In-park enterprise resource sharing platform and method based on artificial intelligence

Abstract

The invention relates to the technical field of resource sharing, in particular to an in-park enterprise resource sharing platform and method based on artificial intelligence. The method comprises the following steps: acquiring an enterprise multi-dimensional shared resource information set, analyzing the enterprise multi-dimensional shared resource information set, and determining a shared resource feature data set; acquiring real-time enterprise demand information, and determining enterprise multi-level nested demand information according to the real-time enterprise demand information; based on enterprise multi-level nested demand information, analyzing the shared resource feature data set, and determining a target resource information set; and determining and outputting a shared resource allocation plan according to the target resource information set. According to the method, the comprehensiveness and flexibility of the shared resource processing process in the park can be improved, a comprehensive and accurate resource sharing scheme is provided for enterprise demands, the docking efficiency of shared resources among enterprises in the park is improved, and the enterprise user experience is improved.

Description

Technical Field

[0001] This application relates to the technical field of resource sharing, and in particular, to an enterprise resource sharing platform and method within a park based on artificial intelligence. Background Art

[0002] In the development and operation of modern enterprise parks, resource sharing has become a key means to improve resource utilization efficiency and enterprise competitiveness. Through resource sharing, various resources of different enterprises can be effectively integrated, the overall utilization rate of resources can be improved, and at the same time, resource costs can be reduced.

[0003] However, existing enterprise resource sharing platforms in parks usually focus on a single resource directly mapped to enterprise needs. The process of sharing resources lacks comprehensiveness and flexibility, and it is difficult to provide a comprehensive and accurate resource sharing plan for enterprise needs, resulting in low efficiency in docking shared resources and poor user experience of the platform. Summary of the Invention

[0004] This application provides an enterprise resource sharing platform and method within a park based on artificial intelligence to solve the above technical problems.

[0005] In a first aspect, this application provides an enterprise resource sharing method within a park based on artificial intelligence, and the method includes: Obtain an enterprise multi-dimensional shared resource information set, analyze the enterprise multi-dimensional shared resource information set, and determine a shared resource feature data set; Obtain real-time enterprise demand information, and determine enterprise multi-level nested demand information according to the real-time enterprise demand information; Based on the enterprise multi-level nested demand information, analyze the shared resource feature data set, and determine a target resource information set; According to the target resource information set, determine and output a shared resource allocation plan.

[0006] Through this solution, analyze the enterprise multi-dimensional shared resource information set to determine a shared resource feature data set that can reflect the quantitative characteristics of each enterprise's resources within the park. At the same time, through the analysis of real-time enterprise demand information, determine enterprise multi-level nested demand information that can reflect the enterprise's resource needs from multiple demand levels. According to the enterprise multi-level nested demand information, perform resource matching on the shared resource feature data set to determine a target resource information set that meets the current enterprise needs, thereby generating a shared resource allocation plan and providing the shared resource allocation plan to the person in charge of the demanding enterprise for reference in resource planning, improving the comprehensiveness and flexibility of the shared resource processing process within the park, providing a comprehensive and accurate resource sharing plan for enterprise needs, improving the docking efficiency of shared resources among enterprises within the park, and improving the enterprise user experience.

[0007] Optionally, the enterprise multi-dimensional shared resource information set includes enterprise basic information, equipment resource information, technology resource information, talent resource information, and workflow information. Analyzing the enterprise multi-dimensional shared resource information set to determine the shared resource feature data set includes: Generating an enterprise feature index according to the enterprise basic information; Generating an equipment feature vector, a technology feature vector, and a talent feature vector according to the equipment resource information, the technology resource information, and the talent resource information respectively; Analyzing the workflow information to extract workflow timing feature data and an implementation node set; Based on the workflow timing feature data, constructing a timing-labeled workflow feature directed graph according to the implementation node set; Based on the enterprise feature index, using the generated equipment feature vector, the technology feature vector, and the talent feature vector as basic resource feature nodes, and using the timing-labeled workflow feature directed graph as a workflow chain feature node, constructing a resource feature directed graph corresponding to each enterprise, and constructing the shared resource feature data set accordingly.

[0008] Through this solution, according to the equipment resource information, the technology resource information, and the talent resource information, an equipment feature vector, a technology feature vector, and a talent feature vector are generated respectively, and at the same time, according to the workflow timing feature data and the implementation node set extracted from the workflow information, a timing-labeled workflow feature directed graph for characterizing the detailed quantitative features in the enterprise workflow is constructed. Further, based on the enterprise feature index, according to the timing-labeled workflow feature directed graph, combined with the above feature vectors, a resource feature directed graph corresponding to each enterprise is constructed, and the shared resource feature data set is constructed accordingly, realizing the comprehensive quantitative analysis of the enterprise resource features and providing a scientific data basis for subsequent resource matching.

[0009] Optionally, the constructing a timing-labeled workflow feature directed graph based on the workflow timing feature data according to the implementation node set includes: Analyzing the implementation node set to determine the total number of implementation nodes and the dependency relationship and dependency strength between any two implementation nodes; Constructing a set of directed edges according to the dependency relationship between all implementation nodes; Constructing a set of edge weights according to the dependency strength between all implementation nodes; Analyzing the workflow timing feature data to extract the task timestamp of each implementation node; Based on the implementation node set, constructing the timing-labeled workflow feature directed graph according to the set of directed edges, the set of edge weights, and the task timestamp.

[0010] Through this solution, analyze the set of implementation nodes, determine the total number of implementation nodes, and extract the dependency relationships and dependency strengths between any two implementation nodes. On this basis, construct a set of directed edges and a set of edge weights respectively, and extract the task timestamps of each implementation node characterized in the workflow timing feature data. Based on the set of implementation nodes, according to the set of directed edges, the set of edge weights, and the task timestamps, realize the automatic construction of the directed graph of the timing-marked workflow features, so as to scientifically and accurately characterize the resource features of the enterprise workflow.

[0011] Optionally, based on the set of implementation nodes, according to the set of directed edges, the set of edge weights, and the set of timing nodes, construct the directed graph of the timing-marked workflow features, specifically as the following formula: ; Wherein, is the directed graph of the timing-marked workflow features, is the set of implementation nodes, is the set of directed edges, is the th dependency relationship between the th implementation node and the is the set of timing nodes, is the th task timestamp corresponding to the is the set of edge weights, is the th dependency strength between the th implementation node and the is the total number of implementation nodes, is the directed graph construction function.

[0012] Through this solution, by means of mathematical analysis, based on the set of implementation nodes, according to the set of directed edges, the set of edge weights, and the set of timing nodes, realize the automatic construction of the directed graph of the timing-marked workflow features, clarify the relationships and timing characteristics between each real-time node, improve the accuracy of the directed graph of the timing-marked workflow features, and thus ensure the accuracy of the subsequent resource matching process based on the directed graph of the timing-marked workflow features.

[0013] Optionally, determining the enterprise multi-level nested demand information according to the real-time enterprise demand information includes: Analyze the real-time enterprise demand information, and extract the demand keyword vector set and the context embedding vector; Based on the demand keyword vector set and the context embedding vector, conduct an extended association on the enterprise demand to determine the enterprise estimated demand set; Generate a demand feature vector corresponding to each demand in the enterprise estimated demand set according to the enterprise estimated demand set; Analyze the demand feature vector corresponding to each demand, and determine the demand type index corresponding to each demand; Construct demand hierarchy structure information according to the demand type index and the demand feature vector corresponding to each demand; Construct the enterprise multi-level nested demand information according to the demand feature vector and the demand hierarchy structure information.

[0014] Through this solution, analyze the real-time enterprise demand information, extract the demand keyword vector set and the corresponding context embedding vector in the real-time enterprise demand information. On this basis, expand and associate the enterprise demand, determine the enterprise estimated demand set, and generate a demand feature vector corresponding to each demand in the enterprise estimated demand set. Furthermore, evaluate the demand type index corresponding to each demand to map the demand type to which each demand belongs. With this demand hierarchy structure information, through the integration of the demand feature vector and the demand hierarchy structure information, the enterprise multi-level nested demand information is constructed, improving the comprehensiveness and accuracy of the mapping of the enterprise multi-level nested demand information to the enterprise demand.

[0015] Optionally, the analysis of the demand feature vector corresponding to each demand to determine the demand type index corresponding to each demand is specifically the following formula: ; where is the demand type index corresponding to the th demand, is the type weight vector, is the transpose operation, is the th demand in the enterprise estimated demand set, is the th demand corresponding demand feature vector, is the bias term, is the activation function, is the total number of demands in the enterprise estimated demand set, is the th demand through obtained linear combination value.

[0016] Through this solution, using mathematical analysis means, according to the demand feature vector corresponding to each demand, through the method of linear transformation combined with bias processing, evaluate the score of each demand under different demand types, and through further normalization processing and maximum value optimization, take the maximum probability distribution value corresponding to each demand as the corresponding demand type index to ensure that the demand type index can accurately map the demand types to which different demands belong.

[0017] Optionally, constructing the requirement hierarchy structure information based on the requirement type index and the requirement feature vector corresponding to each requirement includes: The requirement types mapped by the requirement type index include core requirements, auxiliary requirements, and derivative requirements; Based on the requirement type index, according to the requirement feature vector, determine the requirement relationship index between any two requirements, specifically the following formula: ; Wherein, is the requirement relationship coefficient between the th requirement and the th requirement, is the requirement feature vector corresponding to the th requirement, is the requirement feature vector corresponding to the th requirement, is the requirement type index corresponding to the th requirement; Based on the requirement relationship index between all requirements, perform hierarchical partitioning on all requirements to determine the requirement hierarchy structure information.

[0018] Through this solution, using mathematical analysis means, based on the requirement type index and the requirement feature vector corresponding to each requirement, scientifically quantify the requirement relationship index used as the basis for requirement hierarchical partitioning, and construct requirement hierarchy information with this, improving the analysis accuracy of the requirement hierarchy, and further clarifying the importance of different requirements under the problems to be solved by the current enterprise.

[0019] Optionally, analyzing the shared resource feature data set based on the enterprise multi-level nested requirement information to determine the target resource information set includes: Based on the requirement feature vector in the enterprise multi-level nested requirement information, analyze the shared resource feature data set to determine the basic resource feature nodes and the work chain feature nodes that match the requirement feature vector; According to the requirement hierarchy structure information, perform hierarchical partitioning on the basic resource feature nodes and the work chain feature nodes to determine the hierarchical target resource node information; According to the time-sequence marked workflow feature directed graph, perform time-sequence marking on the work chain feature nodes in the hierarchical target resource node information to determine the target resource time-sequence chain; According to the hierarchical target resource node information and the target resource time-sequence chain, construct the target resource information set.

[0020] Through this solution, based on the demand feature vector in the enterprise multi-level nested demand information, the basic resource feature nodes and work chain feature nodes matching it in the shared resource feature dataset are extracted. Combining the demand hierarchy structure information, the hierarchical target resource node information corresponding to the shared resources reflecting the enterprise's demands is determined. On this basis, according to the time-series marked workflow feature directed graph, the target resource time-series chain is determined, providing a time-series reference for the formulation of subsequent resource allocation plans. By integrating the hierarchical target resource node information and the target resource time-series chain, the target resource information set is constructed.

[0021] Optionally, determining and outputting a shared resource allocation plan according to the target resource information set includes: Determining a set of target cooperative enterprises according to the hierarchical target resource node information in the target resource information set; Based on the set of target cooperative enterprises, according to the target resource time-series chain and the hierarchical target resource node information, constructing a visual resource scheduling chain and visual resource progress tracking information; Constructing and outputting the shared resource allocation plan according to the visual resource scheduling chain and the visual resource progress tracking information.

[0022] Through this solution, according to the hierarchical target resource node information, a set of target cooperative enterprises is determined. On this basis, combining the target resource time-series chain and the hierarchical target resource node information, a visual resource scheduling chain for clarifying the resource scheduling order and scheduling period and visual resource progress tracking information are respectively constructed, and the shared resource allocation plan is constructed and output based on this, providing high-value reference data for the persons in charge of relevant demand enterprises and improving the resource scheduling efficiency.

[0023] In a second aspect, the present application provides an enterprise resource sharing platform in a park based on artificial intelligence. The platform includes: a resource analysis module for obtaining an enterprise multi-dimensional shared resource information set, analyzing the enterprise multi-dimensional shared resource information set, and determining a shared resource feature dataset; a demand analysis module for obtaining real-time enterprise demand information and determining enterprise multi-level nested demand information according to the real-time enterprise demand information; a resource matching module for analyzing the shared resource feature dataset based on the enterprise multi-level nested demand information and determining a target resource information set; and an output module for determining and outputting a shared resource allocation plan according to the target resource information set.

[0024] Optionally, the resource analysis module is specifically configured to: generate an enterprise feature index according to the enterprise basic information; generate an equipment feature vector, a technology feature vector, and a talent feature vector according to the equipment resource information, the technology resource information, and the talent resource information, respectively; analyze the workflow information, extract workflow timing feature data and an implementation node set; based on the workflow timing feature data, construct a timing labeled workflow feature directed graph according to the implementation node set; based on the enterprise feature index, use the generated equipment feature vector, the technology feature vector, and the talent feature vector as basic resource feature nodes, and use the timing labeled workflow feature directed graph as a work chain feature node to construct a resource feature directed graph corresponding to each enterprise, and construct the shared resource feature data set accordingly.

[0025] Optionally, when constructing a timing labeled workflow feature directed graph based on the workflow timing feature data and according to the implementation node set, the resource analysis module is specifically configured to: analyze the implementation node set, determine the total number of implementation nodes and the dependency relationship and dependency strength between any two implementation nodes; construct a set of directed edges according to the dependency relationship between all implementation nodes; construct a set of edge weights according to the dependency strength between all implementation nodes; analyze the workflow timing feature data, and extract the task timestamp of each implementation node; based on the implementation node set, construct the timing labeled workflow feature directed graph according to the set of directed edges, the set of edge weights, and the task timestamp.

[0026] Optionally, when constructing the timing labeled workflow feature directed graph based on the implementation node set and according to the set of directed edges, the set of edge weights, and the timing node set, the resource analysis module is specifically the following formula: ; where is the timing labeled workflow feature directed graph, is the implementation node set, is the set of directed edges, is the th implementation node and the th implementation node, is the timing node set, is the th implementation node corresponding to the task timestamp, is the set of edge weights, is the th implementation node and the th implementation node, is the total number of implementation nodes, is the directed graph construction function.

[0027] Optionally, the requirement analysis module is specifically configured to: analyze the real-time enterprise requirement information, extract a requirement keyword vector set and a context embedding vector; based on the requirement keyword vector set and the context embedding vector, perform expansion and association on the enterprise requirements to determine an enterprise estimated requirement set; according to the enterprise estimated requirement set, generate a requirement feature vector corresponding to each requirement in the enterprise estimated requirement set; analyze the requirement feature vector corresponding to each requirement to determine a requirement type index corresponding to each requirement; according to the requirement type index and the requirement feature vector corresponding to each requirement, construct requirement hierarchy structure information; and construct the enterprise multi-level nested requirement information according to the requirement feature vector and the requirement hierarchy structure information.

[0028] Optionally, when the requirement analysis module analyzes the requirement feature vector corresponding to each requirement to determine the requirement type index corresponding to each requirement, it is specifically the following formula: ; Wherein, is the requirement type index corresponding to the th requirement, is the type weight vector, is the transpose operation, is the th requirement in the enterprise estimated requirement set, is the th requirement feature vector corresponding to the th requirement, is the bias term, is the activation function, is the th requirement, is the linear combination value obtained by the

[0029] Optionally, when the requirement analysis module constructs the requirement hierarchy structure information according to the requirement type index and the requirement feature vector corresponding to each requirement, it is specifically configured to: the requirement types mapped by the requirement type index include core requirements, auxiliary requirements, and derivative requirements; based on the requirement type index, according to the requirement feature vector, determine the requirement relationship index between any two requirements, specifically the following formula: ; Wherein, is the requirement relationship coefficient between the th requirement and the th requirement, is the th requirement feature vector corresponding to the is the demand feature vector corresponding to the th demand, is the demand type index corresponding to the th demand; According to the demand relationship index between all demands, all demands are hierarchically divided to determine the demand hierarchy structure information.

[0030] Optionally, the resource matching module is specifically configured to: analyze the shared resource feature data set based on the demand feature vector in the enterprise multi-level nested demand information, and determine the basic resource feature node and the work chain feature node that match the demand feature vector; according to the demand hierarchy structure information, hierarchically divide the basic resource feature node and the work chain feature node to determine the hierarchical target resource node information; according to the time-sequence marked workflow feature directed graph, perform time-sequence marking on the work chain feature node in the hierarchical target resource node information to determine the target resource time-sequence chain; according to the hierarchical target resource node information and the target resource time-sequence chain, construct the target resource information set.

[0031] Optionally, the output module is specifically configured to: determine the target cooperation enterprise set according to the hierarchical target resource node information in the target resource information set; based on the target cooperation enterprise set, construct a visual resource scheduling chain and visual resource progress tracking information according to the target resource time-sequence chain and the hierarchical target resource node information; construct and output the shared resource allocation plan according to the visual resource scheduling chain and the visual resource progress tracking information. Description of the Drawings

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0033] Figure 1 is a schematic diagram of an application scenario provided by an embodiment of the present application; Figure 2 is a flowchart of a method for sharing enterprise resources in a park based on artificial intelligence provided by an embodiment of the present application; Figure 3 is a schematic structural diagram of a platform for sharing enterprise resources in a park based on artificial intelligence provided by an embodiment of the present application. Detailed Embodiments

[0034] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are some, but not all, of the embodiments of this application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of this application without creative efforts shall fall within the scope of protection of this application.

[0035] In addition, the term "and / or" in this document is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this document generally represents an "or" relationship between the associated objects before and after, unless otherwise specified.

[0036] The embodiments of this application will be further described in detail below with reference to the accompanying drawings of the specification.

[0037] Existing park enterprise resource sharing platforms usually focus on a single resource directly mapped to enterprise needs. The process of sharing resources lacks comprehensiveness and flexibility, making it difficult to provide a comprehensive and accurate resource sharing plan for enterprise needs, resulting in low efficiency in docking shared resources and poor user experience of the platform.

[0038] Based on this, this application provides a park enterprise resource sharing platform and method based on artificial intelligence. Analyze the multi-dimensional shared resource information set of enterprises to determine a shared resource feature data set that can reflect the quantitative characteristics of the resources of each enterprise in the park. At the same time, through the analysis of real-time enterprise demand information, determine the enterprise multi-level nested demand information that can reflect the enterprise resource needs from multiple demand levels. According to the enterprise multi-level nested demand information, perform resource matching on the shared resource feature data set to determine the target resource information set that meets the current enterprise needs, thereby generating a shared resource allocation plan and providing the shared resource allocation plan to the person in charge of the demanding enterprise for reference in resource planning, improving the comprehensiveness and flexibility of the process of sharing resources in the park, providing a comprehensive and accurate resource sharing plan for enterprise needs, improving the docking efficiency of shared resources among enterprises in the park, and improving the user experience of enterprises.

[0039] Figure 1 This is a schematic diagram of an application scenario provided by this application. In the process of sharing enterprise resources in the park, the method provided by this application is applied to improve the comprehensiveness and flexibility of the process of sharing resources, so as to improve the docking efficiency of shared resources.

[0040] Specifically, the method of the present application is applied to any server, which communicates with all enterprises in the park. Through this server, the enterprise multi-dimensional shared resource information set provided by each enterprise in the park is obtained and analyzed to determine a shared resource feature data set that can reflect the quantitative characteristics of the resources of each enterprise in the park. At the same time, through the analysis of real-time enterprise demand information, a multi-level nested demand information of enterprises that can reflect the enterprise resource demand from multiple demand levels is determined. According to the multi-level nested demand information of enterprises, resource matching is performed on the shared resource feature data set to determine a target resource information set that meets the current enterprise demand, thereby generating a shared resource allocation plan and providing the shared resource allocation plan to the person in charge of the demanding enterprise for its reference in resource planning, improving the comprehensiveness and flexibility of the shared resource processing process in the park, providing a comprehensive and accurate resource sharing solution for enterprise demands, improving the docking efficiency of shared resources among enterprises in the park, and improving the enterprise user experience. The specific implementation method can refer to the following embodiments.

[0041] Figure 2 FIG. is a flowchart of a method for sharing enterprise resources in a park based on artificial intelligence provided by an embodiment of the present application. The method of this embodiment can be applied to the server in the above scenario. As Figure 2 shown, the method includes: S201. Obtain an enterprise multi-dimensional shared resource information set, analyze the enterprise multi-dimensional shared resource information set, and determine a shared resource feature data set.

[0042] The enterprise multi-dimensional shared resource information set can be an information set that reflects the shared resources that enterprises in the park can provide from multiple dimensions, such as technology, talents, services, etc. The enterprise multi-dimensional shared resource information set can be provided by each enterprise in the park in the form of an online questionnaire.

[0043] The shared resource feature data set can be a data set used to characterize the quantitative characteristics of enterprise shared resources.

[0044] Specifically, there are potential resources that can be shared among enterprises in the same enterprise park. However, due to the large information gap among enterprises, it is difficult to know the specific work processes of enterprises from the business scope reflected in the basic enterprise information alone, and it is difficult to accurately evaluate the potential shared resources of enterprises. Moreover, enterprise shared resources have multi-dimensional characteristics, and the specific resources mapped under different dimension definitions are different. In the process of resource matching for enterprise demands in the park, it is necessary to analyze whether there are enterprises in the park that can provide corresponding shared resources from different dimensions. Through mathematical analysis means, the enterprise multi-dimensional shared resource information set is analyzed to determine a shared resource feature data set that can reflect the quantitative characteristics of enterprise resources from multiple dimensions, providing accurate data support for the subsequent resource matching process.

[0045] S202. Obtain real-time enterprise demand information, and determine enterprise multi-level nested demand information according to the real-time enterprise demand information.

[0046] The real-time enterprise demand information can be the information provided by the enterprises with resource demands in the current park to describe their demands, and the real-time enterprise demand information can be filled in and submitted by the demanding enterprises through the online information submission interface.

[0047] The enterprise multi-level nested demand information can be the information that reflects the specific resource demand characteristics of the enterprise from multiple demand levels, such as the core demand level, the auxiliary demand level, and the derivative demand level, etc.

[0048] Specifically, when enterprises put forward demands, they usually tend to describe their goals. And in the process of achieving the goals they put forward, they often need the cooperation and intervention of various resources. Although the resources directly mapped by the goals in the enterprise demand information are the core resources to achieve their goals, on this basis, auxiliary resources are usually required for cooperation. And in the process of the cooperation between the auxiliary resources and the core resources, derivative resource demands will also be generated. For example, if an enterprise wants to develop a product, the supply chain resources and technical resources directly related to the product are the core resources. But at the same time, the development of this product requires the cooperation of debugging tools, cost optimization tools and other auxiliary resources. On this basis, the development of this product will also give rise to derivative resource demands such as market promotion and user experience feedback. By means of mathematical analysis, analyze the real-time enterprise demand information, analyze the enterprise demands from multiple levels, and obtain the enterprise multi-level nested demand information to comprehensively and accurately reflect the enterprise resource demands.

[0049] S203. Analyze the shared resource feature dataset based on the enterprise multi-level nested demand information, and determine the target resource information set.

[0050] The target resource information set can be the set of target shared resource information in the park required to meet the current enterprise demands.

[0051] Specifically, after analyzing and obtaining the shared resource feature dataset and the enterprise multi-level nested demand information, according to the resource characteristics in the enterprise multi-level nested demand information, retrieve the shared resource feature dataset, screen and obtain several corresponding shared resource information, and arrange these shared resource information according to the hierarchical structure in the enterprise multi-level nested demand information to construct the target resource information set, so as to clarify the shared resources in the park required for the current enterprise to meet its demands and the hierarchical relationship between these shared resources.

[0052] S204. Determine and output the shared resource allocation plan according to the target resource information set.

[0053] The shared resource allocation plan can be a proposed plan for allocating the shared resources required by the current enterprise.

[0054] Specifically, after determining the target resource information set, since the target resource information set may contain multiple types of resources under multiple enterprises at the same time, according to the temporal and hierarchical structure characteristics of each resource reflected by the target resource information set, the allocation process of shared resources is simulated to generate a corresponding shared resource allocation plan, and through data visualization technology, using a human-computer interaction device, the corresponding shared resource allocation plan is provided to the person in charge of the demanding enterprise for its reference in resource planning.

[0055] Through this solution, the multi-dimensional shared resource information set of enterprises is analyzed to determine a shared resource feature data set that can reflect the quantitative characteristics of the resources of each enterprise in the park. At the same time, through the analysis of real-time enterprise demand information, a multi-level nested demand information of enterprises that can reflect the enterprise resource demand from multiple demand levels is determined. According to the multi-level nested demand information of enterprises, resource matching is performed on the shared resource feature data set to determine the target resource information set that meets the current enterprise demand, thereby generating a shared resource allocation plan, and providing the shared resource allocation plan to the person in charge of the demanding enterprise for its reference in resource planning, improving the comprehensiveness and flexibility of the shared resource processing process in the park, providing a comprehensive and accurate resource sharing solution for enterprise demands, improving the docking efficiency of shared resources among enterprises in the park, and improving the enterprise user experience.

[0056] In some embodiments, according to the enterprise basic information, an enterprise feature index is generated; according to the equipment resource information, technology resource information, and talent resource information, an equipment feature vector, a technology feature vector, and a talent feature vector are respectively generated; the workflow information is analyzed to extract workflow temporal feature data and an implementation node set; based on the workflow temporal feature data, according to the implementation node set, a temporal labeled workflow feature directed graph is constructed; based on the enterprise feature index, with the generated equipment feature vector, technology feature vector, and talent feature vector as basic resource feature nodes, and the temporal labeled workflow feature directed graph as the workflow chain feature node, a resource feature directed graph corresponding to each enterprise is constructed, and a shared resource feature data set is constructed accordingly.

[0057] The multi-dimensional shared resource information set of enterprises includes enterprise basic information, equipment resource information, technology resource information, talent resource information, and workflow information.

[0058] The enterprise basic information can be the basic registration information of each enterprise in the park.

[0059] The equipment resource information can be the equipment resource information that an enterprise can provide externally.

[0060] The technology resource information can be the technology resource information that an enterprise can provide externally.

[0061] The talent resource information can be the talent resource information that the enterprise can provide externally.

[0062] The workflow information can be the specific workflow information corresponding to the core business of the enterprise.

[0063] The enterprise feature index can be the unique identification index used to represent the enterprise.

[0064] The device feature vector can be the vector information used to characterize the features of the enterprise's device resources.

[0065] The technical feature vector can be the vector information used to characterize the features of the enterprise's technical resources.

[0066] The talent feature vector can be the vector information used to characterize the features of the enterprise's talent resources.

[0067] The workflow time-series feature data can be the data used to characterize the features of each key time point in the enterprise's internal workflow.

[0068] The implementation node set can be the set containing each key work implementation node in the enterprise's internal workflow.

[0069] The time-series marked workflow feature digraph can be the digraph used to comprehensively reflect the quantitative features of the enterprise's workflow and with corresponding time marks.

[0070] The basic resource feature node can be the node in the digraph used to point to the resource features with clear types.

[0071] The work chain feature node can be the node in the digraph used to point to the workflow resource features.

[0072] The resource feature digraph can be the digraph used to reflect the comprehensive quantitative features of the enterprise's shared resources.

[0073] Specifically, among the types of shared resources that enterprises can provide, the types of equipment, technology, and talents are clearly defined. However, it is difficult to directly obtain the service resources that can be provided externally through the cooperation of different types of resources within an enterprise. This part of the potential service resources is hidden in the specific workflow of the enterprise. For example, in a cloud computing service company, the server resources, cloud computing technology resources, and software talent resources it owns are clearly defined, but the specific services it provides externally can be very diverse, such as financial services, marketing services, enterprise financial management, etc. Because the workflow of cloud computing services can support the above services, when dividing and extracting resource characteristics, it is necessary to consider that the enterprise workflow can support the services provided externally, bringing more diverse choices to enterprises with needs, and at the same time improving resource utilization. According to the basic information of the enterprise, the enterprise credit code, the unique identifier of the enterprise, is used as the enterprise feature index to facilitate the identification of the enterprise to which the resources belong. Through the TF-IDF algorithm (Term Frequency-Inverse Document Frequency), the equipment resource information, technology resource information, and talent resource information are respectively converted into corresponding equipment feature vectors, technology feature vectors, and talent feature vectors to quantify the characteristics of the above resource information, facilitating the subsequent mathematical analysis process for shared resources. Through the named entity recognition algorithm and event extraction algorithm in natural language analysis technology, the extraction of workflow time series feature data and implementation node sets is realized. Through the DAG algorithm (Directed Acyclic Graph), according to the implementation nodes and the dependency relationships between nodes in the workflow information, combined with the time series tags of each implementation node, a time series tagged workflow feature directed graph is generated to represent the detailed quantitative characteristics in the enterprise workflow, realizing the precise analysis of the enterprise's potential service resources. Furthermore, with the enterprise feature index as the root node, the generated equipment feature vectors, technology feature vectors, and talent feature vectors as the basic resource feature nodes, and the time series tagged workflow feature directed graph as the workflow chain feature node, through the DAG algorithm, a resource feature directed graph corresponding to each enterprise is constructed, and based on this, a shared resource feature data set is constructed to represent the overall quantitative characteristics of the enterprise resources.

[0074] Through this solution, according to the device resource information, technical resource information, and talent resource information, device feature vectors, technical feature vectors, and talent feature vectors are respectively generated. At the same time, based on the workflow timing feature data and implementation node set extracted from the workflow information, a timing-labeled workflow feature directed graph for characterizing the detailed quantitative features in the enterprise workflow is constructed. Further, based on the enterprise feature index, according to the timing-labeled workflow feature directed graph, combined with the above feature vectors, a resource feature directed graph corresponding to each enterprise is constructed, and a shared resource feature data set is constructed based on this, realizing the comprehensive quantitative analysis of enterprise resource features and providing a scientific data basis for subsequent resource matching.

[0075] In some embodiments, analyze the implementation node set, determine the total number of implementation nodes and the dependency relationship and dependency strength between any two implementation nodes; according to the dependency relationship between all implementation nodes, construct a set of directed edges; according to the dependency strength between all implementation nodes, construct a set of edge weights; analyze the workflow timing feature data, and extract the task timestamp of each implementation node; based on the implementation node set, according to the set of directed edges, the set of edge weights, and the task timestamp, construct a timing-labeled workflow feature directed graph.

[0076] The dependency relationship can be a value used to represent whether there is a medical relationship between the corresponding work of two implementation nodes. The dependency strength can be a value used to represent the degree of dependence between two real-time nodes with a dependency relationship.

[0077] The set of directed edges can be the set of edges in the currently to-be-constructed timing-labeled workflow feature directed graph. If one real-time node depends on another real-time node, starting from the dependent implementation node, make it point to the corresponding dependent node through a directed edge to represent the dependency relationship between the two nodes.

[0078] The set of edge weights can be a set containing the weights corresponding to different directed edges.

[0079] The task timestamp can be the timestamp data of the work cycle corresponding to the implementation node.

[0080] Specifically, an enterprise workflow usually contains several work steps. Each work step corresponds to a real-time node in the real-time node set. There are dependency relationships between some work steps, that is, one work step needs to be completed after the corresponding work step or requires the product of the corresponding work step. This dependency relationship is manifested as a directed edge between two implementation nodes in the time-series marked workflow feature directed graph. At the same time, there are differences in the dependency strength (strong dependency or weak dependency), and this dependency strength determines the influence weight of the directed edge. Through the relationship recognition algorithm in natural language analysis technology, analyze whether there is a dependency relationship and the corresponding dependency strength between the work steps corresponding to the implementation node set, and construct the directed edge set and the edge weight set respectively. At the same time, according to the start work node and the end work node corresponding to each implementation node in the workflow time-series feature data, determine the work cycle of the work step corresponding to each real-time node to obtain the corresponding task timestamp, providing a time reference basis for the subsequent resource allocation plan. Through mathematical analysis means, based on the implementation node set, according to the directed edge set, the edge weight set and the task timestamp, realize the automatic construction of the time-series marked workflow feature directed graph.

[0081] Through this solution, analyze the implementation node set, determine the total number of implementation nodes, and extract the dependency relationship and dependency strength between any two implementation nodes. On this basis, construct the directed edge set and the edge weight set respectively, and extract the task timestamp of each implementation node characterized in the workflow time-series feature data. Based on the implementation node set, according to the directed edge set, the edge weight set and the task timestamp, realize the automatic construction of the time-series marked workflow feature directed graph to scientifically and accurately characterize the resource characteristics of the enterprise workflow.

[0082] In some embodiments, based on the implementation node set, according to the directed edge set, the edge weight set and the time-series node set, construct the time-series marked workflow feature directed graph, specifically as the following formula (1): (1); Wherein, is the time-series marked workflow feature directed graph, is the implementation node set, is the directed edge set, is the th implementation node and the th implementation node Dependency relationship between, is the time-series node set, is the th implementation node corresponding task timestamp, is the edge weight set, is the th implementation node and the th implementation node Dependency strength between, is the total number of implementation nodes, is a directed graph construction function.

[0083] The directed graph construction function can be a function for automatically constructing a directed graph, and the directed graph construction function can adopt the DAG algorithm.

[0084] Specifically, through in formula (1) to describe the th implementation node and the th implementation node, indicating the existence of a dependency relationship and pointing out the directionality of the dependency relationship. Through describe the task timestamps associated with each implementation node. Together with the causal relationship between the time information and the work tasks of the implementation nodes, add temporal information to the directed graph. Through describe the strength of the dependency relationship between different implementation nodes, clarify the importance of the dependency, so as to reflect the coupling degree between the implementation nodes. This coupling degree provides a basis for the analysis of derivative resource requirements. Through the above mathematical descriptions, through the directed graph construction function, the accurate construction of the directed graph of the temporal labeled workflow characteristics is realized.

[0085] Through this solution, by using mathematical analysis means, based on the set of implementation nodes, according to the set of directed edges, the set of edge weights, and the set of temporal nodes, the automatic construction of the directed graph of the temporal labeled workflow characteristics is realized, clarifying the relationships and temporal characteristics between each real-time node, improving the accuracy of the directed graph of the temporal labeled workflow characteristics, and further ensuring the accuracy of the subsequent resource matching process based on the directed graph of the temporal labeled workflow characteristics.

[0086] In some embodiments, analyze the real-time enterprise demand information, extract the demand keyword vector set and the context embedding vector; based on the demand keyword vector set and the context embedding vector, conduct an extended association of the enterprise demand to determine the enterprise estimated demand set; according to the enterprise estimated demand set, generate the demand feature vector corresponding to each demand in the enterprise estimated demand set; analyze the demand feature vector corresponding to each demand to determine the demand type index corresponding to each demand; according to the demand type index and the demand feature vector corresponding to each demand, construct the demand hierarchy structure information; according to the demand feature vector and the demand hierarchy structure information, construct the enterprise multi-level nested demand information.

[0087] The demand keyword vector set can be a set containing the vector information corresponding to all demand keywords in the enterprise demand information.

[0088] The context embedding vector can be vector information used to represent the context meaning of the demand keywords in the enterprise demand information.

[0089] The enterprise estimated demand set can be a set of several demands corresponding to the current enterprise inferred from the enterprise demand information.

[0090] The demand type index can be a quantitative value used to map the type to which the current demand belongs.

[0091] The demand feature vector can be vector information used to characterize the quantitative features corresponding to the current demand.

[0092] The demand hierarchy structure information can be structure information used to characterize the corresponding hierarchical relationship between different demands of the current enterprise.

[0093] Specifically, through the keyword extraction algorithm in natural language analysis technology, the keywords mapping enterprise demands in the real-time enterprise demand information are extracted, and through a word vector model, such as Word2vec, the corresponding keywords are converted into corresponding vector information to construct a demand keyword vector set and generate corresponding context embedding vectors. On this basis, according to the demand keyword vector set and the corresponding context embedding vectors, using the corresponding existing large language models, such as GPT, the enterprise demands are expanded and associated to construct an enterprise estimated demand set, and through the word vector model, the demand feature vector corresponding to each demand in the enterprise estimated demand set is generated. Furthermore, through mathematical analysis means, the demand type index corresponding to each demand is evaluated to map the type to which the demand belongs. Based on this, the demand hierarchy structure information is constructed, and by integrating the demand feature vector and the demand hierarchy structure information, the enterprise multi-level nested demand information is determined.

[0094] Through this solution, the real-time enterprise demand information is analyzed to extract the demand keyword vector set and the corresponding context embedding vectors in the real-time enterprise demand information. On this basis, the enterprise demands are expanded and associated to determine the enterprise estimated demand set, and the demand feature vector corresponding to each demand in the enterprise estimated demand set is generated. Furthermore, the demand type index corresponding to each demand is evaluated to map the demand type to which each demand belongs. Based on this demand hierarchy structure information, by integrating the demand feature vector and the demand hierarchy structure information, the enterprise multi-level nested demand information is constructed, improving the comprehensiveness and accuracy of the mapping of the enterprise multi-level nested demand information to the enterprise demands.

[0095] In some embodiments, analyzing the demand feature vector corresponding to each demand to determine the demand type index corresponding to each demand is specifically the following formula (2): (2); Where, is the demand type index corresponding to the th demand, is the type weight vector, is the transpose operation, is the th demand in the enterprise estimated demand set, is the The demand feature vector corresponding to a demand is the bias term is the activation function is the total number of demands in the enterprise's estimated demand set is the th demand through The obtained linear combination value.

[0096] The type weight vector can be the vector information of the influence weights corresponding to the demand types. The influence weights corresponding to different demand types can be set according to historical data.

[0097] The bias term can be the initial offset corresponding to different demand types, and the bias term can be obtained by fitting historical data.

[0098] Specifically, through in formula (2), using mathematical processing means of linear transformation and bias processing, perform linear combination processing on the demand feature vector, evaluate the scores of each demand in different types. The higher the score, the more the current demand conforms to the current demand type. Then, through the activation function , perform normalization processing on the linear combination value obtained after the above processing, so that it is converted into a probability distribution value. Then, perform a maximum value query on the probability distribution values of each demand in different types, and take the maximum probability distribution value as the demand type index corresponding to the current demand.

[0099] Through this solution, using mathematical analysis means, according to the demand feature vector corresponding to each demand, through the method of linear transformation combined with bias processing, evaluate the scores of each demand in different demand types, and through further normalization processing and maximum value optimization, take the maximum probability distribution value corresponding to each demand as the corresponding demand type index to ensure that the demand type index can accurately map the demand types to which different demands belong.

[0100] In some embodiments, the demand types mapped by the demand type index include core demands, auxiliary demands, and derivative demands; based on the demand type index, according to the demand feature vector, determine the demand relationship index between any two demands, specifically the following formula (3): (3); where is the demand relationship coefficient between the th demand and the th demand, is the demand feature vector corresponding to the th demand, is the demand feature vector corresponding to the th demand, is the The demand type index corresponding to a demand; according to the demand relationship index between all demands, hierarchical division is performed on all demands to determine the demand hierarchical structure information.

[0101] The core demand can be the demand directly mapped by the current problems of the enterprise.

[0102] The auxiliary demand can be the demand used to cooperate with the core demand to solve the current problems of the enterprise.

[0103] The derivative demand can be the demand derived during the cooperation process between the core demand and the auxiliary demand.

[0104] Specifically, whether there is a hierarchical relationship between two demands depends on whether there is an association between the two demands. As the demand directly related to the current problems of the enterprise, the core demand is at the highest level. Both the auxiliary demand and the derivative demand are associated with the core demand. When mapping the core demand through the current demand type index and there is an association between the current two demands, the cosine value (reflecting the cosine similarity between the two vectors) between the two demand feature vectors is used as the demand relationship coefficient, thereby indirectly mapping the hierarchical relationship between different demands, and then constructing the demand hierarchical structure information.

[0105] Through this solution, by using mathematical analysis means, according to the demand type index and demand feature vector corresponding to each demand, the demand relationship index used as the basis for demand hierarchical division is scientifically quantified, and the demand hierarchical information is constructed based on this, improving the accuracy of demand hierarchical analysis, and then clarifying the importance of different demands under the current problems to be solved by the enterprise.

[0106] In some embodiments, based on the demand feature vectors in the enterprise multi-level nested demand information, the shared resource feature data set is analyzed to determine the basic resource feature nodes and work chain feature nodes that match the demand feature vectors; according to the demand hierarchical structure information, hierarchical division is performed on the basic resource feature nodes and work chain feature nodes to determine the hierarchical target resource node information; according to the time-series marked workflow feature directed graph, time-series marking is performed on the work chain feature nodes in the hierarchical target resource node information to determine the target resource time series chain; according to the hierarchical target resource node information and the target resource time series chain, a target resource information set is constructed.

[0107] The hierarchical target resource node information can be the target resource node information that matches the enterprise demand and has a clear hierarchical structure.

[0108] The target resource time series chain can be a set of target resources that match the enterprise demand and have clear time-series markings.

[0109] Specifically, according to the demand feature vector in the enterprise multi-level nested demand information, vector similarity retrieval matching is performed on the shared resource feature data set to determine the basic resource feature nodes and work chain feature nodes that match the demand feature vector. Then, according to the hierarchical relationship between the demands in the demand hierarchical structure information, hierarchical division is performed on the basic resource feature nodes and work chain feature nodes corresponding to different demands to obtain hierarchical target resource node information. Further, according to the time-sequence marked workflow feature directed graph, the working cycles of different implementation nodes corresponding to the work chain feature nodes obtained by the above matching are marked to obtain the target resource time-sequence chain. Then, through the integration of the hierarchical target resource node information and the target resource time-sequence chain, the target resource information set is constructed.

[0110] Through this solution, according to the demand feature vector in the enterprise multi-level nested demand information, the basic resource feature nodes and work chain feature nodes that match it in the shared resource feature data set are extracted. Combining with the demand hierarchical structure information, the hierarchical target resource node information reflecting the shared resources corresponding to the enterprise demands is determined. On this basis, according to the time-sequence marked workflow feature directed graph, the target resource time-sequence chain is determined, providing a time-sequence reference for the formulation of subsequent resource allocation plans. Through the integration of the hierarchical target resource node information and the target resource time-sequence chain, the target resource information set is constructed.

[0111] In some embodiments, according to the hierarchical target resource node information in the target resource information set, a set of target cooperative enterprises is determined; based on the set of target cooperative enterprises, according to the target resource time-sequence chain and the hierarchical target resource node information, a visual resource scheduling chain and visual resource progress tracking information are constructed; according to the visual resource scheduling chain and the visual resource progress tracking information, a shared resource allocation plan is constructed and output.

[0112] The set of target cooperative enterprises can be the set of enterprises to which each resource belongs in the hierarchical target resource node information.

[0113] The visual resource scheduling chain can be the visual information of the recommended order of resource scheduling.

[0114] The visual resource progress tracking information can be the visual information of the resource scheduling progress.

[0115] Specifically, according to the enterprise feature index corresponding to the enterprise resources in the hierarchical target resource node information, determine the enterprises to which each resource belongs, and then construct the target cooperation enterprise set. On this basis, according to the target resource time series chain and the levels corresponding to each resource in the hierarchical target resource node information, set the resource scheduling order according to the level, so as to construct the resource scheduling chain, and construct the resource progress tracking information according to the time periods corresponding to different resources. Through the data visualization tool, perform unified visualization processing on the resource scheduling chain and the resource progress tracking information, thereby constructing the shared resource allocation plan, and through the human-computer interaction device, such as a high-definition display screen, provide the shared resource allocation plan to the person in charge of the demand enterprise for reference in resource scheduling.

[0116] Through this solution, according to the hierarchical target resource node information, determine the target cooperation enterprise set. On this basis, combine the target resource time series chain and the hierarchical target resource node information to respectively construct the visual resource scheduling chain and the visual resource progress tracking information for clarifying the resource scheduling order and the scheduling period, and thereby construct and output the shared resource allocation plan, providing high-value reference data for the person in charge of relevant demand enterprises and improving the resource scheduling efficiency.

[0117] Figure 3 This is a schematic structural diagram of an enterprise resource sharing platform in a park based on artificial intelligence provided by an embodiment of the present application. As Figure 3 shown, an enterprise resource sharing platform 300 in a park based on artificial intelligence in this embodiment includes: a resource analysis module 301, a demand analysis module 302, a resource matching module 303, and an output module 304.

[0118] The resource analysis module 301 is used to obtain the enterprise multi-dimensional shared resource information set, analyze the enterprise multi-dimensional shared resource information set, and determine the shared resource feature data set; the demand analysis module 302 is used to obtain the real-time enterprise demand information, and determine the enterprise multi-level nested demand information according to the real-time enterprise demand information; the resource matching module 303 is used to analyze the shared resource feature data set based on the enterprise multi-level nested demand information, and determine the target resource information set; the output module 304 is used to determine and output the shared resource allocation plan according to the target resource information set.

[0119] Optionally, the resource analysis module 301 is specifically configured to: generate an enterprise feature index according to the enterprise basic information; generate an equipment feature vector, a technology feature vector, and a talent feature vector according to the equipment resource information, the technology resource information, and the talent resource information, respectively; analyze the workflow information, extract workflow timing feature data and an implementation node set; based on the workflow timing feature data, construct a timing-marked workflow feature directed graph according to the implementation node set; based on the enterprise feature index, use the generated equipment feature vector, the technology feature vector, and the talent feature vector as basic resource feature nodes, and use the timing-marked workflow feature directed graph as a work chain feature node to construct a resource feature directed graph corresponding to each enterprise, and construct the shared resource feature data set accordingly.

[0120] Optionally, when the resource analysis module 301 constructs a timing-marked workflow feature directed graph based on the workflow timing feature data and according to the implementation node set, it is specifically configured to: analyze the implementation node set, determine the total number of implementation nodes and the dependency relationship and dependency strength between any two implementation nodes; construct a set of directed edges according to the dependency relationship between all implementation nodes; construct a set of edge weights according to the dependency strength between all implementation nodes; analyze the workflow timing feature data, and extract the task timestamp of each implementation node; based on the implementation node set, construct the timing-marked workflow feature directed graph according to the set of directed edges, the set of edge weights, and the task timestamp.

[0121] Optionally, when the resource analysis module 301 constructs the timing-marked workflow feature directed graph based on the implementation node set, according to the set of directed edges, the set of edge weights, and the timing node set, it is specifically the following formula: ; Where is the timing-marked workflow feature directed graph, is the implementation node set, is the set of directed edges, is the th implementation node and the th implementation node is the timing node set, is the th implementation node is the set of edge weights, is the th implementation node and the th implementation node is the total number of implementation nodes, Construct a function for a directed graph.

[0122] Optionally, the requirements analysis module 302 is specifically configured to: analyze the real-time enterprise requirements information, extract a set of requirement keyword vectors and context embedding vectors; based on the set of requirement keyword vectors and the context embedding vectors, perform extended association on the enterprise requirements to determine an enterprise estimated requirements set; according to the enterprise estimated requirements set, generate a requirement feature vector corresponding to each requirement in the enterprise estimated requirements set; analyze the requirement feature vector corresponding to each requirement to determine a requirement type index corresponding to each requirement; according to the requirement type index and the requirement feature vector corresponding to each requirement, construct requirement hierarchy structure information; according to the requirement feature vector and the requirement hierarchy structure information, construct the enterprise multi-level nested requirements information.

[0123] Optionally, when the requirements analysis module 302 analyzes the requirement feature vector corresponding to each requirement to determine the requirement type index corresponding to each requirement, it is specifically the following formula: ; Where, is the requirement type index corresponding to the th requirement, is the type weight vector, is the transpose operation, is the th requirement in the enterprise estimated requirements set, is the th requirement feature vector corresponding to the th requirement, is the bias term, is the activation function, is the th requirement, is the linear combination value obtained by the

[0124] Optionally, when the requirements analysis module 302 constructs the requirement hierarchy structure information according to the requirement type index and the requirement feature vector corresponding to each requirement, it is specifically configured to: the requirement types mapped by the requirement type index include core requirements, auxiliary requirements, and derivative requirements; based on the requirement type index, according to the requirement feature vector, determine the requirement relationship index between any two requirements, specifically the following formula: ; Where, is the requirement relationship coefficient between the th requirement and the th requirement, is the the demand feature vector corresponding to the demand is the demand feature vector corresponding to the demand is the demand type index corresponding to the demand; According to the demand relationship index between all demands, all demands are hierarchically divided to determine the demand hierarchy structure information.

[0125] Optionally, the resource matching module 303 is specifically configured to: analyze the shared resource feature data set based on the demand feature vector in the enterprise multi-level nested demand information, and determine the basic resource feature node and the work chain feature node that match the demand feature vector; According to the demand hierarchy structure information, hierarchically divide the basic resource feature node and the work chain feature node to determine the hierarchical target resource node information; According to the time-series marked workflow feature directed graph, perform time-series marking on the work chain feature nodes in the hierarchical target resource node information to determine the target resource time series chain; According to the hierarchical target resource node information and the target resource time series chain, construct the target resource information set.

[0126] Optionally, the output module 304 is specifically configured to: determine the target cooperation enterprise set according to the hierarchical target resource node information in the target resource information set; Based on the target cooperation enterprise set, construct a visual resource scheduling chain and visual resource progress tracking information according to the target resource time series chain and the hierarchical target resource node information; According to the visual resource scheduling chain and the visual resource progress tracking information, construct and output the shared resource allocation plan.

[0127] The platform of this embodiment can be used to execute the method of any of the above embodiments, and its implementation principle and technical effect are similar, which will not be elaborated here.

Claims

1. An enterprise resource sharing method within a park based on artificial intelligence, characterized in that, Including: Obtain the enterprise multi-dimensional shared resource information set, analyze the enterprise multi-dimensional shared resource information set, and determine the shared resource feature data set; Obtain the real-time enterprise demand information, and determine the enterprise multi-level nested demand information according to the real-time enterprise demand information; Based on the enterprise multi-level nested demand information, analyze the shared resource feature data set, and determine the target resource information set; Determine and output the shared resource allocation plan according to the target resource information set.

2. The method according to claim 1, characterized in that The enterprise multi-dimensional shared resource information set includes enterprise basic information, equipment resource information, technology resource information, talent resource information, and workflow information. Analyzing the enterprise multi-dimensional shared resource information set to determine the shared resource feature data set includes: Generate an enterprise feature index according to the enterprise basic information; Generate an equipment feature vector, a technology feature vector, and a talent feature vector according to the equipment resource information, the technology resource information, and the talent resource information respectively; Analyze the workflow information, and extract the workflow time series feature data and the implementation node set; Based on the workflow time series feature data, construct a time series labeled workflow feature directed graph according to the implementation node set; Based on the enterprise feature index, with the generated equipment feature vector, technology feature vector, and talent feature vector as the basic resource feature nodes, and the time series labeled workflow feature directed graph as the workflow chain feature node, construct a resource feature directed graph corresponding to each enterprise, and construct the shared resource feature data set accordingly.

3. The method according to claim 2, characterized in that, The constructing the time series labeled workflow feature directed graph based on the workflow time series feature data according to the implementation node set includes: Analyze the implementation node set, and determine the total number of implementation nodes and the dependency relationship and dependency strength between any two implementation nodes; Construct a set of directed edges according to the dependency relationship between all implementation nodes; Construct a set of edge weights according to the dependency strength between all implementation nodes; Analyze the workflow time series feature data, and extract the task timestamp of each implementation node; Based on the implementation node set, construct the time series labeled workflow feature directed graph according to the set of directed edges, the set of edge weights, and the task timestamp.

4. The method according to claim 3, characterized in that, The constructing the time series labeled workflow feature directed graph based on the implementation node set according to the set of directed edges, the set of edge weights, and the time series node set is specifically the following formula: ； Among them, is the time-sequence marked workflow feature directed graph, is the set of implementation nodes, is the set of directed edges, is the th dependency relationship between the th implementation node and the is the set of time-sequence nodes, is the th task timestamp corresponding to the implementation node, is the set of edge weights, is the th dependency strength between the th implementation node and the is the total number of implementation nodes, is the directed graph construction function.

5. The method according to claim 3, wherein The determining the enterprise multi-level nested demand information according to the real-time enterprise demand information includes: Analyze the real-time enterprise demand information, and extract the demand keyword vector set and the context embedding vector; Based on the demand keyword vector set and the context embedding vector, conduct an extended association on the enterprise demand, and determine the enterprise estimated demand set; Generate a demand feature vector corresponding to each demand in the enterprise estimated demand set according to the enterprise estimated demand set; Analyze the demand feature vector corresponding to each demand, and determine the demand type index corresponding to each demand; Construct the demand hierarchy structure information according to the demand type index and the demand feature vector corresponding to each demand. Construct the multi-level nested demand information of the enterprise according to the demand feature vector and the demand hierarchy structure information.

6. The method according to claim 5, characterized in that, Analyze the demand feature vector corresponding to each demand, and determine the demand type index corresponding to each demand. Specifically, the formula is as follows: ； Among them, is the demand type index corresponding to the th demand, is the type weight vector, is the transpose operation, is the th demand in the enterprise's estimated demand set, is the th demand feature vector corresponding to the th demand, is the bias term, is the activation function, is the total number of demands in the enterprise's estimated demand set, is the linear combination value obtained by the th demand through 7. The method according to claim 5, characterized in that, Construct the demand hierarchy structure information according to the demand type index and the demand feature vector corresponding to each demand, including: The demand types mapped by the demand type index include core demands, auxiliary demands, and derivative demands; Based on the demand type index, determine the demand relationship index between any two demands according to the demand feature vector. Specifically, the formula is as follows: ； Among them, is the requirement relationship coefficient between the -th requirement and the -th requirement, is the requirement feature vector corresponding to the -th requirement, is the requirement feature vector corresponding to the -th requirement, is the requirement type index corresponding to the -th requirement; Perform hierarchical partitioning on all demands according to the demand relationship index between all demands, and determine the demand hierarchy structure information.

8. The method according to claim 7, wherein Based on the multi-level nested demand information of the enterprise, analyze the shared resource feature data set, and determine the target resource information set, including: Based on the demand feature vector in the multi-level nested demand information of the enterprise, analyze the shared resource feature data set, and determine the basic resource feature node and the work chain feature node that match the demand feature vector; Perform hierarchical partitioning on the basic resource feature node and the work chain feature node according to the demand hierarchy structure information, and determine the hierarchical target resource node information; According to the time-series marked workflow feature directed graph, perform time-series marking on the work chain feature nodes in the hierarchical target resource node information to determine the target resource time-series chain; Construct the target resource information set according to the hierarchical target resource node information and the target resource time-series chain.

9. The method according to claim 8, wherein Determine and output the shared resource allocation plan according to the target resource information set, including: Determine the target cooperation enterprise set according to the hierarchical target resource node information in the target resource information set; Based on the target cooperation enterprise set, construct a visual resource scheduling chain and visual resource progress tracking information according to the target resource time-series chain and the hierarchical target resource node information; Construct and output the shared resource allocation plan according to the visual resource scheduling chain and the visual resource progress tracking information.

10. An enterprise resource sharing platform within a park based on artificial intelligence, characterized in that, Including: A resource analysis module for obtaining the enterprise multi-dimensional shared resource information set, analyzing the enterprise multi-dimensional shared resource information set, and determining the shared resource feature data set; A demand analysis module for obtaining real-time enterprise demand information and determining the multi-level nested demand information of the enterprise according to the real-time enterprise demand information; A resource matching module for analyzing the shared resource feature data set based on the multi-level nested demand information of the enterprise and determining the target resource information set; An output module for determining and outputting the shared resource allocation plan according to the target resource information set.

Images