Intelligent contract-driven consultation data processing method

Abstract

The invention relates to the technical field of consultation data processing, and discloses a consultation data processing method driven by an intelligent contract, which is applied to a consultation data processing system comprising a plurality of consultation data nodes deployed on a block chain network and an intelligent contract engine. According to the method, consultation data nodes collect data to generate a consultation data set, an intelligent contract engine receives real-time data to construct a matrix, consultation demand distribution characteristics are determined, and consultation resources are dynamically scheduled according to the consultation demand distribution characteristics. And when the distribution characteristics are determined, processing the data matrix to extract indexes, carrying out data quality modeling and demand distribution prediction, constructing an intelligent contract model, outputting space-time distribution characteristics, and updating a data set. During dynamic scheduling, partition identifiers and distribution characteristics are mapped, resources are controlled to execute actions, and the resources are dynamically allocated. According to the method, efficient processing of consultation data and dynamic scheduling of resources are realized, and the data processing efficiency and the resource utilization rate are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of consulting data processing, and in particular to a consulting data processing method driven by a smart contract. Background Art

[0002] From the perspective of data collection and processing, traditional systems mostly adopt a centralized data storage and processing model. This model often concentrates data on a small number of servers, which can easily lead to single points of failure. Once a server fails, data processing for the entire system is paralyzed. Furthermore, the scope and efficiency of data collection are severely limited, making it difficult to obtain comprehensive, real-time data on user inquiry requests and responses distributed across different regions. Furthermore, due to inconsistent data formats and complex data sources, data integrity and consistency are difficult to ensure. Data loss and errors often occur during data processing, severely impacting subsequent analysis and application.

[0003] When it comes to demand analysis and resource scheduling, traditional methods lack a precise understanding of the spatiotemporal distribution characteristics of consulting demand. They typically rely on simple statistical analysis based on historical data, failing to dynamically adjust to current data availability. This results in inaccurate forecasts of consulting demand and an inability to promptly identify hotspots and changing trends. Resource scheduling often employs fixed resource allocation strategies, failing to dynamically optimize based on demand distribution, which can easily lead to wasted or insufficient resources. For example, during peak demand periods, resources in certain areas may be severely insufficient, resulting in increased response delays and a reduced user experience. Conversely, during low demand periods, significant amounts of resources may remain idle.

[0004] Furthermore, traditional consulting data processing methods pose significant risks in terms of data security and reliability. Centralized storage makes data vulnerable to hacker attacks and tampering, making it difficult to guarantee data security. Furthermore, due to the lack of transparency in data processing, users are unable to understand the specific data processing procedures and the reliability of the results, which to some extent undermines their trust in consulting services.

[0005] The development of blockchain technology, characterized by decentralization, immutability, and traceability, offers new solutions to these challenges. However, integrating blockchain technology with consulting data processing to fully leverage its advantages and achieve efficient processing of consulting data and intelligent scheduling of consulting resources remains a pressing technical challenge. Currently, most existing blockchain-based consulting data processing solutions are still in the exploratory stage, lacking sufficient data processing efficiency, accurate demand analysis, and flexible resource scheduling to meet the demands of practical applications. Therefore, a new consulting data processing approach is urgently needed to address the challenges inherent in traditional technologies, improve the efficiency and quality of consulting data processing, and achieve dynamic and optimized scheduling of consulting resources. Summary of the Invention

[0006] The purpose of the present invention is to provide a smart contract driven consulting data processing method to solve the problems raised in the above background technology.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a smart contract-driven consulting data processing method, the method comprising:

[0008] Collect user consultation request data and response data through the consultation data node to generate a consultation data set;

[0009] Receive real-time consulting data from multiple consulting data nodes through the smart contract engine and build a consulting data matrix;

[0010] Determining the distribution characteristics of consulting demands based on the consulting data set and the real-time data of each consulting data node, wherein determining the distribution characteristics of consulting demands includes: processing the consulting data matrix, extracting demand characteristics in combination with the consulting data set, and predicting demand distribution based on smart contract execution information and data path information, outputting spatiotemporal distribution characteristics of consulting demands through a smart contract model, and updating the consulting data set based on the spatiotemporal distribution characteristics;

[0011] According to the distribution characteristics, consulting resources are dynamically scheduled.

[0012] Preferably, determining the distribution characteristics of consulting needs includes:

[0013] Processing the consulting data matrix to extract data integrity indicators, response delay characteristics, and demand trend information;

[0014] Performing data quality modeling on the consultation data matrix based on the data integrity indicators and response delay characteristics, dividing the consultation data into multiple data partitions and marking the partition identifiers, performing correlation matching with the consultation data set based on the integrity indicators of the data partitions, and marking the partition identifiers in the consultation data set;

[0015] Calculating a data path gradient according to the location of the consultation data node, predicting the distribution of consultation demand according to the data path gradient and the demand trend information, and calculating demand forecast information for each data partition;

[0016] Constructing a smart contract model, using the demand forecast information as an input parameter of the smart contract model, performing logical association modeling on the demand forecast information through the smart contract model, and outputting the spatiotemporal distribution characteristics of consulting demand;

[0017] The consultation data set is updated according to the spatiotemporal distribution characteristics to obtain the distribution characteristics of the consultation needs.

[0018] Preferably, the processing of the consultation data matrix includes:

[0019] Standardize the consulting data matrix, intercept the demand hotspot areas in the matrix through the time window, perform redundancy filtering on the hotspot areas, and calculate the demand trend information through the eigendecomposition algorithm;

[0020] Calculating logical correlation features of the consultation data matrix, calculating the dependency strength, data stability coefficient, and missing data index between partitions based on the logical correlation features, constructing a feature fusion network, and calculating response delay features through the feature fusion network;

[0021] The time domain features and semantic features collected by each consulting data node are extracted, and the demand feature vector of the node is calculated based on the deviation value of the time domain features and the semantic features. The consulting data nodes at different locations are feature matched according to the demand feature vector to calculate the demand trend information.

[0022] Preferably, the performing data quality modeling on the consultation data matrix includes:

[0023] Extract integrity sampling points from each frame of data based on the data integrity indicator, perform correlation mapping between the sampling points and response delay characteristics, generate a data quality map, logically align the data quality maps collected by multiple nodes, and calculate the data quality distribution of the consulting data;

[0024] Set the integrity threshold value, locate data missing according to the integrity value of the multi-frame consultation data matrix, calculate the integrity difference, if the integrity difference is greater than or equal to the integrity threshold value, it means that there is data missing in the partition, perform data model constraint compensation on the current partition, iteratively correct the data quality distribution of the current partition according to the data verification model corresponding to the current partition, and calculate the quality compensation value of the missing data based on the correction result;

[0025] Data quality modeling is performed on the consulting data matrix according to the data quality distribution, and delay annotation is performed on the partitioned data model according to the response delay characteristics.

[0026] Preferably, the step of calculating the data path gradient according to the location of the consulting data node comprises:

[0027] Extract data change points based on multiple sets of consulting data transmission data, and map the change points to a unified data coordinate system based on the node deployment location. Fit the change points using a logical interpolation algorithm to generate a data path model for the consulting data.

[0028] Performing sampling at equal intervals along the transmission path of the data path model, calculating a data attenuation rate, a dependency fluctuation index, and a demand change slope of the path based on the sampling results, and calculating a path change parameter based on the data attenuation rate, the dependency fluctuation index, and the demand change slope;

[0029] Based on the deployment parameters and collection accuracy of the consultation data nodes, the distribution characteristics of consultation demand in each frame of data are projected onto the data path model. The data path model is partitioned according to the number of nodes along the transmission direction. The changing pattern of consultation demand within the partition is analyzed, and the demand distribution characteristics are calculated based on the changing pattern.

[0030] The data path gradient is calculated based on the path change parameters and the demand distribution characteristics. The calculation process of the data path gradient includes: based on the logical position range from the first consultation data node to the last consultation data node, selecting data coordinate points in the node deployment direction, accumulating the product of the path characteristic weight value and the demand distribution characteristic weight value within the logical resolution range, and superimposing the impact value of the node acquisition frequency on the data path change rate.

[0031] Preferably, the calculating of demand forecast information for each data partition includes:

[0032] Taking the main transmission path of the data path model as the baseline and the peak position of the consultation demand in each frame of data as the reference point, the demand offset is calculated and the demand distribution curve is drawn according to the data coordinates;

[0033] Correcting the growth rate and direction in the demand trend information based on the data path gradient;

[0034] Starting from the most recent demand distribution point, the distribution curve is continuously drawn based on the correction results of growth rate and direction to generate the demand distribution points for the next period until the distribution points cover the entire target partition and generate demand forecast information.

[0035] Preferably, the construction of the smart contract model includes:

[0036] The input layer is used to organize demand forecast information into logically distributed data and perform standardization processing;

[0037] The feature fusion layer is used to extract the logical association features of requirements by processing logically distributed data and build dependency relationships between data units;

[0038] The resource allocation layer is used to integrate the association between consulting needs in logical units and generate consulting resource scheduling strategies.

[0039] Preferably, the distribution characteristics of the obtained consulting needs include:

[0040] According to the spatiotemporal distribution characteristics of the consulting needs output by the smart contract model, the identifiers of the data partitions are matched with the spatiotemporal distribution characteristics;

[0041] The partition data in the consultation dataset are reorganized according to the spatiotemporal characteristics to generate a partition distribution map sorted by the consultation demand intensity;

[0042] According to the reorganized partition distribution map, the optimized consultation demand distribution characteristics are output.

[0043] Preferably, the dynamic scheduling of consulting resources includes:

[0044] Mapping partition identifiers to partitions with distribution characteristics of consulting needs one by one;

[0045] Control the execution of consulting resources based on the spatiotemporal distribution characteristics of consulting needs, including computing resource adjustment, data aggregation, and response allocation operations;

[0046] Based on the spatial distribution of consulting needs and the preset resource scheduling strategy, consulting resources are dynamically allocated to the corresponding data partitions.

[0047] Preferably, the execution action of the control consulting resource includes:

[0048] When the consultation demand reaches a preset intensity threshold in the target partition, a resource promotion instruction for the adjacent computing resources is triggered;

[0049] Dynamically combine available data resources according to data aggregation strategies to generate response allocation vectors;

[0050] A logic parameter of the target partition processing node is adjusted based on the response allocation vector.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] In terms of data collection and processing, the system collects user consultation request data and response data through multiple consultation data nodes deployed on the blockchain network to generate consultation data sets. This decentralized collection method expands the scope of data collection, improves the real-time and comprehensiveness of data collection, and effectively avoids the single point of failure problem in the traditional centralized model. At the same time, the smart contract engine receives real-time consultation data from multiple consultation data nodes, constructs a consultation data matrix, and performs a series of processing on it, including standardization, hot spot area interception, redundant filtering, feature decomposition, etc., which can extract data integrity indicators, response delay characteristics, and demand trend information. It then conducts data quality modeling on the consultation data matrix and generates a data quality map, achieving accurate assessment and management of data quality, improving data integrity and consistency, and providing a reliable data foundation for subsequent demand analysis and resource scheduling.

[0053] In terms of demand analysis, this method can accurately determine the distribution characteristics of consulting demand. By processing the consulting data matrix and modeling data quality, combined with smart contract execution information and data path information, it predicts the distribution of consulting demand. By building a smart contract model and outputting the spatiotemporal distribution characteristics of consulting demand, it can accurately and in real time grasp the spatiotemporal variation patterns of consulting demand, including demand hotspots and changing trends. This precise demand analysis can help consulting firms better understand user needs and provide strong support for decision-making.

[0054] In terms of resource scheduling, consulting resources can be dynamically scheduled based on the distribution characteristics of consulting needs. Partition identifiers are mapped one-to-one with the partitions of the distribution characteristics of consulting needs. The execution actions of consulting resources are controlled according to the spatiotemporal distribution characteristics, including computing resource adjustment, data aggregation, and response allocation operations. Resources are dynamically allocated to corresponding data partitions based on spatial distribution and preset strategies. This dynamic scheduling method can achieve optimal resource allocation, timely enhance adjacent computing resources during peak demand periods, and avoid response delays caused by insufficient resources; reasonably allocate resources during low demand periods, reduce idle resources, improve resource utilization efficiency, and enhance the user's consulting experience.

[0055] In terms of data security and credibility, the decentralized and tamper-proof nature of blockchain technology ensures the security and credibility of consulting data during collection, transmission, and storage. Data processing is automatically executed through smart contracts, providing transparency and traceability. Users can clearly understand the data processing process, enhancing their trust in consulting services. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a working principle diagram of the smart contract-driven consulting data processing method of the present invention;

[0057] Figure 2 Design diagram determined by the distribution characteristics of consulting needs;

[0058] Figure 3 Design drawings for consulting data matrix processing;

[0059] Figure 4 A blueprint for the smart contract model. DETAILED DESCRIPTION

[0060] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0061] See also Figure 1-Figure 4 The present invention relates to a smart contract-driven consulting data processing method, which is applied to a consulting data processing system. The system includes multiple consulting data nodes deployed on a blockchain network and a smart contract engine connecting the consulting data nodes. Adjacent consulting data nodes are separated by a set distance. The specific implementation process of this method is described in detail below.

[0062] The consultation data nodes collect user consultation request and response data to generate a consultation data set. These consultation data nodes are distributed across the blockchain network and deployed at set distances. They can obtain user consultation-related data in real time and perform preliminary collection and organization of this data to form an initial consultation data set.

[0063] The smart contract engine receives real-time consulting data from multiple consulting data nodes and constructs a consulting data matrix. As the core component connecting various consulting data nodes, the smart contract engine can receive consulting data transmitted by each node in real time and organize this data according to certain rules and structures to form a consulting data matrix for subsequent data processing and analysis.

[0064] Based on the consulting dataset and real-time data from each consulting data node, the distribution characteristics of consulting demand are determined. This process includes processing the consulting data matrix, extracting demand characteristics based on the consulting dataset, predicting demand distribution based on smart contract execution information and data path information, outputting the spatiotemporal distribution characteristics of consulting demand through the smart contract model, and finally updating the consulting dataset based on these spatiotemporal distribution characteristics. This process requires comprehensive consideration of multiple factors and in-depth data analysis and processing to accurately determine the distribution characteristics of consulting demand.

[0065] Dynamically dispatch consulting resources based on distribution characteristics. By understanding the distribution characteristics of consulting needs, consulting resources can be more reasonably allocated and dispatched to meet users' consulting needs and improve the efficiency and quality of consulting services.

[0066] Example 1:

[0067] To determine the distribution characteristics of consulting demand, the consulting data matrix must be processed to extract data integrity indicators, response latency characteristics, and demand trend information. The consulting data matrix is ​​standardized, and demand hotspots within the matrix are captured by setting time windows, such as dividing the time windows into minutes. Data within these hotspots is then filtered for redundancy, removing duplicate or invalid data. Feature decomposition algorithms, such as singular value decomposition (SVD), are then applied to the filtered data to calculate demand trend information, reflecting the changing trends of consulting demand over time.

[0068] The logical correlation characteristics of the consulting data matrix are calculated. By analyzing the logical relationships between different data points in the matrix, such as the order of data and causal relationships, the inter-partition dependency strength, data stability coefficient, and missing data index are calculated. Dependence strength measures the degree of association between different data intervals, the data stability coefficient reflects the fluctuation of data over a certain period of time, and the missing data index indicates the degree of data missingness. Based on these characteristics, a feature fusion network is then constructed. This network can be a neural network composed of multiple neuron layers. By performing weighted fusion and nonlinear transformation on the input features, the response delay characteristic is calculated, that is, the time delay characteristics experienced by the data from acquisition to response.

[0069] Extract the time domain features and semantic features collected by each consulting data node. Time domain features include time-related features such as the timestamp and time interval of data collection, while semantic features are the semantic understanding of the data content, such as the topic and keywords of the consulting question. Based on the deviation value of the time domain features and semantic features, that is, the degree of difference between the two, calculate the node's demand feature vector, which comprehensively reflects the node's demand characteristics. Then, based on the demand feature vector, feature matching is performed on consulting data nodes in different locations. By calculating the similarity or distance between the vectors, the degree of association between the nodes is determined, and then the demand trend information is calculated to further verify and supplement the overall demand trend from the node level.

[0070] After completing the above processing, data quality modeling is performed on the consulting data matrix based on data integrity indicators and response delay characteristics. First, based on the data integrity indicators, integrity sampling points are extracted from each frame of data. These sampling points are key locations with high data integrity. Then, the sampling points are correlated and mapped with the response delay characteristics, establishing a corresponding relationship between the two, and generating a data quality map. This map visually displays the data quality distribution. The data quality maps collected by multiple nodes are logically aligned and arranged in chronological order or spatial location. The data quality distribution of the consulting data is calculated, thus comprehensively understanding the quality status of the entire consulting data.

[0071] Set an integrity threshold value, which can be set according to actual needs and data characteristics, for example, set to 80%. Locate data missing data based on the integrity value of the multi-frame consultation data matrix, and determine the location and range of data missing by comparing the integrity value of each frame of data with the threshold value. Calculate the integrity difference, that is, the difference between the actual integrity value and the threshold value. If the integrity difference is greater than or equal to the integrity threshold value, it indicates that there is data missing in the partition. At this time, the data model constraint compensation is performed on the current partition. According to the data verification model corresponding to the current partition, such as a rule-based verification model or a statistical model, the data quality distribution of the current partition is iteratively corrected. Through multiple adjustments and optimizations, the data quality gradually approaches the ideal state. Calculate the quality compensation value of the missing data based on the correction results, which is used for subsequent supplementation and repair of the missing data.

[0072] Data quality modeling is performed on the consulting data matrix based on the data quality distribution. A mathematical model that can accurately describe the data quality is established. The partitioned data model is delayed labeled using the response delay characteristics, and each data partition is marked with its corresponding response delay characteristics to facilitate subsequent optimization of the data processing and response process.

[0073] The data path gradient is calculated based on the location of the consulting data nodes. First, based on multiple sets of consulting data transmission data, data change points are extracted. These change points are locations where significant changes occur during the data transmission process. Based on the node deployment locations, these change points are mapped to a unified data coordinate system to establish a unified spatial reference frame. Logical interpolation algorithms, such as cubic spline interpolation, are used to fit the change points and generate a data path model for the consulting data. This model describes the data transmission paths and patterns between nodes.

[0074] Sampling is performed at regular intervals along the transmission path of the data path model, for example, sampling points are selected at regular intervals. Based on the sampling results, the path's data attenuation rate, dependency fluctuation index, and demand change slope are calculated. The data attenuation rate indicates the degree of data loss during transmission, the dependency fluctuation index reflects the fluctuations in data transmission dependencies, and the demand change slope reflects the rate of demand change along the transmission path. Based on these three parameters, the path change parameter is calculated to comprehensively reflect the changing characteristics of the data path.

[0075] Based on the deployment parameters and collection accuracy of the consulting data nodes, the distribution characteristics of consulting demand in each frame of data are projected onto the data path model, and the demand distribution is correlated with the data transmission path. The data path model is partitioned along the transmission direction based on the number of nodes, dividing the transmission path into multiple zones. The changing patterns of consulting demand within these zones are analyzed, such as demand increase and decrease trends and distribution density. Based on these changing patterns, the demand distribution characteristics are calculated to clarify the demand characteristics of each zone.

[0076] Finally, the data path gradient is calculated based on the path change parameters and demand distribution characteristics. The specific process involves selecting data coordinate points in the node deployment direction based on the logical location range from the first to the last consultation data node. The product of the path characteristic weight value and the demand distribution characteristic weight value within the logical resolution range is cumulatively calculated. The weight value can be set according to actual conditions to reflect the importance of different characteristics. The impact of the node acquisition frequency on the data path change rate is also added. The higher the node acquisition frequency, the greater the impact on the data path change rate. The data path gradient is ultimately obtained, which is used to describe the gradient of demand distribution along the data path.

[0077] After calculating the data path gradient, we calculate demand forecasts for each data partition. Using the data path model's main transmission path as the baseline, we determine the primary direction and path of data transmission. Using the peak position of the consultation demand in each frame of data as the reference point, we calculate the demand offset—the degree to which the peak position deviates from the baseline. We then plot a demand distribution curve based on the data coordinates to visually demonstrate the distribution of demand across the data coordinates.

[0078] Based on the data path gradient, the growth rate and direction of the demand trend information are corrected. The data path gradient reflects the changing trend of the demand distribution. This gradient is used to adjust the growth rate and direction of the original demand trend information to better reflect actual demand changes. Starting from the most recent demand distribution point, the distribution curve is drawn based on the revised growth rate and direction, gradually generating the demand distribution points for the next time period. The demand distribution points for each time period are generated in chronological order until the distribution points cover the entire target zone. Finally, demand forecast information is generated, providing an accurate demand forecast basis for subsequent consulting resource scheduling.

[0079] Example 2:

[0080] When building a smart contract model, the model's input layer is used to organize demand forecast information into logically distributed data and perform standardization. Demand forecast information includes multidimensional data such as demand distribution points, growth rates, and offsets for each data partition. The input layer must first structure this information, for example, by constructing a multidimensional data structure based on dimensions such as spatiotemporal coordinates, demand intensity, and time series to ensure that the data conforms to the model's input format requirements. During standardization, a normalization or standardization algorithm, such as minimum-maximum normalization, is used to map demand forecast information of different dimensions to a unified numerical range, such as [0, 1]. This eliminates the impact of data dimension differences on model calculations and ensures the consistency and comparability of input data.

[0081] The feature fusion layer processes logically distributed data to extract logically correlated features of demand and construct dependency relationships between data units. This layer can employ algorithms such as graph neural networks (GNNs) or attention mechanisms to extract features from the input logically distributed data. Taking GNNs as an example, each data unit is considered a node in a graph. Node attributes include information such as the unit's demand intensity and spatiotemporal coordinates. Edges between nodes represent temporal, spatial, or logical relationships between data units, such as demand transmission between adjacent partitions or demand continuity between different time periods within the same partition. Through the message-passing mechanism of GNNs, nodes exchange information, capturing logical correlation features such as dependency strength and transmission paths between data units. For example, for two spatially and temporally adjacent data partitions, the feature fusion layer analyzes the impact of the demand intensity of partition A in the previous time period on the demand intensity of partition B in the current time period and constructs a dependency weight between the two. For data from different time points within the same partition, it extracts continuity characteristics of demand over time, such as the inertial trend of demand growth.

[0082] When processing logically distributed data, the feature fusion layer also performs multi-dimensional feature transformation and fusion on the data. For example, it cross-combines temporal features (such as time period numbers and weekday / weekend identifiers) with spatial features (such as partition coordinates and node deployment locations) to generate joint spatiotemporal features. It also fuses historical demand distribution features with current forecast features to capture the dynamic patterns of demand changes. Through nonlinear transformations in multi-layer neural networks, low-dimensional raw features are mapped to a high-dimensional feature space, enabling the model to learn more abstract and representative logical association features.

[0083] The resource allocation layer integrates the relationships between consulting demands across logical units and generates consulting resource scheduling strategies. This layer builds a resource allocation optimization model based on the previously extracted logical association features and combines them with pre-set resource constraints (such as total computing resources, data storage capacity, and response time requirements). Logical units can be data partitions, consulting data nodes, or processing modules within smart contract engines. The resource allocation layer analyzes the demand intensity, priority, and relationships of each logical unit with other units to determine the priority and quantity of resource allocation.

[0084] Specifically, the resource allocation layer first sorts the logical units by demand intensity based on the temporal and spatial distribution characteristics of the demand, and prioritizes resource allocation for logical units with high demand intensity. For example, for data partitions where demand peaks, the resource allocation layer will increase the proportion of computing resources corresponding to the partition to ensure that it can handle consultation requests in a timely manner. Secondly, considering the dependencies between logical units, for logical units with strong dependencies, such as upstream data partitions and downstream processing nodes, the resource allocation layer will coordinate the resource allocation of the two to avoid a decrease in overall processing efficiency due to insufficient resources in a certain link. For example, if the demand changes in partition A will significantly affect the processing load of partition B, the resource allocation layer will consider the resource requirements of both partitions when allocating resources to ensure balanced resource allocation.

[0085] The resource allocation layer also generates specific resource scheduling strategies, including computing resource adjustment strategies, data aggregation strategies, and response allocation strategies. Computing resource adjustment strategies involve the dynamic allocation of resources such as CPU and memory between different data partitions or nodes. For example, when the demand for a partition suddenly increases, the resource allocation layer will trigger the resource migration mechanism and temporarily allocate the idle computing resources of adjacent nodes to the partition. The data aggregation strategy is used to determine how to integrate and process consulting data from different sources and types to improve data processing efficiency. For example, based on the spatiotemporal correlation of the data, data from the same region and the same time period can be aggregated to the same processing node. The response allocation strategy specifies how consulting responses are generated and distributed. For example, based on the user's location and the content of the consultation, the response is allocated to the most appropriate consulting data node for processing to shorten the response time.

[0086] When generating resource scheduling strategies, the resource allocation layer employs optimization algorithms, such as linear programming, greedy algorithms, or reinforcement learning algorithms, to maximize resource utilization efficiency and consulting service quality while satisfying resource constraints. For example, when using reinforcement learning algorithms, the resource allocation process is treated as a decision-making process. The agent continuously learns the optimal resource allocation strategy through interaction with the environment (i.e., the consulting data processing system) to minimize overall response latency or maximize resource utilization.

[0087] Furthermore, the resource allocation layer considers the execution rules and constraints of smart contracts. Smart contracts may predefine resource allocation priorities, thresholds, and other rules. For example, when the urgency of a particular type of consultation request reaches a certain level, a specific amount of resources must be allocated to handle it. The resource allocation layer embeds these rules into the generation of resource scheduling policies to ensure that policy execution complies with the requirements of the smart contract.

[0088] Ultimately, the resource allocation layer will output the integrated logical association features and the generated resource scheduling strategy to provide specific guidance for the dynamic scheduling of consulting resources, enabling the system to automatically adjust resource allocation according to real-time changes in consulting needs, thereby improving the system's responsiveness and resource utilization efficiency.

[0089] Example 3:

[0090] When obtaining the distribution characteristics of consulting needs, it is necessary to match the identifiers of data partitions with the spatiotemporal distribution characteristics based on the spatiotemporal distribution characteristics of consulting needs output by the smart contract model. The spatiotemporal distribution characteristics output by the smart contract model contain information such as the demand intensity and distribution range of each data partition in the time and space dimensions. Each data partition is assigned a unique partition identifier, such as a partition ID, during construction. At this point, it is necessary to establish a mapping relationship between the partition identifier and the spatiotemporal distribution characteristics. For example, a two-dimensional table is created with the partition identifier in the first column and the corresponding spatiotemporal distribution feature vector in the second column. This vector contains information such as the demand intensity and spatial coordinate range of the partition at different time points, thereby achieving a precise correspondence between the two and providing a clear index for subsequent data processing and analysis.

[0091] The partitioned data in the consultation dataset is then reorganized according to spatiotemporal characteristics to generate a partitioned distribution map sorted by consultation demand intensity. The consultation dataset consists of raw data from multiple data partitions, each containing user consultation request data and response data. During the reorganization process, all partitioned data within the same time period is extracted based on the temporal dimension of the spatiotemporal distribution characteristics. Next, adjacent partitions or those with similar geographical characteristics are clustered according to the spatial dimension. For example, partitioned data from different regions of the same city can be grouped together to facilitate analysis of demand distribution within the region.

[0092] Based on clustering, the demand intensity in each partition data is calculated. The demand intensity can be comprehensively evaluated by multiple indicators such as the number of consultation requests in the partition, the complexity of the request, and the response time. Assume that the calculation formula for demand intensity is:

[0093] I=α·N+β·C+γ·T

[0094] Where I represents demand intensity, N is the number of consultation requests within the partition, C is the request complexity index (ranging from 1 to 5, with larger values ​​indicating greater complexity), T is the response time (in seconds), and α, β, and γ are weight coefficients, satisfying α + β + γ = 1. These weight coefficients can be set based on actual business needs and historical data experience. For example, if the number of consultation requests has the greatest impact on demand intensity, then the value of α should be relatively large.

[0095] After calculating the demand intensity for each zone, the zones are sorted from highest to lowest by demand intensity to generate a sorted zone list. A two-dimensional coordinate system is then constructed, with spatial location as the horizontal axis and time as the vertical axis. Each zone is annotated within the coordinate system according to its spatial coordinates and the demand intensity at the corresponding time point, forming a zone distribution map. Different colors or graphic symbols can be used to indicate demand intensity, for example, darker colors indicate higher demand intensity, making the distribution map more intuitive and clear.

[0096] Finally, based on the reorganized zoning distribution map, we output optimized consulting demand distribution characteristics. The zoning distribution map intuitively displays the distribution of consulting demand across time and space. By analyzing this map, we can extract more representative demand distribution characteristics. For example, we can identify the time periods and regions where demand peaks, analyze the spatial diffusion of demand, such as the trend from the city center to the suburbs, and analyze demand variations across different time periods, such as the difference in demand between weekdays and weekends.

[0097] During the analysis process, data mining algorithms, such as clustering or association rule algorithms, can be used to further uncover hidden demand distribution characteristics. For example, the K-means clustering algorithm can be used to cluster demand points in the zoning distribution map, grouping demand points with similar spatiotemporal distribution characteristics into categories. Each category represents a specific demand distribution pattern, such as the midday peak demand pattern in commercial areas and the evening demand pattern in residential areas.

[0098] Furthermore, the impact of smart contract execution information and data path information on demand distribution characteristics must be considered. Smart contracts may contain rules and constraints regarding consulting services, such as prioritizing consulting requests from certain regions. These rules can influence the actual distribution and processing of demand. Data path information reflects the transmission path and processing flow of consulting data between various nodes. The efficiency and stability of this path can also affect demand distribution characteristics. For example, partitions with long data paths may experience response delays, indirectly affecting the assessment of demand intensity.

[0099] The optimized consulting demand distribution characteristics must be output in a structured manner, such as a report that includes temporal and spatial distribution patterns, demand peak characteristics, and regional demand differences. This report should detail the specific manifestations and causes of each characteristic, providing an accurate basis for the subsequent dynamic scheduling of consulting resources. For example, if the report indicates that a certain area has a significant demand peak between 9:00 AM and 11:00 AM on weekdays, the system can pre-allocate more computing resources and consulting personnel to that area during this time to ensure service quality.

[0100] Throughout the process of capturing consulting demand distribution characteristics, data accuracy and completeness must be ensured. Data cleaning and repair techniques are employed to address potential data gaps or anomalies. For example, interpolation methods are used to fill in missing demand intensity data, and outlier detection algorithms are used to identify and correct anomalous data points. This ensures that the resulting distribution characteristics truly reflect actual consulting demand. Furthermore, as new consulting data is continuously collected and the smart contract model continues to operate, the consulting demand distribution characteristics will be updated in real time, ensuring that the system can adapt to ever-changing consulting demand scenarios.

[0101] Example 4:

[0102] When dynamically scheduling consulting resources, the first step is to map partition identifiers to the distribution characteristics of consulting demand. For example, a blockchain network deploys multiple consulting data nodes, each of which is divided into different data partitions based on geographic region. For example, Partition A corresponds to a city's central business district, and Partition B corresponds to a suburban residential area. Each partition is assigned a unique partition identifier. Once the smart contract model outputs the spatiotemporal distribution characteristics of consulting demand, the system maps Partition A's identifier to the region's high demand intensity between 9:00 AM and 11:00 AM on weekdays, and Partition B's identifier to the region's high demand intensity during weekend afternoons. This creates a clear mapping between identifiers and characteristics, facilitating the rapid identification of demand areas and characteristics during subsequent resource scheduling.

[0103] Then, based on the spatiotemporal distribution characteristics of consulting needs, the execution actions of consulting resources are controlled, including computing resource adjustment, data aggregation, and response allocation operations. Taking computing resource adjustment as an example, suppose that at 10 a.m. on a weekday, the system detects that the consulting demand intensity in partition A (city center business district) has reached the preset threshold, and at this time, the resource upgrade instruction of the adjacent computing resources is triggered. Specifically, partition A was originally responsible for data processing by 3 nodes, each node was allocated 2GB of memory and 1 CPU core. When the demand intensity exceeded the threshold, the system automatically temporarily allocated the computing resources of 2 idle nodes in the adjacent partition C (adjacent office area) to partition A, increased the node memory of partition A to 4GB, and increased the CPU cores to 2 to cope with the sudden increase in consulting requests.

[0104] During data aggregation operations, the system dynamically combines available data resources based on the data aggregation strategy to generate a response allocation vector. For example, if a user submits a request for financial advice in partition A, the system will aggregate the historical financial-related advice data, real-time data streams, and financial advice rules stored in the smart contract within partition A. Assuming that the aggregated data includes 100 financial advice records from partition A in the past hour, 20 financial advice requests currently being processed, and the financial advice priority rules specified in the smart contract, the system will combine this data in chronological order and by importance to generate a response allocation vector containing information such as data source, processing priority, and response time limit, such as "[Partition A historical data, current real-time request, financial rules; priority 1; time limit 10 minutes]".

[0105] Adjust the logical parameters of the target partition's processing nodes based on the response allocation vector. Taking partition A as an example, the response allocation vector specifies a priority of 1 and a time limit of 10 minutes for financial consultation requests. The processing nodes will adjust their logical parameters based on this vector. The original processing node's request queue cache size was 100, with a processing speed of 20 requests per minute. After the adjustment, the cache size was increased to 150, and the processing speed was increased to 30 requests per minute. At the same time, the scheduling priority of financial consultation requests was set to the highest to ensure that responses were completed within the specified time limit.

[0106] Finally, based on the spatial distribution of consultation demand and the pre-set resource scheduling strategy, consultation resources are dynamically allocated to the corresponding data partitions. For example, analysis of the partition distribution map revealed a significant increase in demand for home service consultations in Partition B (a suburban residential area) between 3:00 PM and 5:00 PM on weekends. The pre-set resource scheduling strategy stipulates that when the demand intensity in this area exceeds a threshold, the number of data aggregation nodes and response distribution nodes must be increased. The system automatically allocates two data aggregation nodes and one response distribution node from the resource pool to Partition B. The data aggregation node is responsible for collecting and integrating home service consultation data in this area, while the response distribution node generates responses based on the aggregated data and distributes them to users.

[0107] For example, a corporate user submits a complex request for legal advice in Zone C (office area). The system determines this request as high-priority based on its spatiotemporal distribution characteristics and requires extensive legal knowledge base data support. The system then dynamically combines the three compute nodes in Zone C, the two data retrieval nodes in Zone D (legal data storage area), and the legal rule execution node in the smart contract engine to form a dedicated resource processing group. The compute nodes are responsible for logical analysis of the request, the data retrieval nodes extract relevant cases and clauses from the legal knowledge base, and the smart contract engine performs rule verification and response generation. Data transmission and collaboration between these nodes are carried out through the blockchain network, ensuring that high-priority consultation requests are processed quickly and accurately.

[0108] During dynamic scheduling, the system monitors demand changes and resource usage for each data partition in real time. When the demand intensity for partition A gradually drops below the threshold after 12:00 noon on weekdays, the system automatically releases the previously allocated additional computing resources back to the resource pool for allocation to other partitions with increasing demand. At the same time, for data aggregation and response allocation operations, the system adjusts the aggregation strategy and allocation vector based on real-time data feedback. For example, if it finds that the response time for a certain type of consultation request is generally long, it will optimize the order of data aggregation and prioritize requests that take less time to process, thereby improving overall response efficiency.

[0109] Furthermore, resource scheduling rules predefined in the smart contract are applied throughout the entire scheduling process. For example, the smart contract stipulates that medical consultation requests take precedence over general consultation requests. When both medical consultation and general consultation requests appear in Zone E (the medical consultation concentration zone), the system prioritizes computing resources, data aggregation nodes, and response distribution nodes to the medical consultation request based on the smart contract rules, ensuring the speed and quality of medical consultation responses.

[0110] The entire dynamic scheduling process for consulting resources is decentralized and collaboratively handled through the blockchain network. Each consulting data node has real-time access to demand distribution characteristics and resource scheduling instructions, eliminating the single point of failure that can arise with centralized scheduling. Furthermore, the blockchain's immutable nature ensures transparency and traceability of the resource scheduling process, facilitating subsequent audits and optimization of scheduling efficiency and service quality.

[0111] Example 5:

[0112] When controlling the execution of consulting resources, when consulting demand in a target zone reaches a preset intensity threshold, the system automatically triggers resource upgrade instructions for neighboring computing resources. For example, suppose consulting data nodes in a blockchain network are divided into multiple zones based on urban area. Zone X corresponds to the downtown financial district, and the preset demand intensity threshold for this zone is 500 consulting requests per hour. At 10:00 AM on a weekday, the number of consulting requests in Zone X suddenly increases to 650 per hour, exceeding the preset threshold. The system detects this change through its smart contract engine and automatically sends resource upgrade instructions to neighboring Zones Y (the adjacent business district) and Z (the municipal office district). These instructions increase the CPU cores of two idle nodes in Zone Y from 2 to 4 cores and expand the memory from 8GB to 16GB. Furthermore, the storage bandwidth of three nodes in Zone Z is temporarily allocated to Zone X to enhance its data processing capacity. This entire process, executed automatically by the smart contract according to preset rules without manual intervention, ensures that the computing resources in the target zone can quickly respond to the demand peak.

[0113] At the same time, the system dynamically combines available data resources based on a data aggregation strategy to generate a response allocation vector. For example, a user submits a request for information about a wealth management product in partition X. This request involves multiple dimensions of data, including the product's historical yield, risk rating, and redemption rules. The data aggregation strategy prioritizes real-time market data within partition X, the product rule base stored in the smart contract, and historical transaction records from the adjacent partition A (the financial data storage area). Assume that the aggregated data includes: current real-time quotes for wealth management products in partition X (10 records / second), 20 product redemption rules stored in the smart contract, and 500 transaction records from the past three months in partition A. The system sorts this data by relevance and timeliness, generates a response allocation vector, and specifies the data sources as "real-time quotes from partition X + smart contract rules + historical transactions from partition A." It sets the processing priority to high and the response time limit to 8 minutes. This vector serves as the basis for subsequent resource scheduling.

[0114] Based on the response allocation vector, the system adjusts the logical parameters of the target partition's processing nodes. Continuing with the aforementioned scenario, the response allocation vector prioritizes financial product inquiries and requires responses within 8 minutes. The processing node in target partition X originally had a "first-come, first-served" request queue priority, a cache size of 200 requests, and a processing rate of 15 requests per minute. Based on the response allocation vector, the system adjusts the processing node's priority to "financial product inquiries first," expands the cache size to 300 requests, and reconfigures the node's processing threads through the smart contract engine, increasing the processing rate to 25 requests per minute. Furthermore, the node's logical parameters also adjust the data retrieval path. For example, the path originally used to retrieve product rules from the local database in partition X is temporarily replaced by a remote database retrieval path from partition A to ensure rapid access to complete historical transaction data.

[0115] For example, a user submits a complex consultation request regarding overseas study applications in Partition M (the education consultation hub). This request requires the integration of multiple data sources, including school databases, visa policies, and language test requirements. The data aggregation strategy dynamically combines the school information database in Partition M, the latest visa rules in Partition N (the immigration policy storage area), and the language test scoring standards stored in the smart contract. The resulting response allocation vector includes data from "Partition M School Database + Partition N Visa Rules + Smart Contract Language Standards," with a medium priority and a 15-minute response time limit. Based on this vector, the processing node in Partition M adjusts its logical parameters, increasing the number of concurrent threads for school data retrieval from 4 to 6 and shortening the visa policy cache time from 30 minutes to 10 minutes to ensure real-time data. It also switches the language test requirement matching algorithm from basic mode to enhanced mode to improve response accuracy.

[0116] In another scenario, during a peak nighttime public health emergency consultation period, demand in partition P (medical consultation area) exceeded twice the preset threshold. The system not only triggered computational resource upgrades in neighboring partitions Q and R, doubling the memory of the four nodes in partition Q and increasing the CPU frequency of the nodes in partition R by 50%, but also, based on a data aggregation strategy, urgently aggregated the latest diagnosis and treatment plans from partition S (medical knowledge base), the emergency response process in the smart contract, and real-time case data from partition T (epidemic data statistics area) to generate a response allocation vector. The processing nodes adjusted their logical parameters based on the vector, setting the highest priority for the emergency response process, disabling unnecessary data caching, and prioritizing all computing resources for processing epidemic consultation requests, ensuring that high-risk consultations received immediate responses.

[0117] Throughout this process, the smart contract's predefined rules guide the adjustment of resource execution. For example, the smart contract explicitly stipulates rules such as "the response time for medical consultation requests must not exceed 10 minutes" and "financial product consultations must utilize the latest market data." The system strictly adheres to these rules when generating response allocation vectors and adjusting logical parameters, ensuring compliance in resource scheduling. Furthermore, the distributed nature of the blockchain network enables each node to synchronize parameter adjustment instructions in real time, avoiding the command latency issues of traditional centralized systems and ensuring consistent and efficient resource execution.

[0118] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0119] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A smart contract-driven consulting data processing method, applied to a consulting data processing system, wherein the system comprises a plurality of consulting data nodes deployed on a blockchain network and a smart contract engine connected to the consulting data nodes, wherein two adjacent consulting data nodes are separated by a set distance, and wherein: The method comprises: Collect user consultation request data and response data through the consultation data node to generate a consultation data set; Receive real-time consulting data from multiple consulting data nodes through the smart contract engine and build a consulting data matrix; Determining the distribution characteristics of consulting demands based on the consulting data set and the real-time data of each consulting data node, wherein determining the distribution characteristics of consulting demands includes: processing the consulting data matrix, extracting demand characteristics in combination with the consulting data set, and predicting demand distribution based on smart contract execution information and data path information, outputting spatiotemporal distribution characteristics of consulting demands through a smart contract model, and updating the consulting data set based on the spatiotemporal distribution characteristics; According to the distribution characteristics, consulting resources are dynamically scheduled.

2. The smart contract-driven consulting data processing method according to claim 1, characterized in that: The determination of the distribution characteristics of consulting needs includes: Processing the consulting data matrix to extract data integrity indicators, response delay characteristics, and demand trend information; Performing data quality modeling on the consultation data matrix based on the data integrity indicators and response delay characteristics, dividing the consultation data into multiple data partitions and marking the partition identifiers, performing correlation matching with the consultation data set based on the integrity indicators of the data partitions, and marking the partition identifiers in the consultation data set; Calculating a data path gradient according to the location of the consultation data node, predicting the distribution of consultation demand according to the data path gradient and the demand trend information, and calculating demand forecast information for each data partition; Constructing a smart contract model, using the demand forecast information as an input parameter of the smart contract model, performing logical association modeling on the demand forecast information through the smart contract model, and outputting the spatiotemporal distribution characteristics of consulting demand; The consultation data set is updated according to the spatiotemporal distribution characteristics to obtain the distribution characteristics of the consultation needs.

3. The smart contract-driven consulting data processing method according to claim 2, characterized in that: The processing of the consultation data matrix comprises: Standardize the consulting data matrix, intercept the demand hotspot areas in the matrix through the time window, perform redundancy filtering on the hotspot areas, and calculate the demand trend information through the eigendecomposition algorithm; Calculating logical correlation features of the consultation data matrix, calculating the dependency strength, data stability coefficient, and missing data index between partitions based on the logical correlation features, constructing a feature fusion network, and calculating response delay features through the feature fusion network; The time domain features and semantic features collected by each consulting data node are extracted, and the demand feature vector of the node is calculated based on the deviation value of the time domain features and the semantic features. The consulting data nodes at different locations are feature matched according to the demand feature vector to calculate the demand trend information.

4. The smart contract-driven consulting data processing method according to claim 3, characterized in that: The data quality modeling of the consulting data matrix includes: Extract integrity sampling points from each frame of data based on the data integrity indicator, perform correlation mapping between the sampling points and response delay characteristics, generate a data quality map, logically align the data quality maps collected by multiple nodes, and calculate the data quality distribution of the consulting data; Set the integrity threshold value, locate data missing according to the integrity value of the multi-frame consultation data matrix, calculate the integrity difference, if the integrity difference is greater than or equal to the integrity threshold value, it means that there is data missing in the partition, perform data model constraint compensation on the current partition, iteratively correct the data quality distribution of the current partition according to the data verification model corresponding to the current partition, and calculate the quality compensation value of the missing data based on the correction result; Data quality modeling is performed on the consulting data matrix according to the data quality distribution, and delay annotation is performed on the partitioned data model according to the response delay characteristics.

5. The smart contract-driven consulting data processing method according to claim 4, characterized in that: Calculating the data path gradient according to the consultation data node position includes: Extract data change points based on multiple sets of consulting data transmission data, and map the change points to a unified data coordinate system based on the node deployment location. Fit the change points using a logical interpolation algorithm to generate a data path model for the consulting data. Performing sampling at equal intervals along the transmission path of the data path model, calculating a data attenuation rate, a dependency fluctuation index, and a demand change slope of the path based on the sampling results, and calculating a path change parameter based on the data attenuation rate, the dependency fluctuation index, and the demand change slope; Based on the deployment parameters and collection accuracy of the consultation data nodes, the distribution characteristics of consultation demand in each frame of data are projected onto the data path model. The data path model is partitioned according to the number of nodes along the transmission direction. The changing pattern of consultation demand within the partition is analyzed, and the demand distribution characteristics are calculated based on the changing pattern. The data path gradient is calculated based on the path change parameters and the demand distribution characteristics. The calculation process of the data path gradient includes: based on the logical position range from the first consultation data node to the last consultation data node, selecting data coordinate points in the node deployment direction, accumulating the product of the path characteristic weight value and the demand distribution characteristic weight value within the logical resolution range, and superimposing the impact value of the node acquisition frequency on the data path change rate.

6. The smart contract-driven consulting data processing method according to claim 5, characterized in that: The calculation of demand forecast information for each data partition includes: Taking the main transmission path of the data path model as the baseline and the peak position of the consultation demand in each frame of data as the reference point, the demand offset is calculated and the demand distribution curve is drawn according to the data coordinates; Correcting the growth rate and direction in the demand trend information based on the data path gradient; Starting from the most recent demand distribution point, the distribution curve is continuously drawn based on the correction results of growth rate and direction to generate the demand distribution points for the next period until the distribution points cover the entire target partition and generate demand forecast information.

7. The smart contract-driven consulting data processing method according to claim 2, characterized in that: The construction of the smart contract model includes: The input layer is used to organize demand forecast information into logically distributed data and perform standardization processing; The feature fusion layer is used to extract the logical association features of requirements by processing logically distributed data and build dependency relationships between data units; The resource allocation layer is used to integrate the association between consulting needs in logical units and generate consulting resource scheduling strategies.

8. The smart contract-driven consulting data processing method according to claim 2, characterized in that: The distribution characteristics of obtaining consulting needs include: According to the spatiotemporal distribution characteristics of the consulting needs output by the smart contract model, the identifiers of the data partitions are matched with the spatiotemporal distribution characteristics; The partition data in the consultation dataset are reorganized according to the spatiotemporal characteristics to generate a partition distribution map sorted by the consultation demand intensity; According to the reorganized partition distribution map, the optimized consultation demand distribution characteristics are output.

9. The smart contract-driven consulting data processing method according to claim 1, characterized in that: The dynamic scheduling of consulting resources includes: Mapping partition identifiers to partitions with distribution characteristics of consulting needs one by one; Control the execution of consulting resources based on the spatiotemporal distribution characteristics of consulting needs, including computing resource adjustment, data aggregation, and response allocation operations; Based on the spatial distribution of consulting needs and the preset resource scheduling strategy, consulting resources are dynamically allocated to the corresponding data partitions.

10. The smart contract-driven consulting data processing method according to claim 9, characterized in that: The execution actions of the control consulting resource include: When the consultation demand reaches a preset intensity threshold in the target partition, a resource promotion instruction for the adjacent computing resources is triggered; Dynamically combine available data resources according to data aggregation strategies to generate response allocation vectors; A logic parameter of the target partition processing node is adjusted based on the response allocation vector.

Images