Business analysis method and system based on big data

Abstract

The invention provides a business analysis method and system based on big data, and the method comprises the steps: obtaining user click stream data, payment stream data and an equipment electromagnetic interference signal; generating multi-source heterogeneous business time sequence data based on the user click stream data and the payment stream data; the device electromagnetic interference signal is converted into a device vibration frequency domain feature, a collaborative data stream is generated based on the device vibration frequency domain feature and the multi-source heterogeneous service time sequence data, and the collaborative data stream comprises a mixed service feature; decoupling the mixed service features by using an independent component analysis algorithm to obtain a flow density index and a bandwidth index; dynamically associating the traffic density index with the bandwidth index to construct a service independent feature matrix; and based on a streaming processing engine, converting the service independent feature matrix into a service feature analysis graph, and generating a decision strategy corresponding to the service feature analysis graph in real time. According to the invention, the real-time response capability of abnormal fluctuation in a high-concurrency service scene is improved.

Description

Technical Field

[0001] The present application relates to the field of big data technology, and in particular to a business analysis method and system based on big data. Background Art

[0002] In high-frequency financial trading and real-time risk control scenarios, systems must simultaneously process heterogeneous data from multiple sources, including transaction order streams, market data, and server operating status, and perform anomaly detection and risk response in sub-second time. These scenarios require analytical methods that can achieve deep integration and real-time processing of multi-dimensional data, while also balancing the coordinated analysis of business logic and hardware status.

[0003] The most advanced solution currently uses a multimodal data analysis framework based on deep learning. This uses a long-short-term memory network to process transaction time series data, coupled with a convolutional neural network to analyze server performance metrics. Finally, an attention mechanism is used to fuse the outputs of these two feature vectors. The system then establishes a dynamic risk assessment model and generates a risk score based on the fused feature vector.

[0004] The training and inference processes of deep learning models have large delays, making it difficult to meet the stringent real-time requirements of financial scenarios; the feature fusion of multimodal data remains at the later decision-making level, failing to achieve deep coupling of underlying signals; the generation of risk scores relies on static weights obtained through historical data training, and has poor adaptability to sudden market fluctuations. Summary of the Invention

[0005] The present application provides a business analysis method and system based on big data to solve the problem of low real-time response capability to abnormal fluctuations in high-concurrency business scenarios in the prior art.

[0006] In a first aspect, the present application provides a business analysis method based on big data, comprising: Obtain user clickstream data, payment flow data, and device electromagnetic interference signals; Generate multi-source heterogeneous business time series data based on the user click stream data and the payment flow data; Converting the device electromagnetic interference signal into a device vibration frequency domain feature, and generating a collaborative data stream based on the device vibration frequency domain feature and the multi-source heterogeneous service time series data, wherein the collaborative data stream includes a mixed service feature; Decoupling the mixed service characteristics using an independent component analysis algorithm to obtain a traffic density index and a bandwidth index; Dynamically associating the traffic density indicator with the bandwidth indicator to construct a service-independent feature matrix; Based on a stream processing engine, the business independent feature matrix is ​​converted into a business feature analysis diagram, and a decision strategy corresponding to the business feature analysis diagram is generated in real time.

[0007] Optionally, the decoupling of the mixed service characteristics by using an independent component analysis algorithm to obtain a traffic density index and a bandwidth index includes: Inputting the separation factor into a pre-trained blind source separation model, and separating the traffic density signal source and the bandwidth signal source from the mixed service characteristics through the blind source separation model; extracting a first fluctuation trajectory corresponding to the number of requests per minute from the traffic density signal source, analyzing the first fluctuation trajectory, and obtaining a traffic density index; A second fluctuation trajectory corresponding to the payment completion time is extracted from the bandwidth signal source, and the second fluctuation trajectory is analyzed to obtain a bandwidth index.

[0008] Optionally, inputting the separation factor into a pre-trained blind source separation model, and separating the traffic density signal source and the bandwidth signal source from the mixed service features by using the blind source separation model, includes: The traffic density weight and bandwidth weight in the separation factor are used as separation criteria and set in the decoupling layer of the blind source separation model; Performing orthogonal projection on the hybrid service characteristics through the decoupling layer to generate a feature vector space, wherein the feature vector space includes a traffic density feature component and a bandwidth feature component; determining a separating hyperplane according to a distribution difference between the traffic density characteristic component and the bandwidth characteristic component; Using the separating hyperplane, the feature vector space is divided into a first feature subspace and a second feature subspace; A flow density signal source is extracted from the first characteristic subspace, and a bandwidth signal source is extracted from the second characteristic subspace.

[0009] Optionally, using the separating hyperplane to divide the feature vector space into a first feature subspace and a second feature subspace includes: Calculating the projection distance of each data point in the feature vector space to the separating hyperplane along the normal vector direction of the separating hyperplane; Determining a boundary area according to the positive or negative nature of the projection distance; Based on the boundary region, dividing the feature vector space into two orthogonal subspaces; Determine the orthogonal subspace containing the traffic density characteristic component aggregation area in the two orthogonal subspaces as the first characteristic subspace; The orthogonal subspace containing the bandwidth characteristic component aggregation area in the two orthogonal subspaces is determined as the second characteristic subspace.

[0010] Optionally, generating a collaborative data stream based on the device vibration frequency domain characteristics and the multi-source heterogeneous service time series data includes: Performing waveform superposition on the device vibration frequency domain characteristics and the multi-source heterogeneous service time series data according to sampling time points to generate superimposed waveform data; Re-encapsulating the superimposed waveform data into a service data packet according to the transmission protocol of the service data stream through a signal injection mechanism; Performing timestamp alignment on the physical layer feature timestamp and the service data timestamp in the service data packet; Based on the aligned business data packets, a collaborative data stream containing mixed business features is output.

[0011] Optionally, dynamically associating the traffic density indicator with the bandwidth indicator to construct a service-independent feature matrix includes: Establishing a sliding time window, and calculating the fluctuation coupling degree between the traffic density indicator and the bandwidth indicator within the sliding time window; weighting the traffic density index value and the bandwidth index value respectively according to the fluctuation coupling degree; The weighted traffic density index values ​​and the weighted bandwidth index values ​​are arranged in time series to generate a service-independent feature matrix.

[0012] Optionally, the stream processing engine-based method of converting the service-independent feature matrix into a service feature analysis diagram and generating a decision strategy corresponding to the service feature analysis diagram in real time includes: Creating a corresponding virtual data channel for the business-independent feature matrix in a stream processing engine; In the virtual data channel, decomposing the service-independent feature matrix into a plurality of data blocks, and mapping each data block to an analysis node; Calculating the association strength between the analysis nodes, and establishing a connection relationship between the analysis nodes based on the association strength; Based on the connection relationship, construct a business feature analysis diagram; According to the topological structure characteristics of the business characteristic analysis diagram, a corresponding decision strategy is generated.

[0013] In a second aspect, the present application provides a business analysis system based on big data, comprising: The acquisition module is used to obtain user click stream data, payment flow data and device electromagnetic interference signals; A first generating module is configured to generate multi-source heterogeneous business time series data based on the user clickstream data and the payment transaction data; a conversion module, configured to convert the device electromagnetic interference signal into a device vibration frequency domain feature, and generate a collaborative data stream based on the device vibration frequency domain feature and the multi-source heterogeneous service time series data, wherein the collaborative data stream includes a mixed service feature; A decoupling module, configured to decouple the mixed service characteristics using an independent component analysis algorithm to obtain a traffic density index and a bandwidth index; A construction module, configured to dynamically associate the traffic density indicator with the bandwidth indicator to construct a service-independent feature matrix; The second generating module is used to convert the business independent feature matrix into a business feature analysis diagram based on a streaming processing engine, and generate a decision strategy corresponding to the business feature analysis diagram in real time.

[0014] In a third aspect, the present application provides a computing device comprising a processor and a memory, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute a big data-based business analysis method as described in any one of the first aspects.

[0015] In a fourth aspect, the present application provides a computer storage medium having computer program instructions stored thereon, which, when executed by a processor, implements a big data-based business analysis method as described in any one of the first aspects.

[0016] The technical solution provided by this application has the following beneficial effects: This application realizes the synchronous collection of business operation data and physical equipment status data, providing a complete data foundation for subsequent multi-dimensional analysis. Unify business data from different sources into a standardized time series format to eliminate the interference of data heterogeneity on analysis. Convert physical layer signals into frequency domain parameters that can be quantified and analyzed, revealing the potential correlation between equipment operation status and business data. Establish a deep coupling relationship between business data and equipment characteristics to form a data carrier containing mixed characteristics. Separate independent business indicators from complex data and eliminate cross-interference between features. Form a structured data representation through dynamic association of indicators, retaining the time dimension characteristics of business operation. Achieve intuitive and visual expression of business status and support real-time decision making.

[0017] Furthermore, this application also drives the blind source separation model through separation factors to accurately separate the traffic density signal source and bandwidth signal source from the mixed business characteristics, and then extract the request number fluctuation trajectory as the traffic density indicator and the payment time fluctuation trajectory as the bandwidth indicator.

[0018] In addition, the essential decoupling of business characteristics is achieved, so that the two core indicators of traffic load and transmission capacity remain independently analyzable, providing pure feature input for subsequent accurate decision-making.

[0019] These and other aspects of the present application will become more readily apparent from the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0021] Figure 1 A flowchart of a big data-based business analysis method provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of a big data-based business analysis system provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0023] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to being different types.

[0024] Current multimodal business analysis methods based on deep learning suffer from three key limitations: First, model inference latency fundamentally conflicts with the real-time requirements of high-concurrency scenarios, resulting in delayed risk response. Second, the shallow feature fusion of business data and physical layer signals struggles to capture the inherent correlation between equipment vibration and business fluctuations, limiting the sensitivity of anomaly detection. Finally, static weighting mechanisms are unable to adapt to sudden market fluctuations, causing a disconnect between risk scores and actual business status. These shortcomings stem from existing technologies' neglect of both deep data coupling and timeliness, necessitating the development of a deep collaborative analysis framework for business and physical signals.

[0025] In response to the above problems, this application proposes a business analysis method based on big data, which constructs a collaborative data stream containing mixed features through the cross-domain fusion of equipment electromagnetic interference signals and business time series data. This method innovatively uses the independent component analysis algorithm to decouple physical vibration characteristics and business indicators, realizes the pure extraction of traffic density and bandwidth indicators, and constructs a business feature matrix through dynamic association. Different from the batch processing mode of the existing technology, this solution converts feature maps in real time based on the streaming engine, directly compensates business data based on the physical layer signal drift, and simultaneously improves the timeliness and accuracy of analysis. This method has made a breakthrough in establishing a deep coupling relationship between equipment status and business indicators, making millisecond-level anomaly detection and adaptive decision-making possible, and completely solving the three major defects of the existing technology: response lag, feature fragmentation and lack of adaptability.

[0026] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0027] Figure 1 A flowchart of a business analysis method based on big data provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the method includes: Step 101: Obtain user clickstream data, payment flow data, and device electromagnetic interference signals.

[0028] In step 101, user clickstream data records users' browsing and clicking behaviors on the platform, including click time, page location, etc. Payment transaction data records successful order information, including amount, payment time, etc. Device electromagnetic interference signals represent electromagnetic radiation signals generated by the server during operation, reflecting the device's operating status.

[0029] In this embodiment, the system collects data from three channels simultaneously: front-end tracking to obtain user click behavior logs, payment gateway interface to obtain transaction records, and electromagnetic sensors to collect electromagnetic signals around the server. These three types of data are aligned with the collection time using a unified time base to ensure the timing consistency of subsequent processing.

[0030] For example, during a flash sale on e-commerce platform A, the system collects click records from users browsing product pages, including click timestamps and product numbers. It also simultaneously obtains the order number, payment amount, and payment success time returned by the payment system. Electromagnetic sensors deployed in server cabinets capture electromagnetic signal strength data at a fixed sampling frequency. All three types of data are timestamped at the moment of collection, and time synchronization errors are kept within acceptable limits.

[0031] Step 102: Generate multi-source heterogeneous business time series data based on the user click stream data and the payment transaction data.

[0032] In step 102, multi-source heterogeneous business time series data represents converting business data from different sources into a unified time series format.

[0033] In an embodiment of the present application, the user click stream data is aggregated and calculated by minute request volume, and the payment flow data is summarized by minute transaction amount. The two types of data are aligned and merged by timestamp to form a time series data table containing two dimensions: request volume and transaction amount per minute.

[0034] For example, the click data collected in step 101 is aggregated by minute to calculate the number of clicks on the product page per minute. Simultaneously, the payment data is aggregated by minute to calculate the total amount of successful transactions per minute. After aligning the two types of data using timestamps, a time series data table is generated containing three fields: time point, number of clicks, and transaction amount.

[0035] Step 103: Convert the device electromagnetic interference signal into a device vibration frequency domain feature, and generate a collaborative data stream based on the device vibration frequency domain feature and the multi-source heterogeneous service timing data, wherein the collaborative data stream includes a mixed service feature.

[0036] In step 103, the device vibration frequency domain characteristics represent the device vibration frequency distribution characteristics obtained by converting the electromagnetic signal. The collaborative data stream represents the fusion data carrier of business data and device characteristics. The hybrid business characteristics represent the comprehensive characteristics that include business indicators and device status.

[0037] In an embodiment of the present application, a spectrum analysis is performed on the electromagnetic signal, and the energy distribution of the main vibration frequency band is extracted as the device feature; the feature is matched with the business timing data according to the time point, and a new data stream containing business and device features is generated through weighted fusion.

[0038] For example, the electromagnetic signal is subjected to fast Fourier transform to obtain a spectrum diagram, and the three frequency bands with the highest energy are selected as features; these feature values ​​are combined with the business data at the corresponding moment, and the business data is multiplied by the adjustment coefficient based on the device characteristics to generate a fused data record.

[0039] Step 104: Decouple the mixed service characteristics using an independent component analysis algorithm to obtain a traffic density index and a bandwidth index.

[0040] In step 104, the traffic density index is a quantitative index reflecting the pressure of service requests, and the bandwidth index is a quantitative index reflecting the processing capacity of the system.

[0041] In an embodiment of the present application, an independent component analysis algorithm is used to process mixed features, separation parameters are set to project the data into a new space, and a projection direction that makes the features independent of each other is found. The components related to the request amount are extracted as traffic density indicators, and the components related to the processing time are extracted as bandwidth indicators.

[0042] For example, a separation model is established for the fused data, and the optimal projection direction is found through iterative calculation, and the data is decomposed into two parts: one part mainly contains information on changes in request volume as a traffic density indicator, and the other part mainly contains information on processing time as a bandwidth indicator.

[0043] Step 105: Dynamically associate the traffic density index and the bandwidth index to construct a service-independent feature matrix.

[0044] In step 105, the business independent feature matrix represents a matrix representation of the indicator association result.

[0045] In an embodiment of the present application, a sliding time window is established, and the covariance relationship between the traffic density index and the bandwidth index is calculated within the window. The index values ​​are dynamically weighted according to the strength of the correlation, and the weighted results are arranged in chronological order to form a feature matrix.

[0046] For example, take a 5-minute sliding window, calculate the covariance between the number of requests per minute and the processing time, use the covariance value as the weight to adjust the original indicator value, and arrange the adjusted values ​​into a matrix in chronological order, with each row representing a time point and each column representing an indicator.

[0047] Step 106: Based on the stream processing engine, the business independent feature matrix is ​​converted into a business feature analysis diagram, and a decision strategy corresponding to the business feature analysis diagram is generated in real time.

[0048] In step 106, the service characteristic analysis diagram represents a graphical representation of the service status. The decision strategy represents an operation instruction generated according to the analysis result.

[0049] In an embodiment of the present application, the feature matrix is ​​input into the stream processing engine, the matrix rows are converted into graph nodes, the similarity between the nodes is calculated to establish edge relationships, abnormal patterns are identified based on the graph structure characteristics, and corresponding expansion or current limiting instructions are generated.

[0050] For example, each row of the feature matrix is ​​regarded as a node, and the similarity of the request volume and processing time between nodes is calculated. Nodes with high similarity are connected to form a business status relationship graph. When an abnormal connection pattern is found, an automatic adjustment strategy is triggered.

[0051] This method achieves coordinated monitoring of business status and equipment operation through synchronous collection and deep integration of multi-source data; adopts feature decoupling technology to obtain pure business indicators and avoid cross-interference; ensures the real-time and accuracy of analysis results through dynamic association and streaming processing; and the final visual analysis chart provides an intuitive basis for decision-making, improving the efficiency of business anomaly detection and processing in high-concurrency scenarios.

[0052] To solve the problem of deep decoupling of service data and device status features and further improve the accuracy of index extraction, in some embodiments, step 104: using an independent component analysis algorithm to decouple the mixed service features to obtain a traffic density index and a bandwidth index includes: Step 201: Input the separation factor into a pre-trained blind source separation model, and separate the traffic density signal source and the bandwidth signal source from the mixed service features through the blind source separation model.

[0053] In step 201, the separation factor is a pre-set combination of weight parameters, which includes two parts: traffic density weight value and bandwidth weight value. Its design basis is the statistical distribution difference between the traffic density component and the bandwidth component in the mixed service characteristics. The specific value is determined by calculating the variance ratio and correlation coefficient of the two types of components. The core meaning of this factor is to guide the blind source separation model to distinguish and extract the decoupling criteria of the traffic density signal source and the bandwidth signal source. The blind source separation model refers to a computational model that can separate independent components from a mixed signal, and its decoupling layer realizes feature space reconstruction through orthogonal transformation. The traffic density signal source refers to the data component that mainly reflects the service request pressure after separation. The bandwidth signal source refers to the data component that mainly reflects the system processing capability after separation.

[0054] In an embodiment of the present application, the pre-calculated separation factor parameters are configured to the model decoupling layer. After the model receives the mixed business feature input, the data is projected into a new feature space through an orthogonal transformation. In this space, the optimal splitting surface is found according to the direction indicated by the separation factor, and the mixed data is separated into two independent data sets, which are output as traffic density signal source and bandwidth signal source respectively.

[0055] Step 202: extracting a first fluctuation trajectory corresponding to the number of requests per minute from the traffic density signal source, analyzing the first fluctuation trajectory, and obtaining a traffic density index.

[0056] In step 202, the first fluctuation trajectory refers to the curve shape of the number of requests in the traffic density signal source changing over time. The traffic density index refers to a quantitative evaluation value obtained by analyzing the curve.

[0057] In an embodiment of the present application, a change curve of the number of requests per minute is extracted from the separated traffic density signal source, and elements such as the peak and trough characteristics and the change trend of the curve are analyzed. A comprehensive evaluation value reflecting the business pressure status is calculated through a specific algorithm as a traffic density indicator.

[0058] Step 203: extracting a second fluctuation trajectory corresponding to the payment completion time from the bandwidth signal source, analyzing the second fluctuation trajectory, and obtaining a bandwidth index.

[0059] In step 203, the second fluctuation trajectory refers to the curve shape of the payment completion time in the bandwidth signal source over time. The bandwidth index refers to the quantitative evaluation value obtained by analyzing the curve.

[0060] In an embodiment of the present application, a curve of payment completion time variation is extracted from the separated bandwidth signal source, and elements such as the time distribution characteristics and abnormal fluctuation points of the curve are analyzed. A comprehensive evaluation value reflecting the system processing capability is calculated as a bandwidth indicator through a specific algorithm.

[0061] Here's a specific example: In the flash sale activity of e-commerce platform A, the system inputs the generated fusion data records into the pre-trained blind source separation model. The separation factor received by the model consists of a traffic density weight of 0.7 and a bandwidth weight of 0.3. These two weights are calculated by calculating the variance ratio of the number of requests and payment completion time in the historical data. The variance ratio calculation formula is: ,in represents the variance ratio, represents the variance of the number of requests, represents the variance of payment completion time; the model first projects the fused data into a new coordinate system, determines the optimal separation direction by calculating the eigenvector of the covariance matrix, and decomposes the data into two parts: the first part mainly contains fluctuating data with the number of requests per minute in the range of 1200-1500 times as the traffic density signal source, and the second part mainly contains fluctuating data with the payment completion time in the range of 200-400 milliseconds as the bandwidth signal source; the per-minute change curve of the number of requests is extracted from the traffic density signal source, and the average fluctuation amplitude of the curve is calculated as the traffic density indicator, specifically taking the average of the difference between the maximum and minimum number of requests per minute; the per-minute change curve of the payment completion time is extracted from the bandwidth signal source, and the fluctuation frequency of the curve is calculated as the bandwidth indicator, specifically counting the number of times the payment completion time exceeds 300 milliseconds per minute; the final traffic density indicator reflects the system load pressure, and the bandwidth indicator reflects the transaction processing efficiency. Both indicators are used for subsequent business status analysis and decision-making.

[0062] In the embodiment of the present application, the method achieves deep decoupling of business characteristics and device characteristics through precise signal separation technology, and the obtained indicator data is purer and more reliable; the dynamic weight mechanism ensures that the decoupling process can adapt to the characteristics of different business scenarios; the final traffic density and bandwidth indicators provide an accurate basis for system status evaluation, effectively supporting subsequent business decisions.

[0063] In order to solve the problem of accurate separation of signal sources in mixed service features and further improve the accuracy of decoupling, in some embodiments, step 201: inputting the separation factor into a pre-trained blind source separation model, and separating the traffic density signal source and the bandwidth signal source from the mixed service features by the blind source separation model, includes: Step 301: The traffic density weight and bandwidth weight in the separation factor are used as separation criteria and set in the decoupling layer of the blind source separation model.

[0064] In step 301, the traffic density weight is a parameter that reflects the importance of service request pressure. A larger value indicates a higher weight for that feature in the separation process. The bandwidth weight is a parameter that reflects the importance of system processing power. Its value determines the priority of the corresponding feature during separation. The decoupling layer is the network layer in the blind source separation model specifically responsible for feature decomposition. It distinguishes different features by adjusting internal parameters.

[0065] In an embodiment of the present application, the calculated traffic density weight and bandwidth weight parameters are input into the configuration unit of the decoupling layer, which adjusts the parameter settings of the internal projection matrix according to the weight ratio, so that the model can distinguish different features according to the preset importance in subsequent processing.

[0066] Step 302: Perform orthogonal projection on the hybrid service characteristics through the decoupling layer to generate a feature vector space, where the feature vector space includes a traffic density feature component and a bandwidth feature component.

[0067] In step 302, orthogonal projection refers to a spatial transformation method that maintains the original relationship of the data. The feature vector space refers to the mathematical space in which the data is located after projection, and each dimension represents a characteristic component. The traffic density characteristic component refers to the mathematical representation of the data representing the request load in the mixed service feature after orthogonal projection in the feature vector space. The bandwidth characteristic component refers to the mathematical representation of the data representing the transmission capacity in the mixed service feature after orthogonal projection in the feature vector space. Both are converted from the original mixed service feature through the orthogonal projection processing of the decoupling layer. The traffic density signal source is a set of pure request load signals extracted from the first characteristic subspace, and its data basis is the distribution of the traffic density characteristic component in this subspace. The bandwidth signal source is a set of pure transmission capacity signals extracted from the second characteristic subspace, and its data basis is the distribution of the bandwidth characteristic component in this subspace. The characteristic component is the manifestation of the signal source in the mathematical space, and the signal source is the physical entity obtained after the characteristic component is spatially segmented. In an embodiment of the present application, a decoupling layer with configured parameters receives mixed business feature input and converts the original data into a new coordinate system through matrix operations. In the new coordinate system, each data point is represented by a set of feature vectors, which mainly include two types of components: vectors reflecting request pressure and vectors reflecting processing capacity.

[0068] Step 303: Determine a separating hyperplane based on the distribution difference between the traffic density characteristic component and the bandwidth characteristic component.

[0069] In step 303, the distribution difference refers to the degree of aggregation and relative position relationship of different feature components in space. A separating hyperplane refers to a mathematical plane that can best distinguish two types of feature components.

[0070] In an embodiment of the present application, the aggregation area and distribution range of two types of feature components in the feature vector space are analyzed, their center point distances and distribution discreteness are calculated, and the plane equation that can separate the two types of features to the greatest extent is found. This plane is the separating hyperplane.

[0071] Step 304: using the separating hyperplane, split the feature vector space into a first feature subspace and a second feature subspace.

[0072] In step 304, the first characteristic subspace refers to a spatial region that mainly contains a flow density characteristic component, and the second characteristic subspace refers to a spatial region that mainly contains a bandwidth characteristic component.

[0073] In an embodiment of the present application, based on the determined separating hyperplane equation, the positional relationship of each point in the space relative to the plane is calculated, and the points on one side of the plane are classified into the first characteristic subspace, and the points on the other side are classified into the second characteristic subspace, thereby realizing the division of the spatial area.

[0074] Step 305: extracting a flow density signal source from the first characteristic subspace, and extracting a bandwidth signal source from the second characteristic subspace.

[0075] In step 305, the traffic density signal source refers to the pure business pressure feature data extracted from the first feature subspace, and the bandwidth signal source refers to the pure system processing capacity feature data extracted from the second feature subspace.

[0076] In an embodiment of the present application, data extraction is performed on the two divided feature subspaces respectively, and the data points in the first feature subspace are converted back to the original feature form as the flow density signal source, and the data points in the second feature subspace are converted back to the original feature form as the bandwidth signal source.

[0077] Here's a specific example: In the promotion activities of e-commerce platform C, the system configures the prepared separation factor parameters into the blind source separation model, where the traffic density weight of 0.65 and the bandwidth weight of 0.35 are calculated by analyzing historical promotion data. The specific calculation method is to take the average of the ratio of the fluctuation amplitude of the number of requests and the payment delay time in the past three promotion activities; after the model receives the mixed data containing the number of requests per minute and the payment time, it first projects the data into a two-dimensional feature space through orthogonal transformation, where the horizontal axis mainly reflects the request volume feature and the vertical axis mainly reflects the processing time feature; by calculating the distribution center distance of the two types of feature components Determine the optimal splitting line, where Represents the distribution center distance of the two types of feature components, is the distribution center of the request volume feature, The distribution center of the processing duration feature; select the straight line based on the calculation results As the separating hyperplane, the feature space is divided into two areas; the data points in the upper left area mainly correspond to the feature combination of 1250-1550 requests / minute and 180-350 milliseconds of payment time, which are converted back to the original data as the traffic density signal source; the data points in the lower right area mainly correspond to the feature combination of 900-1200 requests / minute and 380-450 milliseconds of payment time, which are converted as the bandwidth signal source.

[0078] In the embodiment of the present application, the method achieves efficient decoupling of mixed business features through weight-guided spatial projection and precise division; the separated signal source maintains the timing characteristics and physical meaning of the original data; the feature separation method based on mathematical space avoids the deviation caused by subjective judgment and provides a reliable data basis for subsequent analysis.

[0079] To further improve the accuracy of feature space segmentation, in some embodiments, step 304: using the separating hyperplane to segment the feature vector space into a first feature subspace and a second feature subspace includes: Step 401: Calculate the projection distance from each data point in the feature vector space to the separating hyperplane along the normal vector direction of the separating hyperplane.

[0080] In step 401, the normal vector direction refers to the perpendicular direction vector of the separating hyperplane. It is derived from the optimal separation direction calculated based on the distribution difference between the traffic density characteristic component and the bandwidth characteristic component, and is used to define the reference axis for spatial segmentation. The data points include all projection points of the traffic density characteristic component and the bandwidth characteristic component in the characteristic vector space, which are the mathematical representation of the mixed service characteristics after orthogonal projection. The projection distance refers to the perpendicular distance from the data point to the hyperplane along the normal vector direction. Its numerical value reflects the distance between the data point and the hyperplane.

[0081] In an embodiment of the present application, based on the determined separating hyperplane equation, the vertical distance of each data point in the feature space to the plane along a specific direction is calculated, and this distance value will serve as the basis for subsequent space division.

[0082] Step 402: Determine a boundary area according to the positive or negative value of the projection distance.

[0083] In step 402, the boundary area refers to the transition area where the projection distance value is close to zero, which contains data points that may have both types of features; the positivity refers to the sign of the projection distance value, which is used to preliminarily determine the category of the data point.

[0084] In an embodiment of the present application, the projected distance values ​​of all data points are sign-judged, and positive and negative values ​​are classified separately. At the same time, a narrow area around the hyperplane is defined as a boundary buffer zone, and the data points in this area need to be further analyzed and confirmed.

[0085] Step 403: Based on the boundary region, divide the feature vector space into two orthogonal subspaces.

[0086] In step 403 , orthogonal subspaces refer to mutually perpendicular and non-overlapping spatial regions, and each subspace mainly contains data points of a certain type of features.

[0087] In the embodiment of the present application, the feature space is divided into two non-interfering regions with the demarcation region as the boundary, ensuring that the data points in each region have similar feature attributes.

[0088] Step 404: Determine the orthogonal subspace containing the concentration area of ​​the flow density characteristic component in the two orthogonal subspaces as the first characteristic subspace.

[0089] In step 404, the flow density characteristic component aggregation area refers to the continuous spatial range where the projection point density of the flow density characteristic component in the characteristic vector space is higher than the bandwidth characteristic component, which is obtained by calculating the kernel density estimate of the projection points and detecting the density peak area.

[0090] In an embodiment of the present application, the distribution of data points in the divided subspace is analyzed, and the area where the request volume feature points are concentrated is identified, and the area is determined as the first feature subspace that mainly reflects the business pressure.

[0091] Step 405: Determine the orthogonal subspace containing the bandwidth characteristic component aggregation area in the two orthogonal subspaces as the second characteristic subspace.

[0092] In step 405, the bandwidth characteristic component clustering area refers to a continuous spatial range where the projection point density of the bandwidth characteristic component in the characteristic vector space is higher than that of the flow density characteristic component, and is automatically identified through spectrum energy distribution analysis and spatial clustering algorithm.

[0093] In an embodiment of the present application, the distribution of data points in another subspace is analyzed, and an area where processing time feature points are concentrated is identified, and this area is determined as a second feature subspace that mainly reflects the system processing capability.

[0094] Here's a specific example: In the limited-time flash sale event of e-commerce platform E, the system is based on the determined separating hyperplane To carry out feature space segmentation, first calculate the projection distance of each data point in the feature space to the hyperplane, where the projection distance calculation formula is: , Represents the projection distance of the data point to the separating hyperplane; through analysis, it is found that the projection distance of the request quantity feature data points is mainly concentrated in the range of 0.3 to 1.2, and the projection distance of the processing time feature data points is mainly in the range of -0.4 to -1.1; the area with the absolute value of the projection distance less than 0.15 is defined as the dividing area, which contains about 5% of transitional data points; the feature space is divided into two subspaces with the dividing area as the boundary. The left subspace contains 85% of the request quantity feature data points and these points are mainly concentrated in the horizontal coordinate range of 1.1 to 1.5 and the vertical coordinate range of 0.8 to 1.2, so it is determined to be the first feature subspace; the right subspace contains 82% of the processing time feature data points and are mainly concentrated in the horizontal coordinate range of 0.7 to 1.1 and the vertical coordinate range of 1.3 to 1.7, so it is determined to be the second feature subspace.

[0095] In the embodiment of the present application, the method achieves a clear separation of business features through precise spatial projection and area division; the setting of the boundary area avoids the misclassification of edge data; the two feature subspaces finally obtained completely retain the core features of the original data, providing a high-quality data foundation for subsequent analysis.

[0096] To further improve the fusion effect of business data and device status data, in some embodiments, step 103: generating a collaborative data stream based on the device vibration frequency domain characteristics and the multi-source heterogeneous business time series data includes: Step 501: superimpose the device vibration frequency domain characteristics and the multi-source heterogeneous service time series data on waveforms according to sampling time points to generate superimposed waveform data.

[0097] In step 501, the sampling time point refers to the same time base extracted from the device vibration frequency domain characteristics and multi-source heterogeneous service time series data. This is derived from the synchronized clock signal preset by the original acquisition system of the two types of data, ensuring time axis alignment during waveform superposition. Waveform superposition is a processing method that numerically fuses data from different sources after time alignment. The superimposed waveform data refers to the fused data waveform, which contains both service and device characteristic information.

[0098] In an embodiment of the present application, the device vibration frequency domain characteristic values ​​are corresponded to the business timing data at the same time point, and the two types of data are merged into new waveform data through weighted calculation to ensure that the data at each time point reflects both the business status and the device status.

[0099] Step 502: Through a signal injection mechanism, the superimposed waveform data is re-encapsulated into a service data packet according to a transmission protocol of a service data stream.

[0100] In step 502, the signal injection mechanism refers to a technique for embedding processed data into a business data stream in a specific format. The transmission protocol for the business data stream refers to the data encapsulation rules followed when payment transaction data and user clickstream data are exchanged between the collection and processing ends. This protocol is derived from the communication standards pre-set by the business system (such as ISO 8583 in the financial sector or the HTTP / HTTPS data format for e-commerce platforms). A business data packet is a data unit encapsulated according to the communication protocol, containing valid data and protocol header information.

[0101] In an embodiment of the present application, the superimposed waveform data is reorganized according to the data format specified by the business system, and necessary timestamps and data type identifiers are added to form a standardized data packet that meets the transmission requirements.

[0102] Step 503: Perform timestamp alignment on the physical layer feature timestamp and the service data timestamp in the service data packet.

[0103] In step 503, the service data packet contains both the physical layer feature timestamp (derived from the acquisition time of the device's vibration frequency domain features) and the service data timestamp (derived from the generation time of the user's clickstream / payment flow), both of which exist as independent fields within the same data packet. Timestamp alignment is the process of adjusting the time stamps of data from different sources to achieve time synchronization.

[0104] In an embodiment of the present application, the time stamps of the device characteristics and service characteristics in the data packet are checked, and the timestamps with slight differences are calibrated to ensure that the two types of data are completely synchronized in the time dimension.

[0105] Step 504: Based on the aligned service data packets, a collaborative data stream containing mixed service features is output.

[0106] In an embodiment of the present application, the time-calibrated data packets are arranged in chronological order, and after removing redundant information, they are output as a continuous data stream, which retains all key information of the original business data and device characteristics.

[0107] Here's a specific example: During a limited-time sale on e-commerce platform G, the system fused the vibration characteristic values ​​of 0.7, 1.1, and 0.8 in the three main frequency bands collected by the electromagnetic sensor with 1,500 page clicks per minute and a total transaction volume of 60,000 yuan. The specific processing process is as follows: First, the vibration characteristic values ​​are normalized to fall within the range of 0 to 1. The normalization formula is: ,in represents the normalized vibration eigenvalue, represents the original vibration value, and Represent the minimum and maximum vibration values ​​in the historical data respectively; then the normalized vibration value is multiplied by the adjustment coefficient 0.015 and added to the click volume, and then multiplied by 0.00012 and added to the transaction amount to generate superimposed waveform data; then, according to the platform's customized data transmission protocol, a protocol header is added to each superimposed data, including information such as timestamp, data type identifier and data length; inspection found that the vibration data timestamp is 15 milliseconds faster than the business data on average, and all vibration feature timestamps are calibrated with a delay of 15 milliseconds; the final output is a collaborative data stream containing the calibration timestamp, adjusted click volume, adjusted transaction amount and original vibration feature value, generating a complete record every minute, where adjusted click volume = original click volume + 0.015 × normalized vibration value, and adjusted transaction amount = original transaction amount + 0.00012 × normalized vibration value. These data will be used for subsequent business status analysis and anomaly detection.

[0108] In the embodiment of the present application, the method achieves deep coupling of business data and device status data through precise time alignment and numerical fusion; the generated collaborative data flow not only retains the characteristics of the original data, but also establishes the association relationship between the business and the device; the standardized data format ensures that the subsequent processing modules can efficiently parse and use this data, providing a more comprehensive data foundation for business status analysis.

[0109] To further improve the dynamic correlation analysis capability between service indicators, in some embodiments, step 105: dynamically correlating the traffic density indicator and the bandwidth indicator to construct a service-independent feature matrix includes: Step 601: Establish a sliding time window, and calculate the fluctuation coupling degree between the traffic density index and the bandwidth index within the sliding time window.

[0110] In step 601, the sliding time window refers to the data analysis interval that moves along the time axis and is used to capture the short-term variation characteristics of the indicator. The volatility coupling degree refers to the degree of coordinated variation between two indicators within the time window, reflecting their dynamic correlation.

[0111] In an embodiment of the present application, a time window of fixed length is set, and the coordinated changes of the traffic density index and the bandwidth index are calculated within the window. The coupling degree value between the two is obtained through a specific association algorithm. The larger the value, the stronger the linkage between the indicators.

[0112] Step 602: weighting the traffic density index value and the bandwidth index value respectively according to the fluctuation coupling degree.

[0113] In step 602, weighting refers to the process of adjusting the numerical weights of indicators based on the strength of the relationship between them. The traffic density index value and bandwidth index value are derived from the intensity change trajectory of the traffic density signal source and the spectrum width change trajectory of the bandwidth signal source, respectively, extracted by decoupling the hybrid service characteristics. They are quantized time series data points. The traffic density index value is a specific numerical instance of the traffic density indicator within the sliding time window, and the bandwidth index value is a specific numerical instance of the bandwidth indicator within the sliding time window. Both are quantitative representations of the indicators in the time dimension.

[0114] In an embodiment of the present application, the calculated fluctuation coupling degree is used as an adjustment coefficient to proportionally adjust the traffic density index value and the bandwidth index value within the window, so that the adjusted values ​​can reflect the actual correlation strength between the indicators.

[0115] Step 603: Arrange the weighted traffic density index values ​​and the weighted bandwidth index values ​​in time series to generate a service-independent feature matrix.

[0116] In an embodiment of the present application, the weighted index values ​​calculated in each time window are arranged in chronological order, with the traffic density index as the first column and the bandwidth index as the second column, to construct a structured data matrix for subsequent analysis and processing.

[0117] Here's a specific example: In the limited-time flash sale event of e-commerce platform I, the system uses a 5-minute sliding window to perform dynamic correlation analysis on the traffic density indicator and bandwidth indicator. The specific implementation process is as follows: First, in the first time window, the data series of the number of requests per minute and the payment completion time are obtained. The request number series is 1300, 1400, 1500, 1450, and 1350 times / minute, and the payment completion time series is 220, 240, 260, 250, and 230 milliseconds; the covariance value of the two series is calculated using the formula ,in express and The covariance value of Represents a sequence of request times, represents the payment time series, and represent the means of the two series, is the number of data points in the window, 5, Indicates the number of requests in the sequence elements, Indicates the payment completion time series elements, the calculated covariance value is 1250; the covariance value is converted into fluctuation coupling using the formula ,in and where represents the standard deviation of the two sequences, and the coupling degree is calculated to be 0.68. The coupling degree is used to weight the number of requests, that is, the original value is multiplied by 0.68, and the adjusted number of requests is 884, 952, 1020, 986, and 918 times / minute. The payment completion time is weighted by multiplying it by 0.32, and the adjusted payment completion time is 70.4, 76.8, 83.2, 80.0, and 73.6 milliseconds. These adjusted values ​​are arranged in chronological order to form the first row of data in the business independent feature matrix. After the window slides for 1 minute, the data in the new window is processed in the same way to generate the second row of data in the matrix. This process is continued until all monitoring periods are covered, and finally a complete business independent feature matrix is ​​constructed. Each row of data reflects the dynamic correlation between the request volume and processing time, providing data support for subsequent business status analysis and decision-making.

[0118] In the embodiment of the present application, the method accurately captures the real-time correlation between business indicators through dynamic analysis of the sliding window; the weighted adjustment based on the coupling degree enables the feature matrix to better reflect the actual operating status of the system; the structured matrix form provides a clear and standardized data basis for subsequent business status analysis and decision-making.

[0119] To further improve the real-time performance and visualization of service status analysis, in some embodiments, step 106: converting the service-independent feature matrix into a service feature analysis graph based on a stream processing engine and generating a decision strategy corresponding to the service feature analysis graph in real time includes: Step 701: Create a corresponding virtual data channel for the service-independent feature matrix in a stream processing engine.

[0120] In step 701, the virtual data channel refers to an independent processing path allocated to a specific data stream in the stream processing engine to ensure the isolation and orderliness of data processing.

[0121] In an embodiment of the present application, a dedicated processing channel is created for the input matrix data in the streaming processing engine, and corresponding computing resources and cache space are configured to provide a stable data flow environment for subsequent analysis.

[0122] Step 702: In the virtual data channel, decompose the service-independent feature matrix into multiple data blocks, and map each data block to an analysis node.

[0123] In step 702, a data block refers to a local data unit that divides a matrix into rows or columns. An analysis node refers to a graphical element that represents a single data block and contains all feature information of the data block.

[0124] In an embodiment of the present application, the feature matrix is ​​divided into several data blocks in chronological order, each data block contains indicator data of a fixed time span, and then a corresponding analysis node is created for each data block, and the node attributes include the request volume and processing time characteristics within the time period.

[0125] Step 703: Calculate the association strength between the analysis nodes, and establish a connection relationship between the analysis nodes based on the association strength.

[0126] In step 703, the association strength refers to the similarity of business characteristics between nodes. The connection relationship refers to the visual link between nodes established based on the association strength.

[0127] In an embodiment of the present application, the similarity between every two analysis nodes in terms of request volume and processing time is calculated, and node pairs whose similarity exceeds a set threshold are connected, and the thickness of the connection line reflects the strength of the association.

[0128] Step 704: Construct a service feature analysis diagram based on the connection relationship.

[0129] In an embodiment of the present application, all analysis nodes are arranged in chronological order, and connection lines are drawn according to the established connection relationships to form a feature map that can reflect the law of business status changes.

[0130] Step 705: Generate a corresponding decision strategy based on the topological structure characteristics of the business characteristic analysis graph.

[0131] In step 705, the topological structure characteristics of the business feature analysis graph refer to the network structure characteristics formed by analyzing the connection relationship between nodes, which are derived from the node connection relationship established based on the association strength, and specifically include quantitative indicators such as the connection density between nodes, connection weight distribution, and network clustering characteristics.

[0132] In an embodiment of the present application, special patterns such as densely connected areas and isolated nodes in the feature map are analyzed, and when abnormal connection patterns are detected, corresponding control instructions such as server expansion and traffic restriction are automatically generated.

[0133] Here's a specific example: During a flash sale on e-commerce platform K, the system fed a constructed business-independent feature matrix into a streaming engine for real-time analysis. The matrix contained 15 rows, one per minute, of adjusted request counts and payment completion times over a 15-minute period. The streaming engine first created a dedicated processing channel for the matrix data to ensure that processing was uninterrupted by other tasks. The matrix was then partitioned into three data blocks, each corresponding to a 5-minute span. Each block was converted into an analysis node, whose attributes contained the average request count and payment completion time for that 5-minute period. The first node corresponded to the first 5 minutes of data, with an average request count of 1420 per minute and an average payment completion time of 235 milliseconds. The second node corresponded to the middle 5 minutes of data, with an average request count of 1560 per minute and an average payment completion time of 265 milliseconds. The third node corresponded to the last 5 minutes of data, with an average request count of 1680 per minute and an average payment completion time of 310 milliseconds. The similarity between nodes is calculated using the formula: Similarity = 1 - [|Request Count Difference| / (Maximum Request Count Difference) + |Payment Time Difference| / (Maximum Payment Time Difference)] / 2. The request count difference and payment time difference refer to the differences in the corresponding attributes of the two nodes. The maximum request count difference is the difference between the maximum and minimum request counts in the matrix, 260 requests / minute, and the maximum payment time difference is 90 milliseconds. The calculated similarity between nodes 1 and 2 is 0.72, the similarity between nodes 2 and 3 is 0.65, and the similarity between nodes 1 and 3 is 0.58. A similarity threshold of 0.6 is set, and connections are established between nodes 1-2 and 2-3, with the thickness of the connection line proportional to the similarity. The constructed business feature analysis diagram shows a significant increase in payment completion time for node 3 and a weakening of the connection strength. Upon detecting this abnormal pattern, the system automatically generates decision policies: "Immediately increase server resources by 20%" and "Activate the queuing system." These policies are directly issued to the resource management platform via an interface call for execution.

[0134] In an embodiment of the present application, the method realizes real-time graphical display of business status through a streaming processing engine; the feature map based on node association can intuitively reveal the evolution trend of the system operation status; the automated policy generation mechanism ensures a rapid response to abnormal situations, effectively improving the stability and reliability of the business system.

[0135] Figure 2 A schematic diagram of a business analysis system based on big data provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, the system includes: The acquisition module 21 is used to acquire user click stream data, payment flow data and device electromagnetic interference signals.

[0136] The first generating module 22 is configured to generate multi-source heterogeneous business time series data based on the user clickstream data and the payment transaction data.

[0137] The conversion module 23 is used to convert the device electromagnetic interference signal into a device vibration frequency domain feature, and generate a collaborative data stream based on the device vibration frequency domain feature and the multi-source heterogeneous service timing data, wherein the collaborative data stream includes a mixed service feature.

[0138] The decoupling module 24 is configured to decouple the mixed service characteristics using an independent component analysis algorithm to obtain a traffic density index and a bandwidth index.

[0139] The construction module 25 is configured to dynamically associate the traffic density indicator with the bandwidth indicator to construct a service-independent feature matrix.

[0140] The second generating module 26 is configured to convert the service-independent feature matrix into a service feature analysis diagram based on a stream processing engine, and generate a decision strategy corresponding to the service feature analysis diagram in real time.

[0141] Figure 2 The business analysis system based on big data can be executed Figure 1 The implementation principles and technical effects of the big data-based business analysis method described in the illustrated embodiment are not further elaborated. The specific manner in which each module and unit performs operations in the big data-based business analysis system in the above embodiment has been described in detail in the relevant embodiments of the method and will not be elaborated on here.

[0142] In one possible design, Figure 2 A business analysis system based on big data in the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32; The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0143] The processing component 32 is used to perform the above Figure 1 The embodiment provides a business analysis method based on big data.

[0144] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above method.

[0145] The storage component 31 is configured to store various types of data to support operations on the terminal. The storage component can be implemented by any type of volatile or non-volatile memory device, or a combination thereof, such as static random-access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0146] Of course, a computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0147] The input / output interface provides an interface between the processing component and the peripheral interface module, which can be an output device, an input device, etc.

[0148] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0149] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0150] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The embodiment shown is a business analysis method based on big data.

[0151] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0152] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0153] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0154] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A business analysis method based on big data, characterized in that: include: Obtain user clickstream data, payment flow data, and device electromagnetic interference signals; Generate multi-source heterogeneous business time series data based on the user click stream data and the payment flow data; Converting the device electromagnetic interference signal into a device vibration frequency domain feature, and generating a collaborative data stream based on the device vibration frequency domain feature and the multi-source heterogeneous service time series data, wherein the collaborative data stream includes a mixed service feature; Decoupling the mixed service characteristics using an independent component analysis algorithm to obtain a traffic density index and a bandwidth index; Dynamically associating the traffic density indicator with the bandwidth indicator to construct a service-independent feature matrix; Based on a stream processing engine, the business independent feature matrix is ​​converted into a business feature analysis diagram, and a decision strategy corresponding to the business feature analysis diagram is generated in real time.

2. The method according to claim 1, characterized in that The use of the independent component analysis algorithm to decouple the mixed service characteristics to obtain a traffic density index and a bandwidth index includes: Inputting the separation factor into a pre-trained blind source separation model, and separating the traffic density signal source and the bandwidth signal source from the mixed service characteristics through the blind source separation model; extracting a first fluctuation trajectory corresponding to the number of requests per minute from the traffic density signal source, analyzing the first fluctuation trajectory, and obtaining a traffic density index; A second fluctuation trajectory corresponding to the payment completion time is extracted from the bandwidth signal source, and the second fluctuation trajectory is analyzed to obtain a bandwidth index.

3. The method according to claim 2, characterized in that Inputting the separation factor into a pre-trained blind source separation model, and separating the traffic density signal source and the bandwidth signal source from the mixed service features by the blind source separation model, comprises: The traffic density weight and bandwidth weight in the separation factor are used as separation criteria and set in the decoupling layer of the blind source separation model; Performing orthogonal projection on the hybrid service characteristics through the decoupling layer to generate a feature vector space, wherein the feature vector space includes a traffic density feature component and a bandwidth feature component; determining a separating hyperplane according to a distribution difference between the traffic density characteristic component and the bandwidth characteristic component; Using the separating hyperplane, the feature vector space is divided into a first feature subspace and a second feature subspace; A flow density signal source is extracted from the first characteristic subspace, and a bandwidth signal source is extracted from the second characteristic subspace.

4. The method according to claim 3, characterized in that The method of using the separating hyperplane to divide the feature vector space into a first feature subspace and a second feature subspace includes: Calculating the projection distance of each data point in the feature vector space to the separating hyperplane along the normal vector direction of the separating hyperplane; Determining a boundary area according to the positive or negative nature of the projection distance; Based on the boundary region, dividing the feature vector space into two orthogonal subspaces; Determine the orthogonal subspace containing the traffic density characteristic component aggregation area in the two orthogonal subspaces as the first characteristic subspace; The orthogonal subspace containing the bandwidth characteristic component aggregation area in the two orthogonal subspaces is determined as the second characteristic subspace.

5. The method according to claim 1, wherein The generating of the collaborative data stream based on the device vibration frequency domain characteristics and the multi-source heterogeneous service time series data includes: Performing waveform superposition on the device vibration frequency domain characteristics and the multi-source heterogeneous service time series data according to sampling time points to generate superimposed waveform data; Re-encapsulating the superimposed waveform data into a service data packet according to the transmission protocol of the service data stream through a signal injection mechanism; Performing timestamp alignment on the physical layer feature timestamp and the service data timestamp in the service data packet; Based on the aligned business data packets, a collaborative data stream containing mixed business features is output.

6. The method according to claim 1, characterized in that The dynamically associating the traffic density indicator and the bandwidth indicator to construct a service-independent feature matrix includes: Establishing a sliding time window, and calculating the fluctuation coupling degree between the traffic density indicator and the bandwidth indicator within the sliding time window; weighting the traffic density index value and the bandwidth index value respectively according to the fluctuation coupling degree; The weighted traffic density index values ​​and the weighted bandwidth index values ​​are arranged in time series to generate a service-independent feature matrix.

7. The method according to claim 1, characterized in that The stream processing engine is based on converting the business independent feature matrix into a business feature analysis diagram, and generating a decision strategy corresponding to the business feature analysis diagram in real time, including: Creating a corresponding virtual data channel for the business-independent feature matrix in a stream processing engine; In the virtual data channel, decomposing the service-independent feature matrix into a plurality of data blocks, and mapping each data block to an analysis node; Calculating the association strength between the analysis nodes, and establishing a connection relationship between the analysis nodes based on the association strength; Based on the connection relationship, construct a business feature analysis diagram; According to the topological structure characteristics of the business characteristic analysis diagram, a corresponding decision strategy is generated.

8. A business analysis system based on big data, characterized in that: include: The acquisition module is used to obtain user click stream data, payment flow data and device electromagnetic interference signals; A first generating module is configured to generate multi-source heterogeneous business time series data based on the user clickstream data and the payment transaction data; a conversion module, configured to convert the device electromagnetic interference signal into a device vibration frequency domain feature, and generate a collaborative data stream based on the device vibration frequency domain feature and the multi-source heterogeneous service time series data, wherein the collaborative data stream includes a mixed service feature; A decoupling module, configured to decouple the mixed service characteristics using an independent component analysis algorithm to obtain a traffic density index and a bandwidth index; A construction module, configured to dynamically associate the traffic density indicator with the bandwidth indicator to construct a service-independent feature matrix; The second generating module is used to convert the business independent feature matrix into a business feature analysis diagram based on a streaming processing engine, and generate a decision strategy corresponding to the business feature analysis diagram in real time.

9. A computing device, characterized in that It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a business analysis method based on big data as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that A computer program is stored, and when the computer program is executed by a computer, a business analysis method based on big data as described in any one of claims 1 to 7 is implemented.

Images