Multi-channel digital marketing data management system and method based on big data

Abstract

The invention belongs to the technical field of sample size, and particularly relates to a multi-channel digital marketing data management system and method based on big data, and the system comprises a scene intelligent adaption module which is used for carrying out the intelligent recognition and adaption of a research scene after a user submits a research demand, and carrying out the intelligent recognition and adaption of the research scene; an accurate parameter and algorithm selection basis is provided for subsequent sample size calculation; and the intelligent sample size calculation module is used for firstly identifying key factors in a research scene, comprehensively analyzing the factors, calculating an initial sample size, then monitoring data change and external environment factors in real time, triggering a sample size dynamic adjustment algorithm, recalculating and giving an adjusted sample size suggestion. Through a risk assessment function, various risks possibly faced in research can be analyzed by using an algorithm, risk indexes are quantified, and an assessment report is provided, so that the mode of determining the sample size by researchers' subjective experience in the past is changed, and the decision is more objective and scientific.

Description

Technical Field

[0001] This invention relates to the field of sample size technology, specifically to a multi-channel digital marketing data management system and method based on big data. Background Technology

[0002] In today's era of rapid development in big data and artificial intelligence technologies, the scientific determination of sample size is crucial for various research and decision-making processes. Whether it's medical clinical trials to determine the efficacy of new drugs, market research analyzing consumer preferences, or social science research dissecting group behavior, the appropriate selection of sample size directly impacts the accuracy, reliability, and validity of research results. However, traditional methods for determining sample size primarily rely on simple formula calculations or empirical judgments, which are no longer sufficient to meet the complex and ever-changing needs of modern research and decision-making. Specifically, the following problems exist: 1. Traditional methods lack dynamic adaptability: Traditional sample size calculations are mostly based on fixed formulas and static assumptions, which cannot adapt to dynamic changes in data characteristics, research objectives, and the external environment during the research process. For example, in a multi-year clinical trial of an oncology drug, as the research progresses, new treatment options emerge, patient enrollment rates change, and data from patients in different regions vary. Traditional methods cannot adjust the sample size in real time, which may lead to insufficient or excessive sample sizes. Insufficient sample sizes will result in statistically weak research results and an inability to accurately determine drug efficacy; excessive sample sizes will lead to a waste of medical resources, time, and money.

[0003] 2. Computational difficulties in complex scenarios: In complex research scenarios with multiple interacting factors and significant nonlinear relationships, traditional methods struggle to accurately calculate sample size. For example, when studying the impact of social media use on adolescent mental health, multiple factors are involved, such as age, gender, family environment, and social circles. These factors are interconnected and have complex relationships. Traditional calculation methods cannot comprehensively consider these complex relationships, which can easily lead to biases in sample size calculations and thus affect the scientific validity of the research conclusions.

[0004] 3. High Subjectivity of Human Experience Judgment: When relying on researchers' personal experience to determine sample size, different researchers may arrive at different sample size recommendations due to differences in knowledge, practical experience, and subjective preferences. For example, two market researchers conducting market demand research on the same product may differ in their opinions. One may believe that a larger sample size should be used to obtain more comprehensive data, while the other, based on experience with similar projects in the past, may believe that a smaller sample size is sufficient. This difference in subjectivity may lead to significant biases in the research results, making it impossible to provide a reliable basis for corporate decision-making.

[0005] 4. Insufficient data-driven capability: Traditional methods do not fully utilize historical and real-time data in determining sample size, failing to uncover hidden patterns and value. For example, in urban traffic flow research, a large amount of historical traffic data exists, but traditional methods have failed to effectively utilize this data to predict traffic flow changes in different time periods and under different weather conditions. Consequently, it is impossible to scientifically determine the sample size required for traffic flow monitoring, resulting in monitoring results that do not accurately reflect actual traffic conditions.

[0006] Based on the above, a multi-channel digital marketing data management system and method based on big data is invented. Summary of the Invention

[0007] To address the aforementioned technical problems, according to one aspect of the present invention, the present invention provides the following technical solution: A multi-channel digital marketing data management system and method based on big data, comprising: The scene intelligent adaptation module is used to intelligently identify and adapt the research scene after the user submits the research requirements, providing accurate parameters and algorithm selection basis for subsequent sample size calculation; The intelligent sample size calculation module is used to first identify key factors in the research scenario, and then perform comprehensive analysis on the factors to calculate the initial sample size. Next, it monitors data changes and external environmental factors in real time, and can trigger a dynamic sample size adjustment algorithm to recalculate and provide suggestions for the adjusted sample size. The scheme evaluation module is used to first analyze the potential risks in the research process based on the sample size scheme using risk assessment algorithms, and then conduct a comprehensive comparative analysis and evaluation of different schemes. The results verification module is used to further verify the results after obtaining the sample size calculation results and the scheme evaluation conclusions, so as to ensure the scientificity and accuracy of the final scheme. The results visualization module is used to first display the data, intermediate results and final results in the sample size calculation process in an intuitive chart form. Then, for complex intelligent algorithm models, it uses visualization technology to display the model's structure, parameters and operating mechanism. The report generation module is used to automatically generate detailed sample size determination reports based on user needs and analysis results.

[0008] As a preferred embodiment of the multi-channel digital marketing data management system and method based on big data described in this invention, the scenario intelligent adaptation module includes: The information collection and preliminary analysis module is used to collect and analyze information after the user inputs relevant information such as research description, research objectives, and data types. The multi-dimensional scene recognition module first matches the key information obtained from preliminary analysis against a pre-set basic scene classification library to determine the approximate scene category to which the research belongs. Then, it uses image recognition and knowledge graph technologies to further mine detailed features in the information and subdivide the basic scene. Afterward, to ensure the reliability of scene recognition, it evaluates the confidence of the identified subdivided scenes by calculating the information matching degree and referring to the recognition accuracy of similar historical scenes to give a confidence score. When the confidence score reaches a certain threshold, the scene recognition is considered valid; if it is below the threshold, the user is prompted to supplement information or manual intervention is required. The intelligent parameter and algorithm recommendation module is used to first search the scenario database based on the final determined sub-scenario; then, after retrieving relevant information, it will combine the specific needs and data characteristics of the current research to screen and optimize the parameters and algorithms, so as to finally determine a set of parameters and algorithms that are most suitable for the current research scenario, and pass it to the sample size intelligent calculation module. The dynamic scene learning and optimization module is used to continuously collect new research case data during operation; and when a certain number of new cases are accumulated, machine learning algorithms are activated to optimize the scene recognition model and parameter recommendation strategy. Through continuous learning and optimization, the module's adaptability to various research scenarios is continuously improved, better meeting the diverse needs of users.

[0009] As a preferred embodiment of the multi-channel digital marketing data management system and method based on big data described in this invention, the intelligent sample size calculation module includes: The multi-factor comprehensive analysis module is used to automatically identify key factors in the research scenario after the user inputs relevant information such as research description, research objectives, and data types. Then, it uses intelligent algorithms to comprehensively analyze the factors, establish a mathematical model, and calculate the initial sample size. The dynamic adjustment mechanism module is used to monitor data changes and external environmental factors in real time during the research process. When data characteristics change significantly or external environmental factors have a major impact, it can automatically trigger the dynamic sample size adjustment algorithm, recalculate and provide an adjusted sample size suggestion to ensure the effectiveness and reliability of the research.

[0010] As a preferred embodiment of the multi-channel digital marketing data management system and method based on big data described in this invention, the embodiment evaluation module includes: The risk assessment module is used to analyze the risks that may be faced during the research process based on the calculated sample size plan using risk assessment algorithms. It can also provide users with risk level assessment reports by quantifying risk indicators to help users understand the potential risks of the sample size plan. The scheme comparison module is used to comprehensively compare and analyze different schemes when users propose multiple sample size schemes or generate multiple candidate schemes. It can provide users with objective and accurate analysis of the advantages and disadvantages of the schemes through data comparison and visualization, and help users make the best decision.

[0011] As a preferred embodiment of the multi-channel digital marketing data management system and method based on big data described in this invention, the result verification module includes simulation verification, expert verification, and cross-method verification.

[0012] As a preferred embodiment of the multi-channel digital marketing data management system and method based on big data described in this invention, the specific steps of the simulation verification are as follows: S11, Data Preparation: First, retrieve historical data or generate simulated data; if there is abundant historical data, relevant data will be selected according to the research scenario; if there is insufficient historical data, simulated data that conforms to the research scenario will be generated based on existing data distribution characteristics, research hypotheses, and probability models. S12, Simulation Experiment Execution: Based on the sample size scheme output by the intelligent sample size calculation module, multiple simulation sampling experiments are conducted using the selected data; first, data of the corresponding sample size are extracted from the dataset according to different sampling methods, and the statistical analysis methods used in the sample size calculation are applied to analyze the simulation sample data to obtain the simulation results. After each simulation experiment, key indicators are recorded. S13, Result Stability Assessment: Statistical analysis is performed on the results of multiple simulation experiments, and the statistics of each key indicator are calculated. If the statistics are at a low level, it indicates that the simulation results are highly stable and the sample size scheme is reliable; otherwise, it indicates that there may be problems with the sample size scheme and further adjustments are needed.

[0013] As a preferred embodiment of the multi-channel digital marketing data management system and method based on big data described in this invention, the specific steps of the expert verification are as follows: S21, Solution Submission: Compile the sample size calculation results, solution evaluation conclusions and related analysis processes into a detailed report, and submit it to domain experts through the expert verification interface; S22, Expert Feedback Collection: Experts review the submitted proposals based on their professional knowledge and practical experience, and provide feedback and suggestions through the interface; S23, Feedback Integration and Processing: After receiving expert feedback, integrate it with your own analysis results to evaluate the feedback and determine whether the sample size plan needs to be adjusted.

[0014] As a preferred embodiment of the multi-channel digital marketing data management system and method based on big data described in this invention, the specific steps of the cross-method verification are as follows: S31, Method selection: Select a method that differs from the original calculation method from a library of various sample size calculation methods and statistical analysis methods; S32, Recalculation and Comparison: Using the selected method, recalculate the sample size based on the original or adjusted data, and analyze the results; so as to be able to compare the new calculation results with the original sample size calculation results, and evaluate them in terms of numerical differences and consistency of statistical conclusions.

[0015] As a preferred embodiment of the multi-channel digital marketing data management system and method based on big data described in this invention, the result visualization module includes: The data visualization module is used to display the data, intermediate results, and final results in the sample size calculation process in an intuitive chart format; The model visualization module is used to visualize the structure, parameters, and operating mechanism of complex intelligent algorithm models, enabling users to intuitively understand the principles and logic of sample size calculation and enhancing their trust in the system results.

[0016] Compared with existing technologies: I. The intelligent sample size calculation module accurately addresses dynamic and complex challenges: Traditional methods struggle with dynamic changes and complex scenarios in research, while the intelligent sample size calculation module effectively addresses these issues. The multi-factor comprehensive analysis function automatically identifies key factors after the user inputs relevant research information and uses intelligent algorithms for comprehensive analysis. This solves the problems of traditional methods struggling to comprehensively consider complex relationships between multiple factors and prone to bias in sample size calculations in complex scenarios. For example, in researching the impact of social media use on adolescent mental health, the system can incorporate numerous factors such as age, gender, and family environment to establish an accurate mathematical model for calculating the initial sample size, thus improving the scientific rigor of the research conclusions.

[0017] The dynamic adjustment mechanism compensates for the lack of dynamic adaptability in traditional methods. In long-term studies such as clinical trials of cancer drugs, the system monitors data changes and external environmental factors in real time. When new treatment options emerge or patient enrollment rates change, the dynamic adjustment algorithm is automatically triggered to recalculate the sample size, avoiding insufficient or excessive sample size, ensuring the statistical validity of the research results, and at the same time making rational use of medical resources to reduce research costs and time costs.

[0018] II. Problems with subjectivity and insufficient data utilization in modular solution evaluation: The inherent subjectivity of human experience in judgment and the insufficient data-driven capabilities of traditional methods can be effectively addressed through the scheme evaluation module. The risk assessment function, based on calculated sample size schemes, uses algorithms to analyze various risks that may be encountered in the research, quantifies risk indicators, and provides assessment reports. This changes the previous approach of relying on researchers' subjective experience to determine sample size, making decision-making more objective and scientific. For example, in market research, the system can objectively assess risks such as insufficient statistical power and cost overruns under different sample size schemes based on data, avoiding sample size bias caused by researchers' subjective preferences, and providing a reliable basis for corporate decision-making. The scheme comparison function can evaluate multiple schemes with different sample sizes from multiple dimensions, and analyze their advantages and disadvantages through data comparison and visualization. This function fully leverages the value of data, utilizing historical and real-time data to provide users with objective scheme comparison results, solving the problem of insufficient data utilization in traditional methods. For example, in urban traffic flow research, the system can use historical traffic data to compare the monitoring effects of different sample size schemes under different time periods and weather conditions, scientifically determine the sample size required for traffic flow monitoring, and make the monitoring results more accurately reflect the actual traffic conditions.

[0019] III. Results visualization and report generation modules enhance research transparency and usability. The results visualization module uses data visualization and model visualization to present the sample size calculation process and results in intuitive charts, while also showcasing the structure and operating mechanism of complex algorithm models. This not only helps users understand the data and analysis results more clearly, but also enhances users' trust in the system's results, changing the situation where traditional methods present results in a non-intuitive and difficult-to-understand manner, and improving the transparency of research.

[0020] The report generation module automatically generates detailed sample size determination reports, covering research background, objectives, data sources, and other aspects, and supports multiple output formats. This makes research results more standardized and practical, facilitating user archiving, sharing, and submission for review. Compared to traditional methods, it better meets the needs of different users in recording and applying research results in different scenarios. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the framework of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0023] This invention provides a multi-channel digital marketing data management system and method based on big data. Please refer to [link / reference]. Figure 1 ,include: The scene intelligent adaptation module is used to intelligently identify and adapt the research scene after the user submits the research requirements, providing accurate parameters and algorithm selection basis for subsequent sample size calculation; The intelligent sample size calculation module is used to first identify key factors in the research scenario, and then perform comprehensive analysis on the factors to calculate the initial sample size. Next, it monitors data changes and external environmental factors in real time, and can trigger a dynamic sample size adjustment algorithm to recalculate and provide suggestions for the adjusted sample size. The scheme evaluation module is used to first analyze the potential risks in the research process based on the sample size scheme using risk assessment algorithms, and then conduct a comprehensive comparative analysis and evaluation of different schemes. The results verification module is used to further verify the results after obtaining the sample size calculation results and the scheme evaluation conclusions, so as to ensure the scientificity and accuracy of the final scheme. The results visualization module is used to first display the data, intermediate results and final results in the sample size calculation process in an intuitive chart form. Then, for complex intelligent algorithm models, it uses visualization technology to display the model's structure, parameters and operating mechanism. The report generation module automatically generates a detailed sample size determination report based on user needs and analysis results. The report includes research background, research objectives, data sources, analysis methods, sample size calculation process and results, protocol evaluation conclusions, and risk warnings. The report supports multiple output formats (such as PDF, Word, and Excel) for easy archiving, sharing, and submission for review.

[0024] The scene intelligent adaptation module includes: The information collection and preliminary analysis module is used to collect and analyze information after the user inputs relevant information such as research description, research objectives, and data types. For text-based information, natural language processing technology is used to perform operations such as word segmentation, part-of-speech tagging, and semantic understanding to extract key semantic information. For structured information such as data types, format validation and standardization are performed directly. The multi-dimensional scene recognition module first matches the key information obtained from preliminary analysis against a pre-set basic scene classification library to determine the approximate scene category to which the research belongs. Then, it utilizes image recognition and knowledge graph technologies to further mine detailed features within the information, further subdividing the basic scene. Combining this with knowledge graph information on subdivided medical clinical trials, such as based on drug type, research subject characteristics, and trial objectives, the "medical clinical trial" scene is subdivided into the specific scene of "efficacy verification trial of diabetes drugs for middle-aged and elderly populations." During this process, the system comprehensively considers the relationships between multiple factors to ensure the accuracy of scene recognition. Subsequently, to ensure the reliability of scene recognition, a confidence assessment is performed on the identified subdivided scene. A confidence score is given by calculating the information matching degree and referencing the recognition accuracy of similar historical scenes. When the confidence score reaches a certain threshold (e.g., 80%), the scene recognition is considered valid; if it falls below the threshold, the user is prompted to supplement information or manual intervention is required. The intelligent parameter and algorithm recommendation module first searches the scenario database based on the final determined sub-scenario. The database stores information such as historical success cases, best practice parameters, and applicable algorithms for different scenarios. After retrieving relevant information, it filters and optimizes parameters and algorithms based on the specific needs and data characteristics of the current research. If the amount of data in the current research is small, the system may prioritize recommending statistical methods suitable for small sample studies and make appropriate adjustments to the parameters to improve the accuracy and applicability of sample size calculation. Finally, it determines a set of parameters and algorithms that are most suitable for the current research scenario and passes it to the intelligent sample size calculation module. The dynamic scenario learning and optimization module continuously collects new research case data during operation. When a certain number of new cases are accumulated, machine learning algorithms are activated to optimize the scenario recognition model and parameter recommendation strategy. For example, if a new research scenario, "clinical trial of diabetes drugs based on artificial intelligence-assisted diagnosis," emerges, the system will analyze the characteristics of the scenario, incorporate it into the scenario database, and update the parameters and algorithm recommendation rules. Through continuous learning and optimization, the module's adaptability to various research scenarios is constantly improved, better meeting the diverse needs of users.

[0025] By setting up a scene intelligent adaptation module, not only can accurate scene identification be performed first, providing a more practical foundation for subsequent sample size calculations, but it can also avoid the problem of users choosing incorrect parameters and algorithms due to insufficient professional knowledge, thereby improving the accuracy and efficiency of sample size calculations. At the same time, it lowers the threshold for system use, allowing non-professional users to easily obtain scientific sample size solutions. In addition, it can keep up with the development and changes in the research field, continuously improve its adaptation capabilities, and always maintain its advanced nature and practicality.

[0026] The intelligent sample size calculation module includes: The multi-factor comprehensive analysis module is used to automatically identify key factors in the research scenario after the user inputs relevant information such as research description, research objectives, and data types. These factors include the characteristics of the research subjects, the characteristics of the research methods, and the distribution of data. Then, intelligent algorithms are used to comprehensively analyze the factors, establish a mathematical model, and calculate the initial sample size. The dynamic adjustment mechanism module is used to monitor data changes and external environmental factors (such as policy and regulatory changes, market dynamics, etc.) in real time during the research process. When data characteristics change significantly (such as increased data variance or effect size shift) or when major influencing factors occur in the external environment, the dynamic sample size adjustment algorithm can be automatically triggered to recalculate and provide an adjusted sample size suggestion to ensure the effectiveness and reliability of the research.

[0027] The scheme evaluation module includes: The risk assessment module is used to analyze potential risks during the research process based on the calculated sample size plan using risk assessment algorithms, such as the risk of insufficient statistical power, the risk of research cost overruns, and the risk of time delays. It can also provide users with risk level assessment reports by quantifying risk indicators to help users understand the potential risks of the sample size plan. The scheme comparison module is used to conduct a comprehensive comparative analysis of different schemes when users propose multiple sample size schemes or generate multiple candidate schemes. It evaluates the schemes from multiple dimensions such as sample size, research cost, research period, statistical power, and feasibility. Through data comparison and visualization, it provides users with an objective and accurate analysis of the advantages and disadvantages of the schemes, helping users to make the best decision.

[0028] The result verification module includes simulation verification, expert verification, and cross-method verification. In addition, the result verification module can also make a comprehensive judgment based on the results of simulation verification, expert verification, and cross-method verification. If multiple verification methods support the original sample size scheme, the scheme is confirmed to be feasible. If there are obvious contradictions or problems, the user will be prompted to re-examine the research design, data quality, or calculation method and revise the sample size scheme.

[0029] The specific steps of the simulation verification are as follows: S11, Data Preparation: First, retrieve historical data or generate simulated data. If abundant historical data exists, relevant data will be selected based on the research scenario, such as historical patient data from medical clinical trials or past consumer behavior data from market research. If insufficient historical data is available, simulated data that fits the research scenario will be generated based on existing data distribution characteristics, research hypotheses, and probability models. For example, when simulating the market demand survey sample size for a new type of smartwatch, the system will generate simulated data based on the sales data distribution of similar products and the characteristics of the target user group. S12, Simulation Experiment Execution: Based on the sample size scheme output by the intelligent sample size calculation module, multiple simulated sampling experiments are conducted using the selected data. First, data of the corresponding sample size are extracted from the dataset according to different sampling methods (such as simple random sampling and stratified sampling). Then, the statistical analysis methods used in the sample size calculation are applied to analyze the simulated sample data to obtain the simulation results. After each simulation experiment, key indicators (such as the estimated value of the statistic, confidence interval, etc.) are recorded. S13, Results Stability Assessment: Statistical analysis is performed on the results of multiple simulation experiments. The coefficient of variation, standard deviation, and other statistical measures of stability for each key indicator are calculated. If the statistical measures are at a low level, it indicates that the simulation results are highly stable and the sample size scheme is reliable. Conversely, it indicates that there may be problems with the sample size scheme and further adjustments are needed.

[0030] The specific steps for expert verification are as follows: S21, Proposal Submission: Compile the sample size calculation results, proposal evaluation conclusions, and related analysis processes into a detailed report, and submit it to domain experts through the expert verification interface; the report includes information such as research background, data sources, calculation methods, and risk assessment results, so that experts can fully understand the proposal content; S22, Expert Feedback Collection: Experts review the submitted proposals based on their professional knowledge and practical experience, and provide feedback and suggestions through the interface; experts may offer insights from the perspectives of the rationality of the research design, the applicability of the statistical methods, and the clinical significance; S23, Feedback Integration and Processing: After receiving expert feedback, integrate it with your own analysis results to evaluate the feedback and determine whether the sample size plan needs to be adjusted.

[0031] The specific steps of the cross-method verification are as follows: S31, Method Selection: Select a method different from the original calculation method from a library of various sample size calculation methods and statistical analysis methods; for example, if the original sample size calculation uses a machine learning-based prediction model, a traditional statistical formula calculation method (such as a sample size calculation formula based on hypothesis testing) can be selected for cross-method validation; or a different machine learning model can be selected (such as switching from a neural network model to a random forest model); S32, Recalculation and Comparison: Using the selected method, recalculate the sample size based on the original or adjusted data, and analyze the results; so as to be able to compare the new calculation results with the original sample size calculation results, and evaluate them in terms of numerical differences and consistency of statistical conclusions.

[0032] By setting up a results verification module, potential problems with the sample size scheme, such as statistical bias and inaccurate estimation, can be identified in advance, providing a basis for scheme optimization and adjustment and improving the success rate of actual research and decision-making. Furthermore, it can combine expert feedback with system analysis results to achieve human-machine collaborative verification, compensating for the potential limitations of intelligent algorithms and making the sample size determination scheme more aligned with actual research needs. In addition, it can reduce errors and biases that may arise from a single method, improving the credibility and accuracy of the results and enhancing users' confidence in the system's output.

[0033] The results visualization module includes: The data visualization module is used to display the data, intermediate results, and final results in the sample size calculation process in intuitive chart formats, such as bar charts, line charts, pie charts, and heatmaps. For example, a bar chart can be used to show the comparison of sample size under different schemes, and a line chart can be used to show the dynamic trend of sample size changes with the research process, helping users to understand the data and analysis results more clearly. The model visualization module is used to visualize the structure, parameters, and operating mechanism of complex intelligent algorithm models, enabling users to intuitively understand the principles and logic of sample size calculation and enhancing their trust in the system results.

[0034] In practical use, the specific operating steps are as follows: Step 1: After the user inputs relevant information such as research description, research objectives, and data type, the information collection and preliminary analysis module collects and analyzes the information. Following analysis, the multi-dimensional scene recognition module first matches the key information obtained from the preliminary analysis against a pre-set basic scene classification library to determine the approximate scene category to which the research belongs. Then, image recognition and knowledge graph technologies are used to further mine detailed features in the information and subdivide the basic scene. Afterwards, to ensure the reliability of scene recognition, a confidence assessment is performed on the identified subdivided scenes. A confidence score is given by calculating the information matching degree and referring to the recognition accuracy of similar historical scenes. When the confidence score reaches a certain threshold, the scene recognition is considered valid; if it is below the threshold, the user is prompted to supplement the score. Information is added or manual intervention is performed for correction. After identification, the intelligent parameter and algorithm recommendation module first searches the scene database based on the final determined sub-scene. Then, after retrieving relevant information, it combines the specific needs of the current research and the characteristics of the data to screen and optimize the parameters and algorithms, so as to finally determine a set of parameters and algorithms most suitable for the current research scene, and pass it to the sample size intelligent calculation module. After recommendation, the dynamic scene learning and optimization module can continuously collect new research case data during operation. When a certain number of new cases are accumulated, machine learning algorithms can be launched to optimize the scene recognition model and parameter recommendation strategy. Through continuous learning and optimization, the module's adaptability to various research scenarios is continuously improved, better meeting the diverse needs of users. Step Two: After the user inputs relevant information such as research description, research objectives, and data types, the multi-factor comprehensive analysis module can automatically identify key factors in the research scenario. Then, it uses intelligent algorithms to comprehensively analyze the factors, establish a mathematical model, and calculate the initial sample size. After analysis, the dynamic adjustment mechanism module can monitor data changes and external environmental factors in real time during the research process. When data characteristics change significantly or external environmental factors have significant impacts, the dynamic sample size adjustment algorithm can be automatically triggered to recalculate and provide an adjusted sample size suggestion, ensuring the effectiveness and reliability of the research. Step 3: The risk assessment module analyzes the potential risks during the research process based on the calculated sample size plan using risk assessment algorithms. It can also provide users with a risk level assessment report through quantitative risk indicators, helping users understand the potential risks of the sample size plan. After the assessment, the plan comparison module can conduct a comprehensive comparative analysis of different plans when users propose multiple sample size plans or generate multiple candidate plans. It can also provide users with an objective and accurate analysis of the advantages and disadvantages of different plans through data comparison and visualization, assisting users in making the best decision. Step 4: After obtaining the sample size calculation results and scheme evaluation conclusions through the result verification module, the results can be further verified to ensure the scientificity and accuracy of the final scheme. Step 5: The data visualization module displays the data, intermediate results, and final results of the sample size calculation process in an intuitive chart format. After the display, the model visualization module will use visualization technology to show the structure, parameters, and operating mechanism of the complex intelligent algorithm model, so that users can intuitively understand the principle and logic of the sample size calculation and enhance their trust in the system results. Step Six: The report generation module automatically generates a detailed sample size determination report based on the user's needs and analysis results.

[0035] Although the present invention has been described above with reference to embodiments, various modifications can be made and components can be replaced with equivalents without departing from the scope of the invention. In particular, as long as there is no structural conflict, the features in the disclosed embodiments can be combined with each other in any manner. The lack of an exhaustive description of these combinations in this specification is merely for the sake of brevity and resource conservation. Therefore, the present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A multi-channel digital marketing data management system and method based on big data, characterized in that, include: The scene intelligent adaptation module is used to intelligently identify and adapt the research scene after the user submits the research requirements, providing accurate parameters and algorithm selection basis for subsequent sample size calculation; The intelligent sample size calculation module is used to first identify key factors in the research scenario, and then perform comprehensive analysis on the factors to calculate the initial sample size. Next, it monitors data changes and external environmental factors in real time, and can trigger a dynamic sample size adjustment algorithm to recalculate and provide suggestions for the adjusted sample size. The scheme evaluation module is used to first analyze the potential risks in the research process based on the sample size scheme using risk assessment algorithms, and then conduct a comprehensive comparative analysis and evaluation of different schemes. The results verification module is used to further verify the results after obtaining the sample size calculation results and the scheme evaluation conclusions, so as to ensure the scientificity and accuracy of the final scheme. The results visualization module is used to first display the data, intermediate results and final results in the sample size calculation process in an intuitive chart form. Then, for complex intelligent algorithm models, it uses visualization technology to display the model's structure, parameters and operating mechanism. The report generation module is used to automatically generate detailed sample size determination reports based on user needs and analysis results.

2. The multi-channel digital marketing data management system and method based on big data according to claim 1, characterized in that, The scene intelligent adaptation module includes: The information collection and preliminary analysis module is used to collect and analyze information after the user inputs relevant information such as research description, research objectives, and data types. The multi-dimensional scene recognition module first matches the key information obtained from preliminary analysis against a pre-set basic scene classification library to determine the approximate scene category to which the research belongs. Then, it uses image recognition and knowledge graph technologies to further mine detailed features in the information and subdivide the basic scene. Afterward, to ensure the reliability of scene recognition, it evaluates the confidence of the identified subdivided scenes by calculating the information matching degree and referring to the recognition accuracy of similar historical scenes to give a confidence score. When the confidence score reaches a certain threshold, the scene recognition is considered valid; if it is below the threshold, the user is prompted to supplement information or manual intervention is required. The intelligent parameter and algorithm recommendation module is used to first search the scenario database based on the final determined sub-scenario; then, after retrieving relevant information, it will combine the specific needs and data characteristics of the current research to screen and optimize the parameters and algorithms, so as to finally determine a set of parameters and algorithms that are most suitable for the current research scenario, and pass it to the sample size intelligent calculation module. The dynamic scene learning and optimization module is used to continuously collect new research case data during operation; and when a certain number of new cases are accumulated, machine learning algorithms are activated to optimize the scene recognition model and parameter recommendation strategy. Through continuous learning and optimization, the module's adaptability to various research scenarios is continuously improved, better meeting the diverse needs of users.

3. The multi-channel digital marketing data management system and method based on big data according to claim 1, characterized in that, The intelligent sample size calculation module includes: The multi-factor comprehensive analysis module is used to automatically identify key factors in the research scenario after the user inputs relevant information such as research description, research objectives, and data types. Then, it uses intelligent algorithms to comprehensively analyze the factors, establish a mathematical model, and calculate the initial sample size. The dynamic adjustment mechanism module is used to monitor data changes and external environmental factors in real time during the research process. When data characteristics change significantly or external environmental factors have a major impact, it can automatically trigger the dynamic sample size adjustment algorithm, recalculate and provide an adjusted sample size suggestion to ensure the effectiveness and reliability of the research.

4. The multi-channel digital marketing data management system and method based on big data according to claim 1, characterized in that, The scheme evaluation module includes: The risk assessment module is used to analyze the risks that may be faced during the research process based on the calculated sample size plan using risk assessment algorithms. It can also provide users with risk level assessment reports by quantifying risk indicators to help users understand the potential risks of the sample size plan. The scheme comparison module is used to comprehensively compare and analyze different schemes when users propose multiple sample size schemes or generate multiple candidate schemes. It can provide users with objective and accurate analysis of the advantages and disadvantages of the schemes through data comparison and visualization, and help users make the best decision.

5. The multi-channel digital marketing data management system and method based on big data according to claim 1, characterized in that, The result verification module includes simulation verification, expert verification, and cross-method verification.

6. The multi-channel digital marketing data management system and method based on big data according to claim 5, characterized in that, The specific steps of the simulation verification are as follows: S11, Data Preparation: First, retrieve historical data or generate simulated data; if there is abundant historical data, relevant data will be selected according to the research scenario; if there is insufficient historical data, simulated data that conforms to the research scenario will be generated based on existing data distribution characteristics, research hypotheses, and probability models. S12, Simulation Experiment Execution: Based on the sample size scheme output by the intelligent sample size calculation module, multiple simulation sampling experiments are conducted using the selected data; first, data of the corresponding sample size are extracted from the dataset according to different sampling methods, and the statistical analysis methods used in the sample size calculation are applied to analyze the simulation sample data to obtain the simulation results. After each simulation experiment, key indicators are recorded. S13, Result Stability Assessment: Statistical analysis is performed on the results of multiple simulation experiments, and the statistics of each key indicator are calculated. If the statistics are at a low level, it indicates that the simulation results are highly stable and the sample size scheme is reliable; otherwise, it indicates that there may be problems with the sample size scheme and further adjustments are needed.

7. The multi-channel digital marketing data management system and method based on big data according to claim 5, characterized in that, The specific steps for expert verification are as follows: S21, Solution Submission: Compile the sample size calculation results, solution evaluation conclusions and related analysis processes into a detailed report, and submit it to domain experts through the expert verification interface; S22, Expert Feedback Collection: Experts review the submitted proposals based on their professional knowledge and practical experience, and provide feedback and suggestions through the interface; S23, Feedback Integration and Processing: After receiving expert feedback, integrate it with your own analysis results to evaluate the feedback and determine whether the sample size plan needs to be adjusted.

8. The multi-channel digital marketing data management system and method based on big data according to claim 5, characterized in that, The specific steps of the cross-method verification are as follows: S31, Method selection: Select a method that differs from the original calculation method from a library of various sample size calculation methods and statistical analysis methods; S32, Recalculation and Comparison: Using the selected method, recalculate the sample size based on the original or adjusted data, and analyze the results; so as to be able to compare the new calculation results with the original sample size calculation results, and evaluate them in terms of numerical differences and consistency of statistical conclusions.

9. The multi-channel digital marketing data management system and method based on big data according to claim 1, characterized in that, The results visualization module includes: The data visualization module is used to display the data, intermediate results, and final results in the sample size calculation process in an intuitive chart format; The model visualization module is used to visualize the structure, parameters, and operating mechanism of complex intelligent algorithm models, enabling users to intuitively understand the principles and logic of sample size calculation and enhancing their trust in the system results.

Images