A Method and System for Multi-Platform Live Streaming Room Data Visualization and Optimization Decision Support

Abstract

The present invention proposes a method and system for multi-platform live broadcast room data visualization and optimization decision support. By collecting historical live broadcast data of live broadcast rooms on different platforms and visualizing the live broadcast data, a multi-dimensional benefit evaluation index system is constructed based on heterogeneous data sources. The number of exposures of the live broadcast room is used as a proxy index for generating benefits. The benefit function of the live broadcast room is introduced to model the capital optimization problem as a multi-objective optimization model. The optimal allocation strategy is solved with the maximization of the benefit function as the objective function. The particle swarm optimization algorithm is applied to iteratively solve the number of exposures of the live broadcast room that maximizes the benefit. The live broadcast data is analyzed through quantitative historical data and multi-objective optimization algorithms, and a visualized allocation plan is sorted out by a computer, thereby helping users make reasonable decisions. Through optimization and decision support, the present invention helps decision-makers flexibly adjust strategies in a dynamically changing market environment, avoid blind investment, and improve decision-making accuracy and overall operation efficiency.

Description

Technical Field

[0001] The present invention relates to the field of data analysis, and specifically to a method and system for multi-platform live broadcast room data visualization and optimization decision support. Background Art

[0002] In current cross-platform e-commerce live broadcasts, merchants face the dilemma that live broadcast data is scattered and difficult to summarize. The lack of clear and intuitive live broadcast data display makes it difficult for merchants to effectively rely on data analysis results in operation decisions, resulting in low decision-making efficiency and even affecting sales performance.

[0003] The live broadcast data of each platform is usually stored in different systems or databases, and the data structures and formats are different, which makes data integration, summarization, and analysis very complex. In addition, existing live broadcast data displays usually lack sufficient visualization tools, resulting in merchants being unable to comprehensively and clearly see the operation status of different platforms when making cross-platform operation decisions. System users often rely on cumbersome manual summarization and unintuitive data reports, and it is difficult to quickly grasp the overall trend and key indicators of live broadcast effects, thus affecting the quality of their decisions. Therefore, there is an urgent need for a method that can quantify and visualize the data of each live broadcast platform to help merchants maximize the benefits of live broadcast rooms on each platform through reasonable capital allocation strategies, thereby improving the overall operation effect and sales performance. Summary of the Invention

[0004] The present invention proposes a method and system for multi-platform live broadcast room data visualization and optimization decision support. By collecting historical live broadcast data of live broadcast rooms on different platforms and visualizing the live broadcast data, analyzing the live broadcast data through quantifying historical data and multi-objective optimization algorithms, and sorting out a visual allocation plan, it helps users make reasonable decisions to solve the above-mentioned problems. The technical solutions provided by the present invention are as follows:

[0005] A method for multi-platform live broadcast room data visualization and optimization decision support includes the following steps:

[0006] Step 1, collect real-time data and historical data of live broadcast rooms on multiple platforms, and construct a data matrix of multiple live broadcast rooms;

[0007] Step 2, perform standardization processing on the collected live broadcast room data through data preprocessing and feature extraction techniques to obtain quantitative indicators with unified measurement standards and time dimensions;

[0008] Step 3, set a multi-objective optimization problem with the goal of maximizing the overall benefits of all live broadcast rooms in the live broadcast matrix according to the quantitative indicators, specifically:

[0009] Step 3.1, the total funds invested in the exposure of the live broadcast room is B, for the live broadcast room Li Spend an exposure cost of x i For live broadcast room L i The unit exposure cost is s i This coefficient is determined by the traffic statistics results after the previous live broadcast. For live broadcast room L i The number of purchasable exposure times;

[0010] Step 3.2, Define the resource matrix R:

[0011]

[0012] Among them, r i Is the resource allocation coefficient of the i-th live broadcast room, satisfying Among them, P i Is the priority of live broadcast room L i Set based on user activity and income contribution factors; U i Is user participation; C i Is the load requirement of the live broadcast room, related to the content complexity and the number of audiences of the live broadcast room; α, β, and γ are adjustment factors that control the impact of priority, user participation, and load requirement on resource allocation; P i 、U i And C i Are obtained after a live broadcast, and the resource allocation coefficient of the live broadcast room before the next live broadcast is calculated through the collected historical data. The resource allocation coefficient r i Has a positive correlation with the unit exposure cost s i ;

[0013] Step 3.3, Use the effective exposure times Instead of the benefit function y i The optimization problem is to maximize the overall benefit of all live broadcast rooms in the live broadcast matrix. The objective function is:

[0014] All constraint conditions need to be satisfied:

[0015]

[0016] Step 4, Use the particle swarm optimization algorithm to solve the multi-objective optimization problem, find the optimal resource configuration for each live broadcast room through the particle swarm algorithm, and obtain an optimized allocation plan that maximizes the overall benefit of the live broadcast matrix;

[0017] Step 5, Display the quantization index and the optimized allocation plan to the decision maker through a visualization tool.

[0018] Preferably, the specific steps of step 2 are:

[0019] Step 2.1, collect data through the API interface opened by the live streaming platform, and use the Pandas library in Python to clean the collected data to remove duplicate, missing, and invalid data;

[0020] Step 2.2, extract features from the collected data, and use the Max-Min method to standardize each indicator so that the numerical ranges of each indicator are the same. The steps are as follows:

[0021] Set the data matrix A, which contains n samples and m features. a ij is the data in the i-th row and j-th column of the matrix, where i ∈ [1, m], j ∈ [1, n]. For the j-th column in the data matrix, calculate the minimum and maximum values of the feature:

[0022] min j = min(a ij ), max j = max(a ij )

[0023] Apply the formula: a ij ’ is the standardized data, and the value of each feature is scaled to the range of [0, 1].

[0024] Preferably, step 2 further selects valuable indicators for subsequent decision-making from the quantitative indicators. The specific steps are as follows:

[0025] Step 2.3, calculate the covariance matrix C of the data matrix A. The formula is After that, calculate the eigenvalues λ and eigenvectors v of the covariance matrix C, which satisfy the following equation: C·v = λ·v, where the eigenvector v represents the main direction of the data, and the eigenvalue λ represents the data variance size in that direction;

[0026] Step 2.4, select the eigenvectors corresponding to the k largest eigenvalues from the calculated eigenvalues, and project the data matrix onto the selected k principal components. The projection formula is: Z = A'·V k , where A’ is the quantized data matrix, and V k is the matrix composed of the k principal component eigenvectors selected, and Z is the data matrix after dimensionality reduction.

[0027] Preferably, further set the benefit function y i and there is a non-linear relationship with the effective exposure times : The objective function is:

[0028] where b i represents the basic benefit offset of the i-th live broadcast room.

[0029] Preferably, step 4 is specifically as follows:

[0030] Step 4.1: In the particle swarm, each particle represents a possible optimization solution, and the position of the particle represents the exposure times of each live broadcast room. The velocity v of the particle represents the change in the exposure times.

[0031] Step 4.2: The fitness of the particle is calculated by the total benefit function, that is:

[0032]

[0033] Step 4.3: Set the initial velocity of the particle to 0 so that the particle swarm algorithm explores the solution space at a slower speed in the initial iteration. The particle velocity determines the moving direction and step size of the particle in the solution space and will be updated in subsequent iterations. When the fitness of the particle is better than its historical optimal value, update the individual optimal solution p of the particle. best ; When the fitness of the particle is better than the global optimal value, update the global optimal solution g. best ; The particle velocity update formula and the position update formula are:

[0034]

[0035]

[0036] Where w is the inertia weight, c1 and c2 are acceleration constants, r1 and r2 are random numbers, t is the number of iterations. The particle swarm updates the positions of the particles through multiple iterations until the change in fitness is less than the convergence threshold or the number of iterations reaches the maximum, and the algorithm stops iterating to obtain the optimal optimization allocation scheme.

[0037] Preferably, further add step 6: According to the optimal optimization allocation scheme solved by the particle swarm algorithm in step 4, optimize the resource matrix R by analyzing historical data, continuously update the live broadcast data, analyze and update the optimization model to ensure that the fund allocation scheme and resource allocation are always adjusted based on the latest historical data and market dynamics.

[0038] A multi-platform live broadcast room data visualization and optimization decision support system, characterized in that it is used to implement the steps in the above-mentioned multi-platform live broadcast room data visualization and optimization decision support method, including a data collection module, a data visualization module, an allocation scheme optimization module, a data processing module, and a decision support module.

[0039] A computer-readable storage medium, on which a computer program is stored, characterized in that: when the program is executed by a processor, it implements the steps in the above-mentioned multi-platform live broadcast room data visualization and optimization decision support method.

[0040] A computer device includes a processor, a memory, and a computer program stored in the memory and executable on the processor. It is characterized in that when the processor executes the program, the steps in the above-mentioned multi-platform live broadcast room data visualization and optimization decision support method are implemented.

[0041] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: By analyzing live broadcast data through quantifying historical data, the system can visualize the key operation data of each platform, helping decision-makers comprehensively evaluate the live broadcast effects of different platforms. Through the visualization interface, decision-makers can intuitively feel the optimized effects and adjust the budget or other strategic parameters according to needs, improving operation efficiency.

[0042] At the same time, the system obtains the fund allocation plan with the maximum benefit through the multi-objective optimization application of the particle swarm algorithm. Through allocation optimization and intelligent decision support, decision-makers can flexibly adjust strategies in a dynamically changing market environment, avoid blind investment, and improve decision-making accuracy and overall operation benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The drawings are used to provide further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation to the present invention. In the drawings:

[0044] Figure 1 is the overall flow schematic diagram of the method provided by the present invention;

[0045] Figure 2 is the logical schematic diagram of the particle swarm optimization algorithm adopted by the present invention;

[0046] Figure 3 is the structural schematic diagram of the system provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0048] To make the above objects, features, and effects of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the drawings and specific embodiments.

[0049] Embodiment 1: A multi-platform live broadcast room data visualization and optimization decision support method, as Figure 1 shown, includes the following steps:

[0050] Step 1: Collect the real-time data and historical data of the live rooms on multiple platforms to construct a data matrix of multiple live rooms. Assume that N live rooms are being broadcast simultaneously, and collect and store the real-time data and historical data of each live room. The real-time data includes the number of exposures, the number of viewers, the interaction frequency, user behavior, etc. of each live room, and the historical data includes the viewer behavior data, sales conversion rate, live broadcast duration, etc. of the live room. Optimize the data format according to the different characteristics of the platform, and summarize the live room data into a central database to construct the live matrix L = [L1, L2,... L N T , which is convenient for subsequent processing and analysis.

[0051] Step 2: For the collected live room data, perform standardization processing through data preprocessing and feature extraction techniques to obtain quantitative indicators with a unified measurement standard and time dimension. The specific steps are as follows:

[0052] Step 2.1: Collect data through the API interface opened by the live broadcast platform, and use the Pandas library in Python to clean the collected data to remove duplicate, missing, and invalid data;

[0053] Step 2.2: Use the principal component analysis PCA algorithm to extract features from the collected data. According to the features extracted from the collected data, convert each feature into a specific quantitative indicator; use the Max-Min method to standardize the quantified indicators to ensure that the numerical ranges of the indicators are the same, which is convenient for subsequent comparison and analysis. Specifically:

[0054] Set the data matrix A, which contains n samples and m features, and a ij is the data in the i-th row and j-th column of the matrix, where i ∈ [1, m] and j ∈ [1, n]. For the j-th column in the data matrix, calculate the minimum and maximum values of the feature:

[0055] min j = min(a ij ), max j = max(a ij )

[0056] Apply the Max-Min standardization formula: a ij ’ is the standardized data, and the value of each feature is scaled to the range of [0, 1].

[0057] Furthermore, select the indicators valuable for subsequent decision-making from the above quantitative indicators. The specific steps are as follows:

[0058] Step 2.3: Calculate the covariance matrix C of the data matrix A, and the formula is ​After that, calculate the eigenvalues λ and eigenvectors v of the covariance matrix C, which satisfy the following equation: C·v = λ·v, where the eigenvector v represents the main direction of the data, and the eigenvalue λ represents the data variance in that direction;

[0059] Step 2.4, Select the eigenvectors corresponding to the largest k eigenvalues from the calculated eigenvalues because they can explain more variance in the data. Project the data matrix onto the selected k principal components. The projection formula is: Z = A'·V k , where A’ is the standardized data matrix, and V k is the matrix composed of the eigenvectors of the selected k principal components, and Z is the data matrix after dimensionality reduction.

[0060] Step 3, Based on the quantified data, set a multi-objective optimization problem with the goal of maximizing the overall benefit of all live rooms in the live matrix, and perform intelligent allocation of the resources of the live rooms to optimize the resource input and scheduling of the live rooms. Specifically:

[0061] Step 3.1, The total funds used for the exposure of the live room is B, and the exposure cost for the live room L i is x i , and the unit exposure cost of the live room L i is s i , and this coefficient is determined by the traffic statistics results after the previous live broadcast. is the number of exposures that can be purchased for the live room L i ;

[0062] Step 3.2, Define the resource matrix R:

[0063]

[0064] where r i is the resource allocation coefficient of the i-th live room, which satisfies where P i is the priority of the live room L i , which is set based on factors such as user activity and revenue contribution; U i is the user participation (number of viewers, interaction frequency, etc.); C i is the load requirement of the live room, which is related to the content complexity and the number of viewers of the live room. α, β, and γ are adjustment factors that control the impact of priority, user participation, and load requirement on resource allocation; P i , U i and C i are obtained after a live broadcast, and the resource allocation coefficient of the live room before the next live broadcast is calculated through the collected historical data. The resource allocation coefficient r i is related to the unit exposure cost s iThere is a positive correlation;

[0065] Step 3.3, live broadcast room L i The benefit function is y i , to simplify the model, the effective exposure times are introduced to replace the benefit function y i , since in actual situations, the effective exposure times of the live broadcast room and the benefits usually show a non-linear relationship, so it is further assumed that the benefit function y i and the effective exposure times have the following relationship: where b i represents the basic benefit offset of the i-th live broadcast room to ensure that the effective exposure times can accurately replace the benefit function;

[0066] Step 3.4, the optimization problem is to maximize the overall benefits of all live broadcast rooms in the live broadcast matrix, and the objective function is:

[0067]

[0068] All constraint conditions need to be satisfied:

[0069]

[0070] Step 4, use the particle swarm optimization algorithm to solve this multi-objective optimization problem. The particle swarm algorithm is used to find the optimal resource allocation for each live broadcast room, and an optimization allocation scheme that maximizes the overall benefits of the live broadcast matrix is obtained, as Figure 2 shown. The specific steps are as follows:

[0071] Step 4.1, in the particle swarm, each particle represents a possible optimization scheme, and the position of the particle represents the exposure times of each live broadcast room The velocity v of the particle represents the change amount of the exposure times;

[0072] Step 4.2, the fitness of the particle is calculated by the total benefit function, that is:

[0073]

[0074] Step 4.3, set the initial velocity of the particle to 0 so that the particle swarm algorithm can explore the solution space at a slower speed in the initial iteration. The particle velocity determines the moving direction and step size of the particle in the solution space and will be updated in subsequent iterations; in actual scheduling, the velocity vector can control the change range of the optimization scheme; in the case of tight funds, set a smaller initial velocity to ensure that the optimization scheme is adjusted more precisely; while when the funds are relatively abundant, a larger initial velocity can be set to speed up the search progress; when the fitness of the current particle is better than its historical optimal value, update the individual optimal solution p of the particle best; The fitness of the current particle is better than the global optimal value, and the global optimal solution g of the population is updated best ; Particle velocity update formula and position update formula:

[0075]

[0076] where w is the inertia weight, c1 and c2 are acceleration constants, r1 and r2 are random numbers, and t is the number of iterations. The particle swarm updates the positions of the particles through multiple iterations until the change in fitness is less than the convergence threshold or the number of iterations reaches the maximum, at which point the algorithm stops iterating and obtains the optimal optimization allocation scheme.

[0077] Step 5: Use a visualization tool to display the quantified data and the optimized allocation scheme to the decision maker. The data display includes but is not limited to real-time charts, trend analysis, comparison of allocation schemes, etc. The decision maker can view the real-time data, benefit prediction, and optimized fund allocation scheme of each platform through an interactive interface. This visualization interface allows the decision maker to query data, adjust strategies, and view the optimization effect in real time according to different needs, thus helping the decision maker make more reasonable operation decisions. In the present invention, the Matplotlib and Seaborn libraries are imported in Python to draw charts.

[0078] Step 6: According to the optimal optimization allocation scheme solved by the particle swarm algorithm in Step 4, optimize the resource allocation matrix R by analyzing historical data, continuously update the live data, analyze and update the optimization model to ensure that the fund allocation scheme and resource configuration are always adjusted based on the latest historical data and market dynamics. Through the dynamic optimization process, the decision maker can track the benefit changes of each platform according to the obtained optimal results, and continuously optimize the resource input and allocation efficiency of the live broadcast room to adjust the decision.

[0079] Embodiment 2: A multi-platform live broadcast room data visualization and optimization decision support system, as Figure 3 shown, including a data collection module, a data visualization module, an allocation scheme optimization module, a data processing module, and a decision support module, where:

[0080] Data collection module: Collect multi-platform live broadcast historical data through the open API of the live broadcast platform, construct a multi-dimensional benefit evaluation index system based on heterogeneous data sources, combine real-time interaction data, and form multi-dimensional indicators according to user participation, commodity conversion rate, and server system load to evaluate the priority of the live broadcast rooms in different platforms when allocating resources;

[0081] Data visualization module: After standardizing the collected data, display it in an intuitive way such as charts and trend analysis to assist the allocation decision in a quantitative manner;

[0082] Allocation Scheme Optimization Module: Using the live broadcast room exposure times as the benefit proxy index, introducing the benefit function of the live broadcast room to model the fund optimization problem as a multi-objective optimization model, and sorting out a visual allocation scheme through the computer to solve the optimal allocation strategy with the maximization of the benefit function as the objective function;

[0083] Data Processing Module: Standardize the collected data in Python; and apply the particle swarm optimization algorithm according to the multi-objective optimization model to iteratively solve the exposure times of the live broadcast room that maximize the benefit.

[0084] Decision Support Module: According to the solution results of the particle swarm algorithm, optimize according to the optimal exposure times of the live broadcast rooms on different platforms to assist the users of the system in making decisions.

[0085] Embodiment 3: The computer-readable storage medium of this embodiment stores a computer program, and when the program is executed by a processor, it implements the steps in a multi-platform live broadcast room data visualization and optimization decision support method of Embodiment 1.

[0086] The computer-readable storage medium of this embodiment can be the internal storage unit of the terminal, such as the hard disk or memory of the terminal; the computer-readable storage medium of this embodiment can also be the external storage device of the terminal, such as the plug-in hard disk, smart memory card, secure digital card, flash card, etc. equipped on the terminal; further, the computer-readable storage medium can also include both the internal storage unit and the external storage device of the terminal.

[0087] The computer-readable storage medium of this embodiment is used to store the computer program and other programs and data required by the terminal, and the computer-readable storage medium can also be used to temporarily store the data that has been output or will be output.

[0088] Embodiment 4: The computer device of this embodiment includes a processor, a memory, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the steps in a multi-platform live broadcast room data visualization and optimization decision support method of Embodiment 1.

[0089] In this embodiment, the processor can be a central processing unit, or other general-purpose processors, digital signal processors, application-specific integrated circuits, off-the-shelf programmable gate arrays, or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.; the memory can include a read-only memory and a random access memory, and provide instructions and data to the processor. A part of the memory can also include a non-volatile random access memory. For example, the memory can also store information about the device type.

[0090] Those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solutions, in essence, or the parts that contribute to the prior art can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disks, optical disks, etc., and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0091] The foregoing are only the preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for visualizing multi-platform live broadcast room data and optimizing decision-making support, characterized in that, It includes the following steps: Step 1: Collect the real-time data and historical data of live rooms on multiple platforms, and construct a data matrix of multiple live rooms; Step 2: For the collected live room data, perform standardization processing through data preprocessing and feature extraction techniques to obtain quantitative indicators with unified measurement standards and time dimensions; Step 3: According to the quantitative indicators, set a multi-objective optimization problem with the goal of maximizing the overall benefits of all live rooms in the live matrix. Specifically: Step 3.1, the total funds invested in the live broadcast room for exposure is B, for the live broadcast room L i The exposure cost is x i , the live broadcast room L i The unit exposure cost is s i , this coefficient is determined by the traffic statistics results after the end of the previous live broadcast, is the number of exposures that can be purchased for the live broadcast room L i ; Step 3.2: Define the resource matrix R: Among them, r i is the resource allocation coefficient of the i-th live broadcast room, satisfying r i =α·P i +β·U i +γ·C i , where P i is the priority of the live broadcast room L i , which is set based on factors such as user activity and revenue contribution; U i is the user participation rate; C i is the load requirement of the live broadcast room, which is related to the content complexity and the number of audiences of the live broadcast room; α, β, and γ are adjustment factors that control the influence of priority, user participation rate, and load requirement on resource allocation; P i , U i and C i are obtained after a live broadcast. The resource allocation coefficient of the live broadcast room before the next live broadcast is calculated through the collected historical data. There is a positive correlation between the resource allocation coefficient r i and the cost per unit exposure s i . Step 3.3, using the effective number of exposures to replace the benefit function y i , the optimization problem is to maximize the overall benefit of all live rooms in the live matrix, and the objective function is: All constraints need to be satisfied: Step 4: Use the particle swarm optimization algorithm to solve the multi-objective optimization problem. Through the particle swarm algorithm, find the optimal resource allocation for each live room to obtain an optimized allocation plan that maximizes the overall benefits of the live matrix; Step 5: Use a visualization tool to display the quantitative indicators and the optimized allocation plan to the decision maker.

2. The multi-platform live broadcast room data visualization and optimization decision support method according to claim 1, wherein The specific steps of Step 2 are: Step 2.1: Collect data through the API interface opened by the live platform, and use the Pandas library in Python to clean the collected data to remove duplicate, missing, and invalid data; Step 2.2: Extract features from the collected data, and use the Max-Min method to perform standardization processing on each indicator to make the numerical ranges of each indicator consistent. The steps are: Set the data matrix A, which contains n samples and m features, where a ij is the data at the i-th row and j-th column in the matrix, i ∈ [1, m], j ∈ [1, n]. For the j-th column in the data matrix, calculate the minimum and maximum values of the feature: min j = min(a ij ), max j = max(a ij ) Application formula: a ij ’ is the standardized data, and the value of each feature is scaled to the range of [0, 1].

3. A method for multi-platform live broadcast room data visualization and optimization decision support according to claim 2, characterized in that, Step 2 further selects valuable indicators for subsequent decision-making from the quantitative indicators. The specific steps are: Step 2.3, calculate the covariance matrix C of the data matrix A, and the formula is After that, calculate the eigenvalues λ and eigenvectors v of the covariance matrix C, which satisfy the following equation: C·v = λ·v, where the eigenvector v represents the main direction of the data, and the eigenvalue λ represents the magnitude of the data variance in that direction; Step 2.4: Select the eigenvectors corresponding to the largest k eigenvalues from the calculated eigenvalues, and project the data matrix onto the selected k principal components. The projection formula is: Z = A'·V k , where A’ is the quantized data matrix, and V k is the matrix composed of the eigenvectors of the selected k principal components, and Z is the data matrix after dimensionality reduction.

4. A method for multi-platform live broadcast room data visualization and optimization decision support according to claim 1, characterized in that, Further set the benefit function y i and the number of effective exposures There is a non-linear relationship: The objective function is: where b i represents the basic benefit offset of the i-th live streaming room.

5. A method for multi-platform live broadcast room data visualization and optimization decision support according to claim 4, characterized in that, The specific content of Step 4 is: Step 4.1, in the particle swarm, each particle represents a possible optimization solution, and the position of the particle represents the number of exposures of each live broadcast room. The velocity v of the particle represents the change in the number of exposures. Step 4.2: The fitness of the particle is calculated by the total benefit function, that is: Step 4.3, set the initial velocity of the particles to 0 so that the particle swarm algorithm explores the solution space at a slower speed in the initial iteration. The particle velocity determines the moving direction and step size of the particles in the solution space and will be updated in subsequent iterations. When the fitness of a particle is better than its historical optimal value, update the individual optimal solution p of the particle best ; when the fitness of a particle is better than the global optimal value, update the global optimal solution g best ; the particle velocity update formula and the position update formula are as follows: Where w is the inertia weight, c1 and c2 are acceleration constants, r1 and r2 are random numbers, t is the number of iterations. The particle swarm updates the position of the particle through multiple iterations until the change in fitness is less than the convergence threshold or the number of iterations reaches the maximum, and the algorithm stops iterating to obtain the optimal optimized allocation plan.

6. A method for multi-platform live broadcast room data visualization and optimization decision support according to claim 1, characterized in that Further add Step 6: According to the optimal optimized allocation plan solved by the particle swarm algorithm in Step 4, optimize the resource matrix R by analyzing the historical data, continuously update the live data, analyze and update the optimization model to ensure that the fund allocation plan and resource allocation are always adjusted based on the latest historical data and market dynamics.

7. A multi-platform live broadcast room data visualization and optimization decision support system, characterized in that It is used to implement the steps in a multi-platform live room data visualization and optimization decision support method described in any one of claims 1-6, including a data collection module, a data visualization module, an allocation plan optimization module, a data processing module, and a decision support module.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, it implements the steps in a multi-platform live room data visualization and optimization decision support method described in any one of claims 1-6.

9. A computer device, comprising a processor, a memory, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in a multi-platform live room data visualization and optimization decision support method described in any one of claims 1-6.

Images