scoring results of the news media dissemination effect at preset time intervals. When it is found that the evaluation index of a certain dimension is lower than the preset threshold, the corresponding optimization strategy execution process is automatically triggered.

8. The multi-dimensional evaluation and optimization control system for news media dissemination effects according to claim 7, characterized in that, The optimization strategies implemented by the full-link closed-loop control layer include adjusting the release time of content on different platforms, pushing content to high-influence, authentic dissemination nodes, optimizing content tags and recommended keywords, and adjusting the display format of content.

9. The multi-dimensional evaluation and optimization control system for news media dissemination effects according to claim 1, characterized in that, The full-link closed-loop control layer also includes an abnormal public opinion early warning module. This module is used to monitor the sentiment trend of comments on news content in real time. When it is found that the proportion of negative comments is rising rapidly, it automatically generates a graded response plan and pushes it to relevant staff.

10. The multi-dimensional evaluation and optimization control system for news media dissemination effects according to claim 1, characterized in that, The full-link closed-loop control layer feeds back the effect data after the optimization strategy is executed to the cross-platform heterogeneous data perception layer in real time. The cross-platform heterogeneous data perception layer integrates the feedback effect data with the original collected data and then transmits it to the three-dimensional nine-degree multi-dimensional evaluation layer for re-evaluation, forming a continuously iterative self-optimizing closed loop to continuously improve the evaluation accuracy and optimization effect of the system.

Images