Interactive information interaction system based on digital media technology

Abstract

The invention relates to the technical field of human-computer interaction optimization, and discloses an interactive information interaction system based on a digital media technology. The system comprises a data acquisition module, a preprocessing module, a behavior analysis module, a region division module, a quality evaluation module and a resource regulation module. The data acquisition module acquires equipment operation frequency, interface response delay and a contact coordinate sequence; the preprocessing module performs normalization processing to generate standardized parameters; the behavior analysis module constructs an interactive behavior data flow through space-time correlation analysis, and marks an abnormal interval; the region division module dynamically generates an interaction sensitive region weight distribution diagram by combining an environment noise threshold value; the quality evaluation module calculates the coupling relation between the contact density and the response delay to generate an evaluation matrix; and the resource regulation and control module adjusts the transmission priority and redistributes the computing resources according to the transmission priority. According to the invention, abnormal interaction behaviors can be accurately identified, adaptive resource allocation is realized, and system interaction fluency and response accuracy are improved.

Description

Technical Field

[0001] This invention relates to the field of human-computer interaction optimization technology, specifically to an interactive information system based on digital media technology. Background Technology

[0002] In digital media interactive applications, such as touchscreen terminals, virtual reality environments, and large interactive installations, ensuring the smoothness and accuracy of the interaction process is crucial for improving user experience. Existing technologies typically rely on independent monitoring and threshold alarm mechanisms for basic performance indicators such as device operating frequency and interface response latency. These methods analyze interactive behavior in isolation, separating the temporal and spatial dimensions, and can only identify single performance indicator anomalies that significantly exceed preset thresholds. Existing technologies lack the ability to effectively detect and differentiate abnormal interaction patterns hidden within normal data streams, caused by complex user behavior sequences or subtle environmental changes, or specific areas of reduced response quality due to localized environmental interference.

[0003] Existing technical solutions have shortcomings. At the behavioral analysis level, due to a lack of in-depth exploration of the spatiotemporal correlation of interactive behaviors, the system struggles to accurately depict the quality profile of a complete interactive session and cannot effectively distinguish between user misoperation, momentary device performance fluctuations, and genuine interactive malfunctions. This results in delayed and coarse judgments of interactive anomalies, often triggering a response only after the user has already perceived a lag or malfunction, lacking predictability. In terms of interactive area management, most systems adopt a pre-static approach to dividing sensitive areas, with fixed area boundaries and weights. This method cannot adapt to the dynamic fluctuations of ambient noise during device operation, nor can it be optimized based on the evolution of actual user interaction habits. This leads to potential resource overcapacity in low-noise environments, while insufficient resource allocation in high-noise interference areas exacerbates the deterioration of interactive quality, resulting in insufficient overall system adaptability and robustness. Summary of the Invention

[0004] The purpose of this invention is to provide an interactive information system based on digital media technology to solve the problems mentioned in the background.

[0005] To achieve the above objectives, the present invention provides an interactive information system based on digital media technology, the system comprising: The data acquisition module is used to collect raw operating data generated by user interaction devices and extract device operating frequency, interface response delay and interactive touch point coordinate sequence; The data preprocessing module is used to normalize the raw operating data, eliminate the influence of dimensions caused by equipment differences, and generate a standardized set of equipment operating parameters. The behavior analysis module is used to construct an interactive behavior data stream based on the standardized equipment operating parameter set through spatiotemporal correlation analysis, mark abnormal interaction intervals, and calculate the continuity index of the behavior trajectory. The region segmentation module is used to dynamically segment interactive sensitive regions and generate a region weight distribution map based on the interactive behavior data stream and a preset environmental noise threshold. The quality assessment module is used to spatially reconstruct the coordinate sequence of interactive touch points using the regional weight distribution map, calculate the coupling relationship between touch point distribution density and interface response delay, and generate an interactive quality assessment matrix. The resource regulation module is used to adjust the priority configuration of information transmission paths, reallocate computing resources, and update the regulation strategy of device operating frequency based on the interaction quality assessment matrix.

[0006] Preferably, the process of generating the standardized equipment operating parameter set includes: Establish a reference coordinate system for equipment operating parameters and resample equipment data at different sampling frequencies to a unified time axis; Calculate the parameter offset for each sampling point, and use the sliding window method to smooth the offset; The filtered parameter sequence is standardized to conform to a normal distribution. Establish a rule base for association between parameters and remove abnormal data points that do not conform to the association rules; Feature extraction is performed on the cleaned parameter sequence to generate a standardized set of equipment operating parameters containing time-domain and frequency-domain features.

[0007] Preferably, the method for constructing the interactive behavior data stream includes: Based on the standardized equipment operating parameter set, the spatiotemporal feature vector of the user's operation trajectory is extracted; Calculate the similarity matrix between feature vectors and identify typical interaction patterns through cluster analysis; Establish an interactive behavior state transition model and label the key nodes of state transition; The abnormal interaction intervals are segmented into multiple scales, and the rate of change of curvature of the behavioral trajectory within each interval is calculated. By combining historical interaction data to train a behavior prediction model, an interaction behavior data stream with time-series labels is generated.

[0008] Preferably, the step of generating the regional weight distribution map includes: Based on the abnormal interaction intervals in the interactive behavior data stream, the sensitive frequency bands of environmental noise are determined; Construct a noise impact assessment function based on a sound field propagation model to calculate the degree of noise interference in each area; The interactive interface is divided into several sub-regions using an adaptive grid partitioning method; Calculate the regional importance weighting coefficient based on the interaction frequency and noise interference level of each sub-region; A regional weight distribution map with spatial resolution is generated by combining weighting coefficients.

[0009] Preferably, the method for calculating the interaction quality evaluation matrix includes: Based on the regional weight distribution map, extract the contact point distribution feature values ​​of each sub-region; Calculate the correlation coefficient matrix between contact point distribution density and interface response delay; Establish a multi-dimensional evaluation index system for interaction quality, including response timeliness, trajectory smoothness, and operational accuracy; Principal component analysis was used for dimensionality reduction to extract key quality feature vectors. The quality feature vectors are weighted and fused with regional weights to generate an interactive quality assessment matrix.

[0010] Preferably, the priority configuration method for the information transmission path includes: Analyze the quality feature vectors in the interaction quality assessment matrix to identify key transmission bottleneck areas; Establish a transmission path performance evaluation model to calculate the load capacity and transmission delay of each path; Adjust the path selection strategy according to the regional weight distribution map, and prioritize data transmission in high-weight regions; Dynamically allocate bandwidth resources and establish a differential service mechanism based on quality requirements; Generate a configuration scheme for information transmission paths with priority tags.

[0011] Preferably, the reallocation process of computing resources includes: Monitor the current resource usage status of the system, including processor load, memory usage, and network bandwidth usage; Based on the information transmission path configuration scheme, predict the peak resource demand for each path; Establish a resource allocation optimization model with overall system load balancing as the objective function; A heuristic algorithm is used to find the optimal resource allocation scheme, with a margin reserved for emergency handling. Implement dynamic resource scheduling and adjust the resource quotas of each process in real time.

[0012] Preferably, the method for updating the device operating frequency control strategy includes: The actual operating parameters of the data acquisition device after the reallocation of computing resources; Compare the deviation between the actual operating parameters and the expected target, and calculate the control error coefficient; Establish a fuzzy control rule base for equipment operating frequency and formulate a frequency adjustment step size strategy; The response sensitivity of frequency control is dynamically adjusted based on real-time load change trends. Generate a table of device operating frequency control strategies with adaptive characteristics.

[0013] Preferably, the process of establishing a transmission path performance evaluation model and calculating the load capacity and transmission delay of each path includes: Periodically measure the data packet transmission rate, packet loss rate, and round-trip time for each transmission path; The effective load capacity of each path is calculated based on the packet transmission rate and packet loss rate. The effective load capacity is the product of the maximum theoretical transmission rate and the available bandwidth coefficient, which is adjusted based on the packet loss rate. The transmission delay for each path is calculated based on the round-trip time and jitter value. The transmission delay is the average round-trip time plus the jitter compensation value. A transmission path performance evaluation model is established. The model inputs are effective load capacity and transmission delay, and the output is the path performance score. The path performance score is positively correlated with effective load capacity and negatively correlated with transmission delay. Transmission paths are sorted based on path performance scores for priority configuration.

[0014] Preferably, the process of establishing a resource allocation optimization model with overall system load balancing as the objective function includes: Real-time acquisition of operational status data for each computing node in the system, including processor utilization, memory usage, and network bandwidth utilization; Calculate the real-time load factor for each computing node. The load factor is a weighted sum of processor utilization, memory usage, and network bandwidth utilization. Define the objective function for overall system load balancing. The objective function is the variance of the load coefficients of all computing nodes, and the goal is to minimize this variance. Set resource allocation constraints, including the maximum processing power, memory capacity, and bandwidth limit for each computing node; An iterative optimization method is used to adjust the resource allocation scheme. The objective function value is recalculated in each iteration until the change in the objective function value is less than a set threshold. The iteration stops and the optimal resource allocation scheme is output.

[0015] Compared with the prior art, the beneficial effects of the present invention are: Spatiotemporal correlation analysis technology constructs a unified interactive behavior data stream by fusing interface response latency over time with the coordinates of interactive touchpoints in a spatial sequence. This method can calculate continuity indicators describing the coherence of behavioral trajectories, thereby identifying anomalies that traditional single-dimensional monitoring cannot detect. This enables the system to move from macroscopic performance monitoring to microscopic behavioral pattern diagnosis, achieving early and accurate localization of interaction smoothness issues and improving the sensitivity and foresight of anomaly detection.

[0016] Dynamic region partitioning technology based on environmental noise thresholds allows the boundaries and weights of interaction-sensitive areas to be adaptively adjusted based on real-time device operating data and environmental noise levels, rather than being statically preset. By continuously sensing changes in environmental interference, the system dynamically generates a region weight distribution map, thereby precisely directing computing and monitoring resources to the areas where interaction quality is most critical. This mechanism solves the inherent problem of resource allocation mismatch with the real-time environment in fixed partitioning modes, and can continuously maintain the stability and accuracy of system response when facing dynamically changing external interference.

[0017] By applying dynamically generated regional weight distribution maps to the spatial reconstruction of interactive touchpoint sequences, the quality assessment module can establish a quantitative coupling relationship between touchpoint distribution density and interface response latency. This method directly correlates the thermal distribution of user behavior with the system's response performance at specific spatial locations, giving the quality assessment results a clear spatial orientation. This provides a precise basis for subsequent resource regulation, ensuring that strategies for reallocating computing resources and adjusting device operating frequencies can directly target identified performance bottleneck areas, improving the targeting of resource scheduling and the overall system's energy efficiency. Attached Figure Description

[0018] Figure 1 This is a schematic diagram illustrating the working principle of the interactive information system based on digital media technology described in this invention. Figure 2 Flowcharts generated for standardized equipment operating parameter sets; Figure 3 A flowchart for generating a regional weight distribution map. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please see Figure 1This invention provides an interactive information system based on digital media technology. The system includes: a data acquisition module that first establishes a connection with the user's interactive device and collects raw operating data generated by the device in real time. This data includes, but is not limited to, device operating frequency, interface response latency, and the coordinate sequence of interactive touchpoints generated by user operations. The raw data is then sent to a data preprocessing module, which normalizes the heterogeneous data from devices of different specifications to eliminate inconsistencies in dimensions caused by differences in device hardware, ultimately generating a standardized set of device operating parameters with unified standards. A behavior analysis module receives this parameter set and performs in-depth spatiotemporal correlation analysis on it, thereby constructing an interactive behavior data stream that reflects the continuity of user operations. During this process, the module identifies and marks abnormal interaction intervals that may have problems and calculates the continuity index of the behavior trajectory. A region division module, based on the constructed interactive behavior data stream and combined with a preset environmental noise threshold, dynamically divides regions sensitive to interactive behavior responses and generates a region weight distribution map indicating the importance of each region. The quality assessment module uses this regional weight distribution map to perform spatial reconstruction analysis on the original interactive touchpoint coordinate sequence, calculates the intrinsic coupling relationship between the density of touchpoint distribution and interface response latency, and generates an interaction quality assessment matrix that can quantitatively evaluate interaction quality. The resource control module analyzes this assessment matrix and adjusts the priority configuration of different information transmission paths in the system accordingly, reallocates computing resources, and updates the control strategy for the operating frequency of user interaction devices, thus forming a closed-loop optimization system from perception to control.

[0021] Example 1: See Figure 2The data preprocessing module receives raw operational data from the data acquisition module. This raw operational data includes device operating frequency, interface response latency, and interactive touchpoint coordinate sequences directly read from the user interaction device. The first step in generating a standardized set of device operating parameters is to establish a reference coordinate system for the device operating parameters. This reference coordinate system uses the system initialization completion time as the absolute time origin, uniformly converting the timestamps of all input data to a millisecond-level time axis starting from this origin. For device data with varying sampling frequencies, a linear resampling algorithm is used to map the data sequence to a unified, fixed-frequency, high-precision time axis. The linear resampling algorithm ensures strict alignment of data points from different sources in the time dimension through interpolation calculations. In specific implementation, the parameter offset of each sampling point relative to its preset calibration reference value is calculated. The device operating frequency offset is the difference between the instantaneous frequency value and the rated frequency value specified in the device specifications, and the interface response latency offset is the difference between the actual measured latency and the theoretical minimum latency value of the system. A sliding window with configurable window length is used to smooth the parameter offset sequence. The sliding window moves point by point on the time axis, and the arithmetic mean of all parameter offsets within the window is calculated as the filtered output value at the window center. The smoothed parameter offset sequence is then standardized using the Z-score standardization method. The Z-score standardization method uses the arithmetic mean and standard deviation of the sequence to calculate the standard normal distribution with a mean of zero and a standard deviation of one.

[0022] A rule base for parameter association is established, which predefines the constraints and reasonable ranges that different device operating parameters should satisfy under normal interaction conditions. For example, a rule might stipulate that when the movement speed of the interactive touch point exceeds a certain threshold, the device operating frequency should not be lower than a corresponding value. The parameter combinations of each sampling point are checked, and abnormal data points that clearly violate the physical logic or statistical laws defined in the rule base are removed. Features are extracted from the cleaned and standardized parameter sequences. Time-domain features include calculating the mean, variance, and zero-crossing rate of the sequence within a sliding window, while frequency-domain features are extracted by performing a Fast Fourier Transform on the sequence to obtain the amplitude and phase information of the main frequency components. The final standardized device operating parameter set is a structured dataset containing the timestamp of each sampling point, the processed parameter values, and the extracted time-domain and frequency-domain features. The behavior analysis module receives the standardized device operating parameter set as input and constructs an interactive behavior data stream based on it. The spatiotemporal feature vector of the user's operation trajectory is extracted, which includes the interactive touch point coordinate sequence, the instantaneous velocity vector of the touch point movement, the acceleration vector, and the time interval between consecutive touch points. The similarity between spatiotemporal feature vectors corresponding to consecutive operational behaviors is calculated, and the difference between the vectors is measured using Euclidean distance to construct a similarity matrix reflecting the similarity relationship between different operational behaviors. Typical interaction patterns are identified through cluster analysis. The cluster analysis algorithm groups the similarity matrix, grouping highly similar operational behaviors into the same category. Each category represents a typical interaction pattern, such as a fast swipe pattern or a fine-tap pattern.

[0023] In practical implementation, an interactive behavior state transition model is established. This model uses a Hidden Markov Model (HMM) to represent the transition probabilities between different typical interaction modes. The states of the HMM correspond to different typical interaction modes. Key nodes of state transitions are marked; these key nodes are the moments when the state changes significantly as calculated by the HMM, often indicating a shift in the user's interaction intent. Abnormal interaction intervals are segmented using multiple scales. These abnormal interaction intervals are data segments previously marked as potentially problematic during parameter preprocessing. Multi-scale segmentation uses time windows of different lengths to divide these intervals. The rate of change of curvature of the behavior trajectory within each segment is calculated. This rate of change is obtained by calculating the derivative of the curvature of the interaction touchpoint's movement path over time; an abnormally high rate of change of curvature usually indicates unsmooth operation. A behavior prediction model is trained using historical interaction data. This historical interaction data consists of labeled interaction sequences accumulated over long-term system operation. The behavior prediction model employs a Long Short-Term Memory (LSTM) network structure, which can learn the temporal dependencies of interactive behaviors. The trained Long Short-Term Memory (LSTM) network model can predict future behavioral trends in the short term based on current and past interaction data, ultimately generating an interaction behavior data stream with time-series labels. The interaction behavior data stream includes identifiers of normal and abnormal interaction intervals, labels of typical interaction patterns, state transition node information, and behavior prediction results.

[0024] In some embodiments, the length of the sliding window can be dynamically adjusted according to the response speed requirements of the actual application scenario. For real-time interactive scenarios requiring fast response, a shorter sliding window length is set to reduce the latency introduced by filtering. In some embodiments, the clustering analysis algorithm can be either the K-means algorithm or the DBSCAN algorithm. The choice depends on whether the number of typical interaction patterns is known in advance. When the number of patterns is unknown, the DBSCAN algorithm is used for density clustering. It is understood that establishing a reference coordinate system is the foundation for data time synchronization, and the accuracy of the reference coordinate system directly affects the accuracy of subsequent spatiotemporal correlation analysis. It is also understood that the completeness of the association rule base is crucial for the identification of outlier data points, and the association rule base needs to be continuously updated and improved as the system deployment environment expands.

[0025] Example 2: See Figure 3The region segmentation module receives interactive behavior data streams from the behavior analysis module. These data streams contain interactive interval information marked with anomalies. Determining the sensitive frequency bands for environmental noise requires analyzing additional environmental sensor data within the abnormal interactive intervals. This data may originate from microphone arrays or vibration sensors. By performing a Fast Fourier Transform on the environmental sensor data within the abnormal intervals, frequency components in the power spectrum significantly higher than the background level are identified; these frequency components constitute the sensitive frequency bands for environmental noise. A noise impact assessment function based on a sound field propagation model is constructed. This model abstracts the interactive interface as a two-dimensional plane and assumes the existence of one or more point noise sources. The noise impact assessment function calculates the noise intensity received at any point on the interface. The noise intensity is directly proportional to the noise source intensity and inversely proportional to the square of the propagation distance, taking into account the sound absorption coefficient of the interface material. It is understandable that the accuracy of the noise impact assessment function depends on prior knowledge or real-time estimation of the noise source location and characteristics. In the absence of direct measurement, statistical estimates based on historical data can be used.

[0026] In implementation, an adaptive mesh generation method is used to initially divide the rectangular interactive interface into uniformly sized square sub-regions. The initial mesh size is set based on the physical dimensions of the interactive interface and the expected operational precision. Two key indicators for each initial sub-region within a recent time window are calculated: interaction frequency and noise interference level estimate. Interaction frequency is the number of effective interaction events occurring within the sub-region per unit time, and the noise interference level estimate is calculated using a noise impact assessment function. The mesh generation is dynamically adjusted based on the interaction frequency and noise interference level estimate for each sub-region. Sub-regions with high interaction frequency and high noise interference level estimate are refined, further subdivided into smaller sub-regions to achieve higher spatial resolution; sub-regions with low interaction frequency and low noise interference level estimate may remain unchanged or undergo mesh merging. The regional importance weight coefficient for each finally determined sub-region is calculated using a weighted formula that comprehensively considers interaction frequency and noise interference level. Sub-regions with higher interaction frequency and lower noise interference level have a larger regional importance weight coefficient. The final generated regional weight distribution map is a two-dimensional matrix. The row and column indices of the matrix correspond to the coordinate numbers of the sub-regions, and the matrix element values ​​store the regional importance weight coefficients of the corresponding sub-regions.

[0027] In practical implementation, the quality assessment module utilizes a regional weight distribution map to perform spatial reconstruction analysis of the interactive touchpoint coordinate sequence. It extracts touchpoint distribution characteristic values ​​for each sub-region, including the total number of touchpoints within the sub-region, the variance of touchpoint coordinates, and the entropy value of the touchpoint distribution. The entropy value reflects the uniformity of touchpoint distribution within the sub-region. The coupling relationship between touchpoint distribution density and interface response latency is calculated. Touchpoint distribution density is the number of touchpoints per unit area within the sub-region, and interface response latency is the average delay time of all interactive events within the sub-region. The Pearson correlation coefficient between the two is calculated to form a correlation coefficient matrix. A multi-dimensional evaluation index system for interaction quality is established, comprising three main dimensions: response timeliness is measured by average latency and latency jitter; trajectory smoothness is measured by the curvature continuity of the touchpoint movement path; and operation accuracy is measured by the root mean square deviation between the touchpoint and the target position. Principal component analysis (PCA) is used to reduce the dimensionality of the multi-dimensional evaluation indexes. PCA transforms multiple potentially correlated indicators into a few uncorrelated principal components, which retain most of the variation information of the original data. The extracted key quality feature vectors are the principal components with the highest variance contribution rates, and the key quality feature vectors comprehensively reflect the core aspects of interaction quality.

[0028] In practice, generating the interaction quality assessment matrix involves the fusion of spatial and quality dimensions. The key quality feature vectors are weighted and fused with the regional importance weight coefficients from the regional weight distribution map. This weighting is achieved using matrix multiplication; the row vector formed by the key quality feature vectors is multiplied by the column vector formed by the regional importance weight coefficients, resulting in a comprehensive score scalar. For each sub-region of the interaction interface, such a comprehensive score is calculated. The comprehensive scores of all sub-regions are arranged according to their spatial location to form a two-dimensional matrix, which is the interaction quality assessment matrix. The rows and columns of the interaction quality assessment matrix correspond to the spatial coordinates of the sub-regions, and the matrix element values ​​represent the comprehensive interaction quality score at that spatial location. The mathematical expression of the interaction quality assessment matrix can be represented as: Where: symbol The symbol represents the quality assessment value of the sub-region located in the i-th row and j-th column of the interaction quality assessment matrix. The number of dimensions representing the key quality feature vector, with the sign... The weighting coefficient representing the k-th quality characteristic, with the sign... Represents the value of the k-th key quality feature vector extracted from the sub-region in the i-th row and j-th column, with the sign... The regional importance weight coefficient of the sub-region in the i-th row and j-th column of the regional weight distribution map represents the regional importance weight coefficient.

[0029] In some embodiments, the analysis of sensitive frequency bands for environmental noise can combine data from multiple sensors, and cross-validation can be used to improve the accuracy of frequency band identification. In some embodiments, the triggering condition for adaptive grid partitioning can be set to automatically perform refinement when the interaction frequency within a sub-region exceeds twice the global average frequency. Optionally, the noise impact assessment function can consider the time-varying characteristics of noise, and introduce a time decay factor to give higher weight to noise data from the most recent moment. Optionally, the construction of a multi-dimensional evaluation index system can introduce user subjective ratings as a supervisory signal, and determine the weight of each objective index through regression analysis. The interaction quality assessment matrix provides a quantitative decision-making basis for subsequent resource regulation, and the areas with lower scores in the interaction quality assessment matrix identify the spatial locations where the system needs to be prioritized for optimization. The dynamic generation mechanism of the regional weight distribution map enables the system to adapt to different usage scenarios and user habits, focusing on areas that users frequently use and are susceptible to interference.

[0030] Example 3: The resource control module receives the interaction quality assessment matrix from the quality assessment module. This matrix quantifies the comprehensive interaction quality score for different spatial regions of the interface. The module analyzes the quality feature vectors within the matrix, which are key principal components extracted from the original multidimensional evaluation indicators through principal component analysis. Key transmission bottleneck regions are identified; these are sets of sub-regions in the interaction quality assessment matrix whose scores are significantly below the average level, typically corresponding to shortcomings in user experience. Establishing a transmission path performance evaluation model requires acquiring real-time performance data for each candidate transmission path. This performance data is obtained by periodically sending probe data packets and statistically analyzing the feedback results. The probe data packets are sent to the target node at fixed time intervals via different network paths.

[0031] In practical implementation, the data packet transmission rate, packet loss rate, and round-trip time (RTT) of each transmission path are periodically measured. The data packet transmission rate is the amount of data successfully transmitted per unit time; the packet loss rate is the proportion of data packets sent without receiving an acknowledgment; and the RTT is the time difference between sending and receiving an acknowledgment signal. The effective load capacity of each path is calculated based on the data packet transmission rate and packet loss rate. The effective load capacity is the product of the path's theoretical maximum transmission rate and an available bandwidth coefficient dynamically adjusted based on the current packet loss rate. The available bandwidth coefficient reflects the path's reliability; the higher the packet loss rate, the smaller the available bandwidth coefficient. The transmission delay of each path is calculated based on the RTT and jitter value. The transmission delay is the arithmetic mean of multiple RTT measurements plus a jitter compensation value. The jitter compensation value is typically a multiple of the standard deviation of the RTT to reflect the volatility of network delay. A transmission path performance evaluation model is established. The input parameters of the model are the effective load capacity and transmission delay of each path, and the output is a path performance score. The path performance score is positively correlated with the effective load capacity and negatively correlated with the transmission delay. The formula for calculating the path performance score can be expressed as: Where: symbol The path performance score represents the m-th transmission path, with the symbol... and It is a preset weighting coefficient and satisfies ,symbol Represents the effective payload capacity of the m-th transmission path, symbol Represents the maximum payload capacity across all evaluated transmission paths, with the symbol [symbol missing]. Represents the transmission delay of the m-th transmission path, symbol This represents the maximum transmission delay across all evaluated transmission paths.

[0032] In implementation, transmission paths are ranked according to path performance scores, with paths boasting higher performance scores appearing at the top of the priority list. The path selection strategy is adjusted in conjunction with a regional weight distribution map, which identifies the importance of each sub-region of the interactive interface. For interactive data streams originating from high-weight regions, the system prioritizes transmission paths with higher performance scores. Bandwidth resources are dynamically allocated, establishing a differential service mechanism based on quality requirements. This mechanism assigns different service levels to different data streams based on the scores of their associated regions in the interactive quality assessment matrix, ensuring high-priority data streams receive guaranteed bandwidth and low-latency forwarding. A priority-marked information transmission path configuration scheme is generated, clearly defining the set of transmission paths that data streams of different service levels should use, as well as the load ratio of each path.

[0033] In some embodiments, path performance measurement can employ a combination of active probing and passive monitoring. Active probing involves sending dedicated test data packets, while passive monitoring analyzes the transmission statistics of actual service data packets. In some embodiments, the differential service mechanism can be implemented as a multi-queue scheduling system, where data packets in high-priority queues are always sent before data packets in low-priority queues. Optionally, the transmission path performance evaluation model can consider path stability factors, incorporating historical fluctuations in path performance as a dimension of the evaluation. It is understood that transmission path performance evaluation needs to be continuous to adapt to dynamic changes in network conditions, and the evaluation frequency needs to strike a balance between the response speed to network changes and the measurement overhead. It is also understood that path priority configuration is a dynamic process; when network topology or load conditions change significantly, the path performance score needs to be recalculated and the configuration scheme updated.

[0034] Example 4: The resource control module monitors the current resource usage status of the system. Resource usage status data includes processor utilization, memory usage, and network bandwidth usage of all computing nodes in the system. Monitoring is achieved by periodically collecting performance counter data through an agent program deployed on each computing node. The module predicts the peak resource demand for each path based on the information transmission path configuration scheme, which clarifies the transmission paths to be used by data streams of different priorities. Peak resource demand prediction is based on historical data analysis and current data stream characteristics; for example, high-priority video stream data typically requires stable CPU computing resources for encoding and decoding and consumes significant network bandwidth. A resource allocation optimization model is established, mathematically describing how computing resources are allocated among multiple computing nodes. The objective function of the resource allocation optimization model is to pursue overall system load balancing, specifically by minimizing the dispersion of the real-time load coefficients of all computing nodes.

[0035] In practical implementation, real-time operational status data of each computing node in the system is collected. This data includes processor utilization percentage, memory usage in megabytes, and network bandwidth utilization percentage. A real-time load coefficient for each computing node is calculated. This coefficient is a weighted sum of processor utilization, memory usage, and network bandwidth utilization, with weights set based on the importance of different resource types in specific application scenarios. An objective function for overall system load balancing is defined, specifically as the variance of the real-time load coefficients of all computing nodes. A smaller variance indicates more similar load levels across nodes and a more balanced overall system load. Resource allocation constraints are set, including physical resource limits for each computing node. For example, the maximum processing power of a single computing node is the number of cores multiplied by the single-core clock speed; the maximum memory capacity is the installed physical memory size; and the bandwidth limit is the nominal speed of the network interface. A resource allocation optimization model is then implemented. The mathematical expression can be represented as: Where: symbol The variance represents the objective function value for overall system load balancing, with the sign... Represents the total number of computing nodes in the system, symbol Represents the real-time load factor of the i-th computing node, with the symbol... This represents the average real-time load factor of all computing nodes.

[0036] In practical implementation, referring to Table 1, a heuristic algorithm is used to solve for the optimal resource allocation scheme. Heuristic algorithms, such as genetic algorithms, search the solution space. The genetic algorithm encodes the resource allocation scheme as chromosomes and iteratively generates new candidate schemes through selection, crossover, and mutation operations. During each iteration, the genetic algorithm evaluates the objective function value corresponding to each candidate scheme in the current population. Candidate schemes with smaller objective function values ​​are considered to have higher fitness and are more likely to be selected for the next generation. The genetic algorithm continues to run until a termination condition is met, which is usually that the change in the objective function value is less than a set threshold or the maximum number of iterations is reached. The optimal resource allocation scheme obtained will reserve emergency processing margin for the system. This emergency processing margin is achieved by setting a safety retention threshold based on the upper limit of the computing node's resource capacity; for example, the actual allocated resources will not exceed 80% of the node's theoretical capacity. Dynamic resource scheduling is implemented through the scheduler component of the resource manager. The scheduler sends instructions to the computing nodes according to the optimal resource allocation scheme to adjust the resource quotas of each process or virtual machine, such as adjusting CPU time slice allocation weights, setting memory usage limits, or adjusting network bandwidth limits.

[0037] Table 1: Computing Node Resource Allocation Scheme In some embodiments, the resource allocation optimization model can be solved using a particle swarm optimization algorithm, which searches for the optimal solution in the solution space by simulating bird flock foraging behavior. In some embodiments, the weights of the real-time load coefficients of computing nodes can be dynamically adjusted according to the application type, assigning higher weights to processor utilization for computationally intensive tasks and higher weights to memory usage for data-intensive tasks. Optionally, peak resource demand prediction can incorporate machine learning methods, using time series forecasting models to predict future resource demands based on historical load patterns. Optionally, dynamic resource scheduling can leverage the capabilities of a container orchestration platform to achieve finer-grained resource allocation by adjusting the number of container replicas and their distribution among nodes. It is understood that the resource allocation optimization model needs to be periodically resolved to adapt to dynamic changes in system load, and the frequency of resolving needs to be balanced between optimization effectiveness and computational overhead. It is understood that reserving contingency capacity is crucial for maintaining system stability and coping with sudden load spikes, but excessive capacity can lead to reduced resource utilization.

[0038] Example 5: The actual operating parameters of the data acquisition device after the reallocation of computing resources are collected. These parameters include the real-time device operating frequency, interface response latency, and interactive touchpoint coordinates under the current resource allocation scheme. The data acquisition module continuously acquires these parameter values ​​at a fixed sampling period. The deviation between the actual operating parameters and the expected target is compared. The expected target is derived from the quality benchmark value or historical best performance index set in the interaction quality evaluation matrix. The deviation is quantified by calculating the absolute difference or relative error between the actual value and the target value. A control error coefficient is calculated. This coefficient is a comprehensive indicator reflecting the overall deviation between the current device operating state and the desired state. The control error coefficient comprehensively considers multiple dimensions, including device operating frequency deviation and interface response latency deviation.

[0039] In practical implementation, a fuzzy control rule base for equipment operating frequency is established. This base contains a series of fuzzy rules based on an "if-then" format. The antecedent (if part) of each fuzzy rule uses the control error coefficient and its rate of change as input variables, while the consequent (then part) outputs a suggested adjustment to the equipment operating frequency. A frequency adjustment step size strategy is formulated, specifying the exact magnitude of frequency adjustment based on the output of the fuzzy control rule base. This strategy defuzzifies the fuzzy output into precise frequency adjustment values, typically in megahertz. The response sensitivity of frequency control is dynamically adjusted based on real-time load trends. These trends are determined by analyzing the slope of recent system resource usage data. When a rapid increase in load is detected, the response sensitivity of frequency control is increased, allowing for larger frequency adjustment steps to respond quickly to changes. An adaptive equipment operating frequency control strategy table is generated. This table is a two-dimensional lookup table where row indices correspond to discrete control error coefficient levels, column indices correspond to discrete control error coefficient rate of change levels, and table elements store suggested equipment operating frequency adjustments.

[0040] The mathematical expression for the equipment operating frequency regulation strategy can be represented as: Where: symbol The adjustment amount representing the operating frequency of the equipment, symbol Represents the proportional control coefficient, symbol Represents the control error coefficient, symbol Represents the differential control coefficient, symbol This represents the rate of change of the control error coefficient over time.

[0041] The construction of a fuzzy control rule base relies on expert experience or learning from historical optimization data. A typical fuzzy rule might be expressed as "If the control error coefficient is negatively large and the rate of change of the control error coefficient is positively small, then the adjustment amount of the equipment operating frequency is positively large." In fuzzy logic processing, the input variables, the control error coefficient and the rate of change of the control error coefficient, need to be fuzzified first, converting their precise values ​​into membership degrees of linguistic variables such as "negatively large," "negatively small," "zero," "positively small," and "positively large." The fuzzy inference engine performs a synthesis operation based on all activated fuzzy rules, resulting in a fuzzy output quantity, i.e., a fuzzy set of equipment operating frequency adjustment amounts. The defuzzification process converts the fuzzy output set into a precise numerical value of the equipment operating frequency adjustment amount. Defuzzification often uses the centroid method to calculate the centroid of the fuzzy set as the final output.

[0042] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0043] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An interactive information system based on digital media technology, characterized in that, Includes the following modules: The data acquisition module is used to collect raw operating data generated by user interaction devices and extract device operating frequency, interface response delay and interactive touch point coordinate sequence; The data preprocessing module is used to normalize the raw operating data, eliminate the influence of dimensions caused by equipment differences, and generate a standardized set of equipment operating parameters. The behavior analysis module is used to construct an interactive behavior data stream based on the standardized equipment operating parameter set through spatiotemporal correlation analysis, mark abnormal interaction intervals, and calculate the continuity index of the behavior trajectory. The region segmentation module is used to dynamically segment interactive sensitive regions and generate a region weight distribution map based on the interactive behavior data stream and a preset environmental noise threshold. The quality assessment module is used to spatially reconstruct the coordinate sequence of interactive touch points using the regional weight distribution map, calculate the coupling relationship between touch point distribution density and interface response delay, and generate an interactive quality assessment matrix. The resource regulation module is used to adjust the priority configuration of information transmission paths, reallocate computing resources, and update the regulation strategy of device operating frequency based on the interaction quality assessment matrix.

2. The interactive information system based on digital media technology according to claim 1, characterized in that, The process of generating the standardized equipment operating parameter set includes: Establish a reference coordinate system for equipment operating parameters and resample equipment data at different sampling frequencies to a unified time axis; Calculate the parameter offset for each sampling point, and use the sliding window method to smooth the offset; The filtered parameter sequence is standardized to conform to a normal distribution. Establish a rule base for association between parameters and remove abnormal data points that do not conform to the association rules; Feature extraction is performed on the cleaned parameter sequence to generate a standardized set of equipment operating parameters containing time-domain and frequency-domain features.

3. The interactive information system based on digital media technology according to claim 2, characterized in that, The method for constructing the interactive behavior data stream includes: Based on the standardized equipment operating parameter set, the spatiotemporal feature vector of the user's operation trajectory is extracted; Calculate the similarity matrix between feature vectors and identify typical interaction patterns through cluster analysis; Establish an interactive behavior state transition model and label the key nodes of state transition; The abnormal interaction intervals are segmented into multiple scales, and the rate of change of curvature of the behavioral trajectory within each interval is calculated. By combining historical interaction data to train a behavior prediction model, an interaction behavior data stream with time-series labels is generated.

4. The interactive information system based on digital media technology according to claim 3, characterized in that, The steps for generating the regional weight distribution map include: Based on the abnormal interaction intervals in the interactive behavior data stream, the sensitive frequency bands of environmental noise are determined; Construct a noise impact assessment function based on a sound field propagation model to calculate the degree of noise interference in each area; The interactive interface is divided into several sub-regions using an adaptive grid partitioning method; Calculate the regional importance weighting coefficient based on the interaction frequency and noise interference level of each sub-region; A regional weight distribution map with spatial resolution is generated by combining weighting coefficients.

5. The interactive information system based on digital media technology according to claim 4, characterized in that, The calculation method for the interaction quality assessment matrix includes: Based on the regional weight distribution map, extract the contact point distribution feature values ​​of each sub-region; Calculate the correlation coefficient matrix between contact point distribution density and interface response delay; Establish a multi-dimensional evaluation index system for interaction quality, including response timeliness, trajectory smoothness, and operational accuracy; Principal component analysis was used for dimensionality reduction to extract key quality feature vectors. The quality feature vectors are weighted and fused with regional weights to generate an interactive quality assessment matrix.

6. The interactive information system based on digital media technology according to claim 5, characterized in that, The priority configuration method for the information transmission path includes: Analyze the quality feature vectors in the interaction quality assessment matrix to identify key transmission bottleneck areas; Establish a transmission path performance evaluation model to calculate the load capacity and transmission delay of each path; Adjust the path selection strategy according to the regional weight distribution map, and prioritize data transmission in high-weight regions; Dynamically allocate bandwidth resources and establish a differential service mechanism based on quality requirements; Generate a configuration scheme for information transmission paths with priority tags.

7. The interactive information system based on digital media technology according to claim 6, characterized in that, The process of reallocating computing resources includes: Monitor the current resource usage status of the system, including processor load, memory usage, and network bandwidth usage; Based on the information transmission path configuration scheme, predict the peak resource demand for each path; Establish a resource allocation optimization model with overall system load balancing as the objective function; A heuristic algorithm is used to find the optimal resource allocation scheme, with a margin reserved for emergency handling. Implement dynamic resource scheduling and adjust the resource quotas of each process in real time.

8. The interactive information system based on digital media technology according to claim 7, characterized in that, The method for updating the device operating frequency control strategy includes: The actual operating parameters of the data acquisition device after the reallocation of computing resources; Compare the deviation between the actual operating parameters and the expected target, and calculate the control error coefficient; Establish a fuzzy control rule base for equipment operating frequency and formulate a frequency adjustment step size strategy; The response sensitivity of frequency control is dynamically adjusted based on real-time load change trends. Generate a table of device operating frequency control strategies with adaptive characteristics.

9. The interactive information system based on digital media technology as described in claim 6, characterized in that, The process of establishing a transmission path performance evaluation model and calculating the load capacity and transmission delay of each path includes: Periodically measure the data packet transmission rate, packet loss rate, and round-trip time for each transmission path; The effective load capacity of each path is calculated based on the packet transmission rate and packet loss rate. The effective load capacity is the product of the maximum theoretical transmission rate and the available bandwidth coefficient, which is adjusted based on the packet loss rate. The transmission delay for each path is calculated based on the round-trip time and jitter value. The transmission delay is the average round-trip time plus the jitter compensation value. A transmission path performance evaluation model is established. The model inputs are effective load capacity and transmission delay, and the output is the path performance score. The path performance score is positively correlated with effective load capacity and negatively correlated with transmission delay. Transmission paths are sorted based on path performance scores for priority configuration.

10. The interactive information system based on digital media technology as described in claim 7, characterized in that, The process of establishing a resource allocation optimization model with overall system load balancing as the objective function includes: Real-time acquisition of operational status data for each computing node in the system, including processor utilization, memory usage, and network bandwidth utilization; Calculate the real-time load factor for each computing node. The load factor is a weighted sum of processor utilization, memory usage, and network bandwidth utilization. Define the objective function for overall system load balancing. The objective function is the variance of the load coefficients of all computing nodes, and the goal is to minimize this variance. Set resource allocation constraints, including the maximum processing power, memory capacity, and bandwidth limit for each computing node; An iterative optimization method is used to adjust the resource allocation scheme. The objective function value is recalculated in each iteration until the change in the objective function value is less than a set threshold. The iteration stops and the optimal resource allocation scheme is output.

Images