Forestry informatization platform construction method based on business process modeling

Abstract

The invention relates to the technical field of business process management, in particular to a forestry informatization platform construction method based on business process modeling. The method comprises the following steps of obtaining forestry business process data, performing integrity filling and feature extraction, and generating business process feature data; on the basis, performing process mode mining and systematic modeling to generate an initial business process model; performing sub-domain cutting on the model based on the flow demand data to generate a resource management sub-domain and a production management sub-domain, and performing correlation analysis and structure construction on the resource management sub-domain and the production management sub-domain; the method comprises the following steps: constructing a progressive relationship of a sub-domain association network through actual forestry service data; and according to the progressive relationship, carrying out branch management processing on the initial model, and establishing a process informatization management platform. According to the invention, more accurate, more scientific and more efficient forestry informatization platform construction can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of business process management, and in particular to a method for building a forestry information platform based on business process modeling. Background Art

[0002] In the process of forestry informatization, the complexity and diversity of business processes have gradually emerged. Traditional forestry management methods usually rely on a single system or manual operation, which is difficult to cope with complex business process requirements. With the expansion of forestry management business scope and the increase of diversified needs, how to efficiently integrate and manage these business process data has become an urgent problem to be solved. The existing forestry informatization system lacks the ability to fully process business process data, which makes it difficult to ensure the integrity and accuracy of the data, thus affecting the scientificity and effectiveness of management decisions. In traditional forestry management systems, the modeling and systematization of business processes are often based on static or single perspectives. The result is that it is difficult to flexibly respond to changes in business needs and cannot fully reflect the complex correlations between different business process subdomains. Especially in the two highly related subdomains of resource management and production management, the lack of effective correlation analysis and progressive evolution processing has led to inefficient business processes and delayed decision-making. Summary of the invention

[0003] Based on this, it is necessary to provide a method for building a forestry information platform based on business process modeling to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a forestry information platform construction method based on business process modeling includes the following steps: Step S1: Acquire forestry business process data; perform integrity filling processing on the forestry business process data, and perform feature extraction to obtain business process feature data; Step S2: mining the business process feature data for process patterns, and systematically modeling the process to generate an initial business process model; Step S3: Perform process requirement analysis on the business process feature data to obtain process requirement data; perform subdomain segmentation on the initial business process model based on the process requirement data to generate a resource management subdomain and a production management subdomain; Step S4: performing correlation analysis on the resource management subdomain and the production management subdomain to obtain subdomain correlation data; constructing a correlation structure for the resource management subdomain and the production management subdomain according to the subdomain correlation data to generate a subdomain correlation network; Step S5: Acquire forestry actual business data; construct a progressive relationship for the subdomain association network based on the forestry actual business data to obtain a subdomain progressive relationship; Step S6: Perform process division processing on the initial business process model according to the sub-domain progressive relationship to obtain the divided business process data; build a management platform for the divided business process data to generate a process information management platform.

[0005] The present invention can effectively process incomplete or missing data through integrity filling processing, thereby improving the integrity and accuracy of business process data. Feature extraction technology can identify and extract key features in business processes, ensuring the effectiveness and accuracy of subsequent analysis and modeling. By mining the process pattern of business process feature data, the system can identify common business process patterns and perform systematic modeling, and the generated initial business process model has strong representativeness and adaptability. This process lays a solid foundation for subsequent subdomain cutting and process optimization, and improves the reliability and accuracy of the model. By analyzing the correlation between the resource management subdomain and the production management subdomain, subdomain association data is obtained, and a subdomain association network is constructed based on this. This structure not only reveals the inherent relationship between the subdomains, but also provides theoretical support for the subsequent progressive relationship construction, ensuring the logic and scientificity of platform construction. By constructing the progressive relationship of the subdomain association network based on actual business data, the order and hierarchical relationship of each subdomain in the business process are clarified. This technical means ensures the orderliness and hierarchy of business process management, which helps to improve the operating efficiency and user experience of the platform. By dividing the initial business process model, complex business processes are divided and managed by subdomains, which improves the management efficiency of the process. At the same time, the information management platform built based on the divided business process data can realize the all-round digital management of business processes and improve the response speed and management efficiency of the overall system. The platform integrates the business data of each subdomain to ensure the consistency of information and the efficiency of management. This makes the entire forestry information platform more accurate, scientific and efficient, so as to better serve the management and optimization of forestry business processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0006] Figure 1 It is a schematic diagram of the steps of a forestry information platform construction method based on business process modeling; Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart; Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG. The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0007] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are 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 technicians in this field without creative work are within the scope of protection of the present invention.

[0008] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0009] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0010] To achieve this, please refer to Figures 1 to 3 , a forestry information platform construction method based on business process modeling, including the following steps: Step S1: Acquire forestry business process data; perform integrity filling processing on the forestry business process data, and perform feature extraction to obtain business process feature data; Step S2: mining the business process feature data for process patterns, and systematically modeling the process to generate an initial business process model; Step S3: Perform process requirement analysis on the business process feature data to obtain process requirement data; perform subdomain segmentation on the initial business process model based on the process requirement data to generate a resource management subdomain and a production management subdomain; Step S4: performing correlation analysis on the resource management subdomain and the production management subdomain to obtain subdomain correlation data; constructing a correlation structure for the resource management subdomain and the production management subdomain according to the subdomain correlation data to generate a subdomain correlation network; Step S5: Acquire forestry actual business data; construct a progressive relationship for the subdomain association network based on the forestry actual business data to obtain a subdomain progressive relationship; Step S6: Perform process division processing on the initial business process model according to the sub-domain progressive relationship to obtain the divided business process data; build a management platform for the divided business process data to generate a process information management platform.

[0011] The present invention can effectively process incomplete or missing data through integrity filling processing, thereby improving the integrity and accuracy of business process data. Feature extraction technology can identify and extract key features in business processes, ensuring the effectiveness and accuracy of subsequent analysis and modeling. By mining the process pattern of business process feature data, the system can identify common business process patterns and perform systematic modeling, and the generated initial business process model has strong representativeness and adaptability. This process lays a solid foundation for subsequent subdomain cutting and process optimization, and improves the reliability and accuracy of the model. By analyzing the correlation between the resource management subdomain and the production management subdomain, subdomain association data is obtained, and a subdomain association network is constructed based on this. This structure not only reveals the inherent relationship between the subdomains, but also provides theoretical support for the subsequent progressive relationship construction, ensuring the logic and scientificity of platform construction. By constructing the progressive relationship of the subdomain association network based on actual business data, the order and hierarchical relationship of each subdomain in the business process are clarified. This technical means ensures the orderliness and hierarchy of business process management, which helps to improve the operating efficiency and user experience of the platform. By dividing the initial business process model, complex business processes are divided and managed by subdomains, which improves the management efficiency of the process. At the same time, the information management platform built based on the divided business process data can realize the all-round digital management of business processes and improve the response speed and management efficiency of the overall system. The platform integrates the business data of each subdomain to ensure the consistency of information and the efficiency of management. This makes the entire forestry information platform more accurate, scientific and efficient, so as to better serve the management and optimization of forestry business processes.

[0012] In the embodiment of the present invention, reference Figure 1 The above is a schematic diagram of the steps of a method for building a forestry information platform based on business process modeling of the present invention. In this example, the method for building a forestry information platform based on business process modeling includes the following steps: Step S1: Acquire forestry business process data; perform integrity filling processing on the forestry business process data, and perform feature extraction to obtain business process feature data; In this embodiment, forestry business process data is fully acquired through the data acquisition module deployed in the forestry management system. The data includes but is not limited to detailed records of multiple links such as forest resource monitoring, felling management, vegetation protection, and ecological restoration. In order to ensure the integrity of the data, a data cleaning and filling algorithm is used to interpolate and fill missing values ​​to ensure the integrity of each process data set. Next, a feature extraction tool is used to extract features from the integrity-processed business process data. The feature extraction tool extracts key features in the business process based on classification algorithms and association rule mining technology, and finally obtains business process feature data. These feature data include workflow patterns, time distribution, resource consumption, etc. of key nodes, providing a basis for subsequent process modeling and analysis.

[0013] Step S2: mining the business process feature data for process patterns, and systematically modeling the process to generate an initial business process model; In this embodiment, after obtaining the business process characteristic data, a process mining tool is used to conduct an in-depth analysis of the characteristic data. The tool mines different business process patterns through the Petri net model and identifies common process patterns, such as sequential patterns, concurrent patterns, selection patterns, etc. In the mining process, the frequency and correlation of each pattern are quantitatively analyzed in combination with statistical methods to ensure that the mining results are highly accurate and representative. Subsequently, based on the mined process pattern, a business process modeling tool (such as a BPMN modeling tool) is used for systematic modeling. By integrating the sub-processes of different patterns, a comprehensive initial business process model is constructed with a clear model structure, covering all key links of forestry business.

[0014] Step S3: Perform process requirement analysis on the business process feature data to obtain process requirement data; perform subdomain segmentation on the initial business process model based on the process requirement data to generate a resource management subdomain and a production management subdomain; In this embodiment, a process demand analysis is performed on the business process feature data, and a demand analysis tool is used to match and compare the process characteristics with the actual business needs to identify the demand points and improvement points of the process. The result of the analysis generates process demand data, which specifically includes the bottleneck nodes of the process, the links with uneven resource allocation, and the business flow paths that need to be optimized. Based on the process demand data, the initial business process model is sub-domain cut. Using the sub-domain partitioning algorithm, the entire business process is divided into a resource management sub-domain and a production management sub-domain. The division of these two sub-domains is based on the core functions of the business process. The resource management sub-domain mainly includes resource allocation and monitoring, and the production management sub-domain covers the actual forestry production operations and execution.

[0015] Step S4: performing correlation analysis on the resource management subdomain and the production management subdomain to obtain subdomain correlation data; constructing a correlation structure for the resource management subdomain and the production management subdomain according to the subdomain correlation data to generate a subdomain correlation network; In this embodiment, when performing correlation analysis on the resource management subdomain and the production management subdomain, a correlation calculation algorithm is used. The algorithm obtains subdomain correlation data by calculating the node correlation, data flow frequency, and resource sharing between subdomains. The correlation data specifically reflects the dependency and degree of interaction between the two subdomains. Based on these data, a network analysis tool is used to construct a correlation structure between the resource management subdomain and the production management subdomain. By drawing a correlation network diagram, the interaction relationship and information flow path between subdomains are displayed, and finally a subdomain correlation network is generated. This network clearly expresses the complex correlation structure between subdomains, providing a basis for the subsequent progressive relationship construction.

[0016] Step S5: Acquire forestry actual business data; construct a progressive relationship for the subdomain association network based on the forestry actual business data to obtain a subdomain progressive relationship; In this embodiment, after obtaining the authority, the actual business data of forestry is obtained by querying the actual business, and the data is cleaned and standardized to ensure data quality. Then, based on these processed business data, the progressive relationship of the subdomain association network generated in the early stage is constructed. The progressive relationship construction uses a multi-level progressive analysis method to compare the actual business data with the nodes and edges in the subdomain association network one by one, and analyze how the business data affects the progressive relationship between subdomains. Finally, the subdomain progressive relationship data is obtained, which contains the priority of each subdomain in the process and the specific path of progression, and clarifies the advancement order and interaction between the resource management subdomain and the production management subdomain in the overall business process.

[0017] Step S6: Perform process division processing on the initial business process model according to the sub-domain progressive relationship to obtain the divided business process data; build a management platform for the divided business process data to generate a process information management platform.

[0018] In this embodiment, the initial business process model is divided into different processes according to the progressive relationship of subdomains. Using the process optimization tool, each step in the business process is reconfigured and divided according to the progressive relationship of subdomains. The result of the division of responsibilities is to classify different business process links into the resource management subdomain and the production management subdomain, and adjust the process sequence and dependencies to achieve more efficient process management. Next, the divided business process data is integrated, and the management platform is built on this basis. The platform development uses a service-oriented architecture (SOA) design to modularize the process management function, including functional modules such as process monitoring, resource allocation, and data analysis. Generate an information management platform covering the entire process of forestry business. The platform supports real-time data processing and decision support, and realizes comprehensive information management of forestry business processes.

[0019] Preferably, step S1 comprises: Step S11: Acquire forestry business process data; Step S12: Perform outlier detection on the forestry business process data to obtain business process outliers; Step S13: Perform integrity filling processing on the forestry business process data based on the business process abnormal values ​​to generate complete business process data; Step S14: Perform text vectorization processing on the complete business process data to generate a high-dimensional feature vector of the process; Step S15: Perform nonlinear principal component analysis on the high-dimensional feature vector of the process to obtain business process feature data.

[0020] The present invention ensures the integrity and accuracy of business process data through outlier detection and completeness filling processing, reduces the impact of data noise and missing on subsequent analysis, and improves the reliability of the platform. Text vectorization and nonlinear principal component analysis of complete business process data can efficiently extract the key features of the business process, generate representative high-dimensional feature vectors, and ensure the accuracy of the model. Through nonlinear principal component analysis, complex nonlinear relationships in business process data can be captured, providing richer and more accurate feature data for subsequent process modeling, and enhancing the intelligent analysis capabilities of the platform. Through precise data processing and feature extraction, the platform can adapt to diverse forestry business processes, has a high degree of adaptability and scalability, and provides strong support for complex forestry business management. It improves the accuracy, reliability and applicability of the forestry information platform in business process modeling and management.

[0021] In this embodiment, historical and real-time data related to forestry business processes are obtained through a data collection module deployed in a forestry information management system. The data set includes but is not limited to production operation plans, resource allocation records, equipment operation logs, personnel management records, etc. The data collection module automatically extracts relevant data from the business process management system (BPM), resource management system (ERP), and ground inspection system (SCADA), and imports these data into the central database through a data interface. The Z-score method is used to standardize each indicator in the data so that its mean is 0 and its variance is 1. Then, a density-based outlier detection algorithm (such as DBSCAN) is applied to the standardized data to analyze the outliers in the data distribution. During the detection process, by setting the minimum number of samples and the distance threshold, the abnormal data points in the business process are accurately located, and finally an abnormal value data set containing abnormal data points and corresponding indicators is generated. Using an interpolation method based on Bayesian inference, the abnormal data points are interpolated multiple times to generate multiple sets of alternative values. Then, the final alternative value is determined by a weighted average or voting mechanism and filled into the original data set. The filled data set is verified by the integrity check tool to ensure the logical consistency of the data, and the processing results are saved as complete business process data, laying the foundation for subsequent feature extraction and analysis. For text-type data, the Bag of Words model or the Word Embedding model is used to convert the text into a vector representation. For numerical data, it is directly converted into a vector form after standardization. Then, the TF-IDF (term frequency-inverse document frequency) method is used to weight the text vector to generate a high-dimensional feature vector with semantic information. Finally, all business process data are mapped into a unified high-dimensional space to form a high-dimensional feature vector that can be used for further analysis. A kernel function (such as a Gaussian kernel function) is constructed to capture the nonlinear relationship in the feature vector, and the feature vector is mapped to the new feature space by calculating the kernel matrix. Then, the gradient descent-based optimization algorithm is used to perform principal component analysis on the mapped data, extract the main feature dimensions, and filter out low-correlation or redundant features. Finally, feature data sets with high explanatory power are retained, which will be used to build a more accurate forestry business process model and subsequent information platform design.

[0022] Preferably, step S2 comprises: Step S21: performing repeated process mining on the business process feature data to obtain frequent process fragments; Step S22: performing process pattern recognition on frequent process fragments to generate a basic business process pattern; Step S23: Performing module structure analysis on the basic model of the business process to obtain a sub-process model; Step S24: Integrate the sub-process models as a whole to generate an initial business process model.

[0023] The present invention can identify frequently occurring process fragments through repeated process mining of business process feature data, help streamline and optimize business processes, remove redundant operations, and improve the efficiency of business processes. By performing process pattern recognition on frequent process fragments and generating the basic model of the business process, the standardization and consistency of the business process are ensured, which helps to improve the execution quality and management standardization of the business process. By performing modular structured analysis on the basic model of the business process, complex business processes can be decomposed into manageable sub-process models, which improves the flexibility and operability of process design and facilitates modular adjustment and optimization for different business scenarios. The sub-process models are integrated as a whole to generate the initial business process model, which realizes the organic integration of each sub-process, ensures the consistency and systematicness of the business process, and helps to improve the comprehensive management capabilities and business collaboration efficiency of the platform. The platform is ensured to be efficient and adaptable in process design, optimization, management and integration, and provides a solid technical foundation for forestry information management.

[0024] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes: Step S21: performing repeated process mining on the business process feature data to obtain frequent process fragments; In this embodiment, the principal component analysis (PCA) technique is applied to perform multidimensional feature decomposition. This technique maps high-dimensional data to low-dimensional space through linear transformation to retain the main variability of the data. During the processing, the business process feature data is first standardized so that each feature has the same mean and variance. Subsequently, the covariance matrix of the feature data is calculated, and the eigenvalue decomposition is performed on it to extract the principal component. The first n principal components are selected according to the contribution rate of the principal component, and the original feature data is projected onto these components to generate a feature vector matrix. Each row of the feature vector matrix represents a low-dimensional feature representation of a business process instance, providing a data basis for subsequent analysis. Based on the feature vector matrix, DBSCAN (density-based spatial clustering algorithm) is used for density clustering analysis. The DBSCAN algorithm identifies high-density and low-density areas in the data by setting the neighborhood radius (ε) and the minimum number of samples (MinPts). The algorithm first calculates the distance between each pair of samples in the eigenvector matrix, and classifies the samples whose distance is within the neighborhood radius into the same cluster. For each density-connected sample group, a potential process pattern is formed. The output of density clustering includes the center, range and feature vector set of each cluster. The feature vector of each potential process pattern is compared with the feature vector in the actual business process data to determine the degree of match. Then, all matching results are summarized and the number of occurrences of each process pattern is counted. In the frequency analysis process, data normalization technology needs to be applied to ensure the consistency of frequency comparisons between different process patterns. Finally, the repetition frequency data of each potential process pattern is obtained as the basis for judging the importance and stability of the process. According to the preset repetition frequency threshold (such as the percentage of occurrences to the total number of times), the process patterns are screened. Those process patterns that appear frequently and have a repetition frequency higher than the threshold are marked as frequent process fragments. Experiments are conducted by setting different thresholds to ensure the recognition accuracy and business relevance of frequent process fragments. Finally, a set of process fragments with high repetition frequency is obtained. These frequent process fragments represent the key parts that are important and stable in the business process, providing a basis for further process modeling and optimization.

[0025] Step S22: performing process pattern recognition on frequent process fragments to generate a basic business process pattern; In this embodiment, frequent process fragments are mapped to the basic structure of the Petri net model, including elements such as nodes, transitions, and arcs. Then, through the process pattern recognition algorithm, the commonalities and differences between different frequent fragments are analyzed to identify the process structure with significant pattern characteristics. The recognition process extracts the basic pattern representing the entire business process by comparing the node connection mode, event sequence, and resource usage in the fragment. Finally, the basic business process pattern is output in the form of a Petri net model, which includes key process paths and control structures, providing a basis for subsequent modular analysis.

[0026] Step S23: Performing module structure analysis on the basic model of the business process to obtain a sub-process model; In this embodiment, each key path or functional block in the basic model is regarded as an independent module, and a detailed analysis is performed on its functions, input and output, resource usage, etc. Then, by applying object-oriented modeling methods, these modules are structured into sub-process models to ensure that each sub-process has clear functional boundaries and interface definitions. The sub-process model ensures that it can run independently and has high reusability in different business scenarios through process decomposition and reorganization. Finally, the sub-process model is output in a modular form and lays the foundation for overall process integration.

[0027] Step S24: Integrate the sub-process models as a whole to generate an initial business process model.

[0028] In this embodiment, by defining the service interface and message passing mechanism, the input and output relationships between different sub-process models are clearly mapped to ensure seamless data transmission between modules. Then, BPMN (Business Process Model and Notation) is used to model each sub-process model as a whole to construct a panoramic view of the initial business process model. During the model integration process, the business rule engine is introduced to ensure that each process module is executed in accordance with the predefined business logic sequence to ensure the coherence and consistency of the entire business process. Finally, the initial business process model is generated, and the consistency and integrity are checked through the model verification tool, which provides a solid foundation for the construction of the forestry information platform.

[0029] Preferably, step S21 includes: Step S211: performing multi-dimensional feature decomposition on the business process feature data to generate a feature vector matrix; Step S212: performing density cluster analysis on the feature vector matrix to generate a potential process pattern; Step S213: Performing pattern frequency analysis on the potential process pattern to obtain the process repetition frequency; Step S214: Based on the process repetition frequency and a preset repetition frequency threshold, the business process feature data is mined for highly repetitive processes to generate frequent process fragments.

[0030] The present invention can conduct a comprehensive and detailed analysis of business process features by multi-dimensional feature decomposition and generating feature vector matrices. This method can effectively extract key features in business processes and improve the accuracy and comprehensiveness of feature extraction. Using density cluster analysis to identify potential process patterns can reveal potential structures and relationships in business processes, help discover business process patterns that are implicit in data and not explicitly identified, and improve the depth and breadth of pattern recognition. By analyzing the frequency of potential process patterns, the repetition frequency of the process is obtained, and it can be clear which process fragments appear frequently. This helps to identify and optimize common operations in business and improve the standardization and consistency of business processes. Based on the preset repetition frequency threshold, highly repeated process mining can effectively screen out frequently occurring process fragments. This not only improves the mining efficiency of process data, but also ensures that the mined process fragments have high practical application value, and supports the optimization and improvement of business processes. The platform's capabilities in business process feature extraction, pattern recognition, frequency analysis and process optimization are enhanced, providing scientific and efficient technical support for forestry information management.

[0031] In this embodiment, the principal component analysis (PCA) technique is applied to perform multidimensional feature decomposition. This technique maps high-dimensional data to low-dimensional space through linear transformation to retain the main variability of the data. During the processing, the business process feature data is first standardized so that each feature has the same mean and variance. Subsequently, the covariance matrix of the feature data is calculated, and the eigenvalue decomposition is performed on it to extract the principal component. The first n principal components are selected according to the contribution rate of the principal component, and the original feature data is projected onto these components to generate a feature vector matrix. Each row of the feature vector matrix represents a low-dimensional feature representation of a business process instance, providing a data basis for subsequent analysis. Based on the feature vector matrix, DBSCAN (density-based spatial clustering algorithm) is used for density clustering analysis. The DBSCAN algorithm identifies high-density and low-density areas in the data by setting the neighborhood radius (ε) and the minimum number of samples (MinPts). The algorithm first calculates the distance between each pair of samples in the eigenvector matrix, and classifies the samples whose distance is within the neighborhood radius into the same cluster. For each density-connected sample group, a potential process pattern is formed. The output of density clustering includes the center, range and feature vector set of each cluster. The feature vector of each potential process pattern is compared with the feature vector in the actual business process data to determine the degree of match. Then, all matching results are summarized and the number of occurrences of each process pattern is counted. In the frequency analysis process, data normalization technology needs to be applied to ensure the consistency of frequency comparisons between different process patterns. Finally, the repetition frequency data of each potential process pattern is obtained as the basis for judging the importance and stability of the process. According to the preset repetition frequency threshold (such as the percentage of occurrences to the total number of times), the process patterns are screened. Those process patterns that appear frequently and have a repetition frequency higher than the threshold are marked as frequent process fragments. Experiments are conducted by setting different thresholds to ensure the recognition accuracy and business relevance of frequent process fragments. Finally, a set of process fragments with high repetition frequency is obtained. These frequent process fragments represent the key parts that are important and stable in the business process, providing a basis for further process modeling and optimization.

[0032] Preferably, step S3 comprises: Step S31: performing semantic clustering processing on the business process feature data to obtain a process requirement cluster; Step S32: fine-grained classification of process requirement clusters to generate process requirement data; Step S33: performing demand matching on the initial business process model based on the process demand data to obtain demand matching data; Step S34: The initial business process model is divided into sub-domains according to the demand matching data to generate a resource management sub-domain and a production management sub-domain.

[0033] The present invention can effectively summarize business process requirements into different process requirement clusters by semantic clustering of business process feature data. This clustering method can capture potential demand patterns in business processes and improve the accuracy and depth of demand analysis. Fine-grained classification of process requirement clusters can make more accurate divisions of business process requirements. This helps to identify more detailed demand features, thereby improving the adaptability and flexibility of business process models. Demand matching of the initial business process model based on the generated process requirement data can ensure the consistency between the business process model and the actual demand. This matching process improves the accuracy and practicality of the business process model and better supports the realization of business goals. Sub-domain cutting of the initial business process model according to the demand matching data can effectively generate resource management sub-domains and production management sub-domains. This sub-domain division method can better organize and manage business processes and improve the modularity and maintainability of the information platform. It can accurately identify and meet business process requirements, thereby improving the overall performance of the forestry information platform and the optimization effect of the business process.

[0034] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes: Step S31: performing semantic clustering processing on the business process feature data to obtain a process requirement cluster; In this embodiment, the business process feature data is converted into word embedding vectors, for example, using the Word2Vec or GloVe model to convert the keywords and phrases in the business process into high-dimensional vector representations. Next, the K-means algorithm is applied to cluster these embedded vectors, where the selection of the K value is based on the elbow rule or the silhouette coefficient to determine the optimal number of clusters. During the clustering process, the distance between each feature vector and the cluster center is calculated, and the feature vector is assigned to the cluster closest to it. Ultimately, several semantically similar process requirement clusters are generated, each cluster containing feature data with similar business requirements, providing a basis for subsequent fine-grained classification.

[0035] Step S32: fine-grained classification of process requirement clusters to generate process requirement data; In this embodiment, the business process features in each cluster are extracted and a feature matrix is ​​constructed. Then, the feature data are trained using a decision tree classification algorithm to construct a decision tree model, which divides the data into more detailed categories based on different attributes of the feature data (such as business functions, execution frequency, etc.). Each category represents a specific process requirement. During the decision tree training process, the accuracy of the model is evaluated using cross-validation technology, and the decision rules are adjusted to optimize the classification results. Finally, the business requirements in each cluster are subdivided into specific process requirement data to provide clear classification information for further demand matching.

[0036] Step S33: performing demand matching on the initial business process model based on the process demand data to obtain demand matching data; In this embodiment, the demand data is compared with the process elements in the initial business process model. By defining a matching algorithm (for example, a matching algorithm based on similarity), the similarity between the process demand data and each element in the process model is calculated. The similarity calculation can be based on the cosine similarity or Euclidean distance between feature vectors. The matching results are summarized to record the degree of match between each process demand and the corresponding element in the business process model. Finally, the demand matching data is obtained, which includes the correspondence between each process demand and the process elements in the initial business process model and their matching degree, providing data support for subsequent sub-domain cutting.

[0037] Step S34: The initial business process model is divided into sub-domains according to the demand matching data to generate a resource management sub-domain and a production management sub-domain.

[0038] In this embodiment, the boundaries and functional allocation of each process module are determined according to the matching degree between the process requirements in the demand matching data and the process model elements. A demand-based cutting algorithm is used to divide the relevant process elements in the initial business process model into different subdomains. The specific operations include: identifying demand points with high matching degrees in the demand data, defining the cutting rules of the subdomains, and cutting in the business process model to generate resource management subdomains and production management subdomains. The boundaries and functions of each subdomain are adjusted according to the demand matching results to ensure the functional integrity and consistency of the subdomains. Finally, two subdomains are obtained - the resource management subdomain and the production management subdomain, which represent the resource management part and the production management part in the business process, respectively, laying the foundation for the subsequent development of the information management platform.

[0039] Preferably, step S33 includes: Step S331: Perform heterogeneous embedding processing on the process demand data and the initial business process model to obtain a multi-dimensional embedding feature vector; Step S332: Perform fuzzy correlation projection on the multi-dimensional embedded feature vector to obtain a demand-process fuzzy mapping; Step S333: performing self-organizing feature mapping on the demand-process fuzzy mapping to obtain a topology preserving projection; Step S334: construct a multi-layer complex network based on topology-preserving projection to obtain demand matching data.

[0040] The present invention obtains a multi-dimensional embedded feature vector by performing heterogeneous embedding processing on process demand data and an initial business process model. This processing method can convert business demand and process model into vectors with high-dimensional features, accurately capture the subtle differences between demand and process, and provide an accurate data basis for subsequent matching. Fuzzy correlation projection is performed on the multi-dimensional embedded feature vector to obtain a demand-process fuzzy mapping. The fuzzy correlation projection technology can handle the uncertainty and ambiguity between demand and process, thereby generating a more robust mapping result and enhancing the matching ability between demand and process. By performing self-organizing feature mapping on the demand-process fuzzy mapping, a topology-preserving projection is obtained. This method can retain the topological structure of the data, ensure that the relative relationship between demand and process is maintained during the process of data dimensionality reduction, and provide a more realistic and reliable representation of the business process. A multi-layer complex network is constructed based on topology-preserving projection to obtain demand matching data. This complex network construction method can reveal the complex relationship and dependency structure between demand and process, support in-depth analysis and optimization of business processes, and provide a solid data foundation for the decision support system of the information platform. It can match needs and business processes more accurately and effectively, thereby improving the intelligence level of the forestry information platform and the optimization effect of business processes.

[0041] In this embodiment, heterogeneous embedding processing is performed on the process demand data and the initial business process model, and the feature data of the process demand data and the initial business process model are converted into a unified multi-dimensional embedding feature vector. For the process demand data, the word embedding technology (such as Word2Vec or BERT model) is used to embed the demand description and convert it into a high-dimensional vector. For the initial business process model, the nodes and edges in the process model are converted into embedded features using a graph embedding algorithm (such as Node2Vec or GraphSAGE). The specific operation includes: text preprocessing (such as word segmentation, removal of stop words) of the demand data, and using a pre-trained word embedding model to generate a demand feature vector. For the business process model, an adjacency matrix of the graph is constructed, and the feature vector of the node is calculated by applying the graph embedding algorithm. Finally, the feature vectors in the demand data and the process model are standardized and normalized to ensure that the embedded features are compared on the same scale to obtain a multi-dimensional embedded feature vector. The fuzzy correlation between the demand feature vector and the business process model feature vector is calculated, and the similarity between the two is quantified using a fuzzy logic function (such as a membership function). Then, a fuzzy mapping algorithm (such as fuzzy C-means clustering) is applied to associate and project the demand feature vector and the process feature vector, and each demand feature vector is mapped to the corresponding process feature vector area. The fuzzy mapping result shows the degree of association and uncertainty of each demand with the corresponding part in the process model. Finally, a demand-process fuzzy mapping is generated, which contains the fuzzy relationship between the demand feature vector and the process feature vector, providing basic data for subsequent demand matching. The neurons in the SOM network are initialized and trained according to the fuzzy mapping data. During the training process, the learning rate and neighborhood function are gradually adjusted to enable the neurons to form a mapping of the low-dimensional grid in the high-dimensional embedding space. At each iteration, the demand-process fuzzy mapping data is input into the SOM network, and the weights of the neurons are updated to minimize the distance between the input data and the neurons. The topology-preserving projection obtained through training shows the structural relationship between the demand characteristics and the business process model characteristics, ensuring that the topology and similarity of the original data are retained during the mapping process. Finally, the generated topology-preserving projection diagram can accurately reflect the relationship between the demand and the process, and provides basic data for subsequent network construction. According to the node and connection relationship in the topology-preserving projection, the hierarchical structure of the complex network is defined. Each layer represents a different network dimension or data attribute, and the connection between nodes represents the relationship between demand and process characteristics. By setting the weights and connection rules of the network layer, the nodes in the topology-preserving projection are mapped to different levels of the network, and the connection relationship between layers is established. Use complex network analysis methods (such as network centrality analysis and community discovery) to analyze the network and identify key nodes and important relationships. Ultimately, the generated demand matching data contains the relationship between nodes in the network and the network structure characteristics, providing a clear view and basis for subsequent process optimization and demand docking.

[0042] Preferably, step S4 comprises: Step S41: extracting features from the data of the resource management subdomain and the production management subdomain to obtain a subdomain feature vector matrix; Step S42: performing correlation analysis on the resource management subdomain and the production management subdomain according to the subdomain feature vector matrix to generate subdomain correlation data; Step S43: Structural processing is performed on the resource management subdomain and the production management subdomain based on the subdomain association data to obtain a subdomain association network.

[0043] The present invention generates a subdomain feature vector matrix by extracting features from the data of the resource management subdomain and the production management subdomain. This process can fully capture the core features and attributes of each subdomain, and provide a deep understanding and accurate description of each subdomain in the business process. The subdomain feature vector matrix is ​​used to perform correlation analysis to generate subdomain correlation data. This analysis method can reveal the actual relationship and degree of dependence between the resource management subdomain and the production management subdomain, identify their interaction mode in the business process, and optimize resource allocation and production scheduling strategies. Based on the subdomain correlation data, the resource management subdomain and the production management subdomain are structured to generate a subdomain correlation network. This structured processing displays the complex relationship between the subdomains in the form of a network structure, making the dependency relationship between the subdomains more intuitive and systematic, and providing clear visual and data support for decision support and business optimization. By constructing a subdomain correlation network, the optimization and coordination of the resource management and production management subdomains in the business process can be achieved. This networked structure helps identify potential bottlenecks and optimization points, improves the efficiency and effectiveness of the business process, and enables the information platform to adjust and manage the operations of each subdomain more intelligently.

[0044] In this embodiment, relevant features are extracted from the data set of each subdomain, including but not limited to operation frequency, resource consumption, task completion time, etc. For the resource management subdomain, the extracted features include resource allocation efficiency, resource utilization, etc.; for the production management subdomain, the extracted features include production efficiency, equipment failure rate, etc. The data is processed using feature extraction techniques (such as principal component analysis (PCA) or feature selection algorithms) to select the most representative features. Then, these features are converted into vector form, and the feature vectors of each subdomain are organized into a matrix, each row represents the data features of a subdomain, and each column represents a feature. Finally, the generated subdomain feature vector matrix will be used for subsequent correlation analysis and structured processing. The similarity between different subdomains in the feature vector matrix is ​​calculated, and the relationship between subdomains is evaluated using a distance measurement method (such as cosine similarity, Euclidean distance). For each pair of subdomains, the similarity values ​​between their feature vectors are calculated, which reflect the proximity of the subdomains in the feature space. The calculated similarity values ​​are summarized to form an association matrix, each element of which represents the strength of association between different subdomains. Next, the subdomain pairs with significant associations are screened out according to the set threshold, and subdomain association data is generated. Finally, the data provides quantitative analysis results for the relationship between subdomains, laying the foundation for subsequent structured processing. Based on the association strength in the subdomain association data, the nodes and edges in the network are constructed. Each subdomain corresponds to a node in the network, and the edges between nodes are defined according to the similarity values ​​in the association data. The weight of the edge is proportional to the similarity between subdomains. Next, network analysis techniques (such as the shortest path algorithm and community discovery algorithm in graph theory) are applied to analyze the subdomain association network to identify important subdomain nodes and key relationships between subdomains. The constructed subdomain association network visualizes the association structure between the resource management subdomain and the production management subdomain, providing a comprehensive understanding of the subdomain relationship. Finally, the generated subdomain association network provides a basis for further process optimization and resource allocation.

[0045] Preferably, step S5 comprises: Step S51: Acquire forestry actual business data; Step S52: Identify the dependency relationship of the subdomain association network to obtain subdomain dependency data; Step S53: performing dependency sequence analysis on forestry actual business data and subdomain dependency data to generate subdomain progressive sequence data; Step S54: constructing a progressive relationship for the subdomain association network based on the subdomain progressive sequence data to obtain a subdomain progressive relationship.

[0046] The present invention obtains actual forestry business data to ensure that the platform uses the most authentic and latest data during the construction process. The high quality and accuracy of the data provide a reliable basis for subsequent analysis and model construction, and improve the effectiveness and accuracy of the overall system. Subdomain dependency data is generated by identifying the dependency relationship of the subdomain association network. This process helps to understand the dependency and interaction relationship between the subdomains in detail, provides clear data support for further analysis and optimization, and ensures that the subdomain management in the business process is more scientific and accurate. Dependency sequence analysis is performed on the actual forestry business data and subdomain dependency data to generate subdomain progressive sequence data. This systematic analysis method can accurately identify the progressive relationship between subdomains, which helps to optimize the order and structure of the business process, thereby improving the coordination and efficiency of the overall business process. Based on the subdomain progressive sequence data, the subdomain association network is constructed with a progressive relationship to obtain a subdomain progressive relationship. This construction method can help to achieve a more reasonable arrangement of the business process, so that each subdomain is processed in an optimized order, thereby improving the fluency of the business process and the overall operation efficiency.

[0047] In this embodiment, data is extracted from the forestry business management system and the field collection equipment. These data include forestry production records, resource management logs, equipment operation status, meteorological data, etc. After data collection, the data is exported into a structured format (such as CSV, Excel or database table) using a data collection tool (such as a database query tool). Next, the collected data is preprocessed, including removing duplicate data, filling missing values, and data standardization. Data standardization converts data from different sources into a unified metric to ensure the consistency and accuracy of the data. Ultimately, the sorted and saved forestry actual business data will serve as the basis for subsequent analysis and modeling. Load the subdomain association network to identify the dependencies between nodes (i.e., subdomains) in the network. Dependency analysis techniques (such as causal analysis or network flow analysis) are used to evaluate the dependency strength and direction between subdomains. For each pair of subdomains, record their mutually dependent properties in the business process (such as resource flow, information transmission, etc.). The generated dependency will be represented by a dependency matrix or a chart, where each element in the matrix represents the dependency strength between subdomains. The dependency data provides detailed information about the relationship between subdomains for subsequent sequential analysis and progressive relationship construction. Analyze the time series and business process patterns in the actual business data of forestry, and identify the order and dependency between subdomains in actual operations. Use data analysis techniques (such as time series analysis or sequence mining) to extract the progressive relationship of each subdomain in the business process. Combined with the subdomain dependency data, identify and construct the logical order and dependency path between subdomains. The generated subdomain progressive sequence data will be presented in the form of a sorted list or tree structure, clarifying the advancement order and priority of the subdomains in the business process. This data provides a basis for the subsequent progressive relationship construction. Map the subdomain progressive sequence data to the subdomain association network to determine the progressive position and relationship of each subdomain. According to the progressive sequence data, adjust the node connection and edge weight in the subdomain association network to reflect the advancement order of the subdomain in the business process. Apply a progressive relationship construction algorithm (such as a topological sorting algorithm or a network rearrangement algorithm) to rearrange the subdomains in a progressive order and establish a network structure of progressive relationships. Finally, the generated subdomain progressive relationship network provides the position and relationship of each subdomain in the overall business process, which helps to optimize and manage business processes.

[0048] Preferably, step S53 includes: Perform tensor decomposition processing on actual forestry business data to obtain actual business feature tensors; Perform density peak clustering on subdomain dependency data to generate a subdomain dependency key point set; Perform algebraic connectivity analysis on the actual business feature tensor and subdomain dependency key point set to obtain the data dependency framework; Construct a directed acyclic graph of the data dependency framework to generate candidate progressive sequence data; The candidate progressive sequence data and the reverse multi-criteria strategy are reordered to obtain the subdomain progressive sequence data.

[0049] The present invention generates an actual business feature tensor by performing tensor decomposition processing on actual forestry business data. This method can capture the multi-dimensional information of business data, effectively reveal the complex features and potential patterns in the business process, and lay the foundation for further analysis and optimization. The subdomain dependency data is clustered by density peaks to generate a subdomain dependency key point set. This process identifies key dependency points through density clustering technology, which helps to accurately locate important relationships between subdomains, thereby better understanding and managing subdomain dependency structures. The actual business feature tensor and the subdomain dependency key point set are analyzed for algebraic connectivity to obtain a data dependency framework. This framework can clearly describe the dependency relationship between each data point and provide clear data support for subsequent progressive sequence construction. A directed acyclic graph is constructed for the data dependency framework to generate candidate progressive sequence data, and reordered through a retrograde multi-criteria strategy to obtain the final subdomain progressive sequence data. This method ensures that the processing order between subdomains is accurately optimized, effectively improving the efficiency and smoothness of the business process.

[0050] In this embodiment, the actual business data of forestry is subjected to tensor decomposition processing. The business data is converted from a structured table format to a tensor format to capture the multidimensional characteristics in the data, which is completed by using a tensor decomposition algorithm (such as a high-order singular value decomposition or CANDECOMP / PARAFAC decomposition). The specific steps include organizing the data into a three-dimensional tensor (for example, the dimensions can be time, subdomain, and feature), and then applying a tensor decomposition algorithm to decompose the tensor into a set of low-rank matrices. These matrices represent the potential structure of the data, and the actual business feature tensors generated will reveal the main patterns and relationships in the business data, providing a basis for subsequent analysis. A density peak clustering algorithm (such as DPCLustering) is applied to process the subdomain dependency data. The algorithm identifies density peaks in the data through density analysis and uses these peaks as key points. These key points represent important dependencies or high-density areas between subdomains. The generated subdomain dependency key point set provides a set of core dependencies between subdomains, helps simplify and summarize dependency data, and lays the foundation for subsequent algebraic connectivity analysis. The connectivity between the feature tensor and the key point set is evaluated using algebraic connectivity analysis techniques. This includes calculating the connectivity of each eigenvector in the feature tensor and comparing it with the data of the dependency key points. Algebraic connectivity analysis can be done by constructing an algebraic connectivity matrix, which contains the connectivity strengths between different eigenvectors and dependency points. The generated data dependency framework reveals the mutual relationship and dependency structure between the feature tensor and the set of dependency key points, providing a data basis for the subsequent construction of a directed acyclic graph. The nodes in the dependency framework (i.e., subdomains and features) are used as vertices of the graph, and the dependencies between the nodes are added to the graph as directed edges. The direction of the directed edge indicates the direction of the dependency, thereby ensuring that there are no loops in the graph. A graph theory algorithm (such as topological sorting) is used to construct the graph and generate candidate progressive order data. The graph shows the potential advancement order of each subdomain in the business process, reflects the dependencies and priorities between subdomains, and provides a basis for subsequent order reordering. The candidate progressive order data is evaluated and adjusted according to the criteria defined by the retrograde multi-criteria strategy (such as resource utilization, business priority, etc.). A sorting algorithm (such as a weighted sorting algorithm) is applied to score and sort the candidate orders according to each criterion. The retrograde strategy will take into account the rollbacks and adjustments in the business process and incorporate these situations into the sorting criteria. The generated subdomain progressive order data will be combined with actual business needs and resource conditions to provide the final progressive order, ensuring that the order of advancement of subdomains in the business process meets actual needs and priorities.

[0051] Preferably, step S6 comprises: Step S61: Perform subdomain business mapping on the initial business process model according to the subdomain progressive relationship to obtain the corresponding subdomain business link; Step S62: performing process division processing on the initial business process model based on the corresponding sub-domain business links to generate division business process data; Step S63: construct an independent business management framework for the business process data in charge, and obtain a business process management platform framework; Step S64: Integrate and map the business process data based on the business process management platform framework to generate a process information management platform.

[0052] The present invention can clarify the specific business links of each subdomain by mapping the initial business process model to subdomain business according to the progressive relationship of subdomains. This mapping ensures the clear positioning of each link in the business process, which helps to improve the transparency and operability of the business process. The initial business process model is processed by process division based on the corresponding subdomain business link to generate the business process data. This processing method helps to optimize the management and execution efficiency of the business process by decomposing complex business processes into more manageable parts. An independent business management framework is constructed for the business process data to obtain a business process management platform framework. This independent framework can provide a modular management structure, so that each business process part can be independently managed and optimized, thereby improving the flexibility and adaptability of overall management. Based on the business process management platform framework, the business process data is integrated and mapped to generate a process information management platform. Through integrated mapping, the scattered business data and processes can be effectively integrated to build a comprehensive and integrated information management platform to improve the utilization efficiency and management level of business data.

[0053] In this embodiment, the initial business process model and its subdomain progressive relationship data are extracted. The business links in the initial business process model are mapped according to the subdomain progressive relationship. The specific operation includes using the subdomain order in the progressive relationship data to map each business link to the corresponding subdomain business link. Use a mapping tool (such as a mapping matrix or an association diagram) to associate each business link with the subdomain to ensure that each link in the model is accurately mapped to the specific business function of the subdomain. This step will generate a business process model containing the corresponding subdomain business link, clarify the position and role of the business link in each subdomain, and provide basic data for subsequent process division processing. Group the business links in the business process model and divide the business areas according to the subdomain. The processing of each subdomain business link includes identifying and extracting all business processes related to the subdomain, and modularizing them. Use a process division tool (such as a process segmentation algorithm or a division tool set) to separate and manage the business process data according to the subdomain to generate the division business process data. Define the management requirements and operation specifications of each subdomain business link. Then, use framework design methods (such as modular design or functional architecture design) to build the business process data into a business management framework. Framework construction includes setting the management interface, data flow path and control logic of each business module. Integrate these modules through management framework design tools (such as architecture design software or framework modeling tools) to generate a systematic business process management platform framework. This framework will ensure that the business processes of each subdomain can run independently and coordinate with each other in the management platform. Apply integration mapping tools (such as integration modeling tools or information integration platforms) to map the business process data of each subdomain to the business process management platform framework. The integration process includes integrating business module data, operation specifications and data flow paths into a unified management platform to ensure that all business links can work together on an information platform. The integration mapping will generate a comprehensive process information management platform, which includes dynamic management and monitoring functions of business processes. Provide a centralized view to manage and optimize forestry business processes, improve management efficiency and business responsiveness.

[0054] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0055] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for building a forestry information platform based on business process modeling, characterized in that: The following steps are involved: Step S1: Acquire forestry business process data; perform integrity filling processing on the forestry business process data, and perform feature extraction to obtain business process feature data; Step S2: mining the business process feature data for process patterns, and systematically modeling the process to generate an initial business process model; Step S3: Perform process requirement analysis on the business process feature data to obtain process requirement data; perform subdomain segmentation on the initial business process model based on the process requirement data to generate a resource management subdomain and a production management subdomain; Step S4: performing correlation analysis on the resource management subdomain and the production management subdomain to obtain subdomain correlation data; constructing a correlation structure for the resource management subdomain and the production management subdomain according to the subdomain correlation data to generate a subdomain correlation network; Step S5: Acquire forestry actual business data; construct a progressive relationship for the subdomain association network based on the forestry actual business data to obtain a subdomain progressive relationship; Step S6: Perform process division processing on the initial business process model according to the sub-domain progressive relationship to obtain the divided business process data; build a management platform for the divided business process data to generate a process information management platform.

2. The method for building a forestry information platform based on business process modeling according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: Acquire forestry business process data; Step S12: Perform outlier detection on the forestry business process data to obtain business process outliers; Step S13: Perform integrity filling processing on the forestry business process data based on the business process abnormal values ​​to generate complete business process data; Step S14: Perform text vectorization processing on the complete business process data to generate a high-dimensional feature vector of the process; Step S15: Perform nonlinear principal component analysis on the high-dimensional feature vector of the process to obtain business process feature data.

3. The method for building a forestry information platform based on business process modeling according to claim 1 is characterized in that: Step S2 includes the following steps: Step S21: performing repeated process mining on the business process feature data to obtain frequent process fragments; Step S22: performing process pattern recognition on frequent process fragments to generate a basic business process pattern; Step S23: Performing module structure analysis on the basic model of the business process to obtain a sub-process model; Step S24: Integrate the sub-process models as a whole to generate an initial business process model.

4. The method for building a forestry information platform based on business process modeling according to claim 3 is characterized in that: Step S21 includes the following steps: Step S211: performing multi-dimensional feature decomposition on the business process feature data to generate a feature vector matrix; Step S212: performing density cluster analysis on the feature vector matrix to generate a potential process pattern; Step S213: Performing pattern frequency analysis on the potential process pattern to obtain the process repetition frequency; Step S214: Based on the process repetition frequency and a preset repetition frequency threshold, the business process feature data is mined for highly repetitive processes to generate frequent process fragments.

5. The method for building a forestry information platform based on business process modeling according to claim 1 is characterized in that: Step S3 includes the following steps: Step S31: performing semantic clustering processing on the business process feature data to obtain a process requirement cluster; Step S32: fine-grained classification of process requirement clusters to generate process requirement data; Step S33: performing demand matching on the initial business process model based on the process demand data to obtain demand matching data; Step S34: The initial business process model is divided into sub-domains according to the demand matching data to generate a resource management sub-domain and a production management sub-domain.

6. The method for building a forestry information platform based on business process modeling according to claim 5 is characterized in that: Step S33 includes the following steps: Step S331: Perform heterogeneous embedding processing on the process demand data and the initial business process model to obtain a multi-dimensional embedding feature vector; Step S332: Perform fuzzy correlation projection on the multi-dimensional embedded feature vector to obtain a demand-process fuzzy mapping; Step S333: performing self-organizing feature mapping on the demand-process fuzzy mapping to obtain a topology preserving projection; Step S334: construct a multi-layer complex network based on topology-preserving projection to obtain demand matching data.

7. The method for building a forestry information platform based on business process modeling according to claim 1 is characterized in that: Step S4 includes the following steps: Step S41: extracting features from the data of the resource management subdomain and the production management subdomain to obtain a subdomain feature vector matrix; Step S42: performing correlation analysis on the resource management subdomain and the production management subdomain according to the subdomain feature vector matrix to generate subdomain correlation data; Step S43: Structural processing is performed on the resource management subdomain and the production management subdomain based on the subdomain association data to obtain a subdomain association network.

8. The method for building a forestry information platform based on business process modeling according to claim 1 is characterized in that: Step S5 includes the following steps: Step S51: Acquire forestry actual business data; Step S52: Identify the dependency relationship of the subdomain association network to obtain subdomain dependency data; Step S53: performing dependency sequence analysis on forestry actual business data and subdomain dependency data to generate subdomain progressive sequence data; Step S54: constructing a progressive relationship for the subdomain association network based on the subdomain progressive sequence data to obtain a subdomain progressive relationship.

9. The method for building a forestry information platform based on business process modeling according to claim 8 is characterized in that: Step S53 includes the following steps: Perform tensor decomposition processing on actual forestry business data to obtain actual business feature tensors; Perform density peak clustering on subdomain dependency data to generate a subdomain dependency key point set; Perform algebraic connectivity analysis on the actual business feature tensor and subdomain dependency key point set to obtain the data dependency framework; Construct a directed acyclic graph of the data dependency framework to generate candidate progressive sequence data; The candidate progressive sequence data and the reverse multi-criteria strategy are reordered to obtain the subdomain progressive sequence data.

10. The method for building a forestry information platform based on business process modeling according to claim 1 is characterized in that: Step S6 includes the following steps: Step S61: Perform subdomain business mapping on the initial business process model according to the subdomain progressive relationship to obtain the corresponding subdomain business link; Step S62: performing process division processing on the initial business process model based on the corresponding sub-domain business links to generate division business process data; Step S63: construct an independent business management framework for the business process data in charge, and obtain a business process management platform framework; Step S64: Integrate and map the business process data based on the business process management platform framework to generate a process information management platform.

Images