Project management big data dynamic optimization method

Abstract

The invention belongs to the technical field of project management, and discloses a project management big data dynamic optimization method, which comprises the following steps: acquiring project optimization information; constructing a project knowledge graph, and performing graph traversal on the project knowledge graph based on the real-time task set of the project needing to be managed to obtain a corresponding dynamic sub-graph; image reasoning is carried out based on the obtained dynamic subgraph, corresponding project expected information is obtained, whether project risks exist in the project needing to be managed or not is judged based on the project expected information, and if the risks exist, corresponding risk early warning information is generated; and constructing a standby coping strategy set based on the risk early warning information, performing strategy selection optimization based on the dependency relationship among the data nodes in the dynamic subgraph to obtain a corresponding optimal coping strategy, performing real-time monitoring on the execution process of the corresponding optimal coping strategy, and performing information feedback based on the monitoring result. According to the invention, the overall level and efficiency of project management are significantly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of project management. More specifically, the present invention relates to a method for dynamically optimizing project management big data. Background Art

[0002] In today's complex and ever-changing business environment and the wave of rapidly developing information technology, project management is facing unprecedented challenges. The limitations of traditional project management methods are becoming increasingly prominent, and innovative solutions are urgently needed, which are specifically reflected in the following aspects: 1. Traditional data collection methods are often limited to single or a few data sources. The data obtained is incomplete and untimely, making it difficult to reflect the true situation of the project. Moreover, in the data processing process, there is a lack of effective integration and standardization means, resulting in chaotic data formats and uneven data quality. This makes it difficult for managers to extract valuable information from it and unable to provide strong support for project decision-making, increasing the risk of decision-making errors. 2. There is a lack of tools for intuitively and comprehensively displaying the overall project architecture and task relationships. Managers are difficult to grasp the overall picture of the project, and have insufficient understanding of the project progress, potential problems, and the synergistic effects between tasks. This not only affects the communication and collaboration efficiency among team members, but also easily leads to problems such as poor task connection and unreasonable resource allocation, thereby delaying the project progress and increasing the project cost. 3. Traditional risk assessment methods mainly rely on subjective experience and simple analysis models, unable to accurately predict potential risks, often reacting passively after the risk occurs and missing the best prevention and control opportunity. In addition, in the face of risks, there is a lack of a systematic mechanism for generating and optimizing response strategies, and the response measures taken are often not targeted and ineffective, making it difficult to effectively reduce risk losses and ensure the smooth progress of the project. 4. Traditional project management methods lack the ability to optimize strategies in real time and make dynamic adjustments, and cannot respond in a timely manner according to the actual situation of the project and environmental changes. This makes the response strategies lag behind when the project faces unexpected situations or new problems, unable to adapt to the dynamic development needs of the project and affecting the achievement of project goals.

[0003] In view of this, the present invention proposes a method for dynamically optimizing project management big data to solve the above problems. Summary of the Invention

[0004] In order to overcome the above defects of the prior art and achieve the above object, the present invention provides the following technical solutions: A method for dynamically optimizing project management big data, comprising: Step 1: Obtain project management information, and perform data preprocessing on the corresponding project management information based on a pre-constructed ontology database to obtain corresponding project optimization information; Step 2: Construct a project knowledge graph, and traverse the project knowledge graph based on the real-time task set of the project to be managed to obtain corresponding dynamic subgraphs; Step 3: Perform image reasoning based on the obtained dynamic subgraphs to obtain corresponding project expected information, and determine whether there are project risks in the project to be managed based on it. If there are risks, generate corresponding risk warning information; Step 4: Construct a set of backup response strategies based on the risk warning information, optimize the strategy selection based on the dependency relationships between various data nodes in the dynamic subgraph to obtain corresponding optimal response strategies, monitor the execution process of the corresponding optimal response strategies in real time, and perform information feedback based on the monitoring results.

[0005] Furthermore, the process of obtaining project optimization information includes: Extract data entities from the obtained project management information to obtain corresponding source data entities, where the source data entities include entity names and word vectors; Construct an ontology database, compare the obtained source data entities with the project management entity ontology in the corresponding ontology database to obtain corresponding entity similarity metrics; Compare the obtained entity similarity metrics with a pre-set similarity threshold, and construct entity combinations based on the comparison results; Map the project management information to the semantic and data format unified with the ontology database based on the obtained entity combinations; and synchronously extract entity attribute information and relationship information; After the mapping is completed, construct a corresponding project Gantt chart based on the mapped project management information; Obtain the dependency relationships corresponding to project tasks at corresponding time nodes based on the project Gantt chart, and determine the number of dependency edges based on it; construct a corresponding dynamic window based on the number of dependency edges; Obtain the local project management information corresponding to the corresponding dynamic window, input it into a pre-selected variational autoencoder to obtain corresponding reconstructed project information; and perform anomaly assessment on the task data of project tasks at corresponding time nodes based on it to obtain corresponding data scores; evaluate whether the task data at corresponding time nodes is abnormal based on the data scores. If there is an abnormality, correct the data; Repeat the above process until the data correction of all abnormal data points is completed to obtain corresponding project optimization data.

[0006] Furthermore, the process of constructing the project knowledge graph includes: Based on the project optimization data corresponding to the corresponding project set, divide the corresponding project set into several sub-projects according to functional modules, business areas or organizational structures; obtain the data entities corresponding to each sub-project and record them as project entities; obtain the communication records corresponding to each sub-project based on the project optimization information; Construct data nodes, map the project entities to the corresponding data nodes, and store the entity attribute information of the corresponding project entities in the corresponding data nodes; Connect the data nodes based on the task relationships and communication records in the obtained project Gantt chart; Obtain project document data, extract text features from it to obtain corresponding text feature vectors, synchronously convert the time information in the corresponding project Gantt chart into time vectors, store the corresponding text feature vectors and time vectors in the corresponding data nodes, and obtain corresponding project sub-graphs; Combine the project sub-graphs corresponding to different sub-projects to obtain the corresponding project knowledge graph.

[0007] Furthermore, the process of obtaining the dynamic sub-graph includes: Obtain real-time project optimization information, and based on it, obtain the project phase and the set of real-time tasks being executed by the project to be managed at the current moment; Based on the project phase and the set of real-time tasks, match the corresponding data nodes in the corresponding project knowledge graph and use them as the path starting points, and recursively extract all other data nodes and edges directly or indirectly related to the project tasks based on them to obtain the corresponding connected sub-graph; obtain the time attributes corresponding to each project task in the set of real-time tasks, and based on them, perform temporal arrangement on the extracted data nodes to transform the static project sub-graph into a dynamic sub-graph with temporal characteristics.

[0008] Furthermore, the process of performing image inference based on the obtained dynamic sub-graph includes: Input the obtained dynamic sub-graph into a pre-constructed graph neural network inference model to obtain corresponding project expected information, where the project expected information includes task index prediction data and potential node risk information; Furthermore, based on the project knowledge graph, structurally organize the corresponding project expected information to obtain corresponding project prediction information, where the project prediction information includes the predicted values, confidence intervals, risk point descriptions, risk levels, influence ranges and corresponding time information of each key indicator.

[0009] Furthermore, the backbone network of the graph neural network inference model is a hierarchical neural network, which is composed of a graph neural network and a temporal neural network, and the basic framework of the hierarchical neural network includes an input layer, a convolutional layer, a prediction layer, a detection layer and an output layer; The input layer is used to receive the input dynamic subgraph and perform image preprocessing on it; The convolutional layer is used to perform graph convolution operations on the dynamically subgraph after image preprocessing to obtain the corresponding node feature matrix; The prediction layer is used to read the historical task parameter sequences of each data node in the dynamic subgraph, perform data prediction based on the historical task parameter sequences and the node feature matrix, obtain the predicted states corresponding to the data nodes at several future time nodes, perform state analysis on the corresponding predicted states to obtain the corresponding feature vectors, and obtain the predicted values of each key indicator of the project task corresponding to the corresponding data node based on the feature vectors; at the same time, the Monte Carlo dropout method is used to predict the confidence intervals corresponding to the corresponding key indicators; The detection layer respectively performs anomaly scoring on the corresponding data nodes and the edges corresponding to the nodes based on the feature vectors in the prediction layer, and determines whether the corresponding data node is an abnormal node based on this. If it is an abnormal node, the risk level is divided based on the anomaly score to obtain the corresponding node risk level; The output layer is used to receive the predicted values of the key indicators, the anomaly scores and risk levels of each data node, and map them into a pre-set data space to obtain the corresponding project prediction information.

[0010] Furthermore, the process of determining whether there is an abnormal risk in the project to be managed based on the project prediction information includes: Based on the obtained project prediction information, calculate the deviation between the key indicator parameters corresponding to each project task in the corresponding project prediction information and the corresponding confidence interval, obtain the parameter deviation degree of each project task, and perform a comprehensive anomaly assessment on the project to be managed based on this to obtain the corresponding project risk score; Set a risk threshold, compare the obtained project risk score with the corresponding risk threshold, and determine whether the project risk is in a controllable state based on this. If it is in a controllable state, continue to monitor the project; if it is not in a controllable state, generate the corresponding risk warning information.

[0011] Furthermore, the process of the set of alternative countermeasures includes: When a risk warning information is recognized, obtain the data nodes with risks in the dynamic subgraph and mark them as risk nodes; Based on the project prediction information, obtain the risk types corresponding to the corresponding risk nodes and the triggering probabilities of each risk type; Based on the project prediction information, obtain the risk feature vectors of the project tasks corresponding to the corresponding risk nodes, and combine the task parameters and risk feature vectors of the project tasks to obtain several similar historical project task cases from the historical project database, and use the risk countermeasures of the corresponding historical project task cases as reference countermeasures; Obtain the task differences between the current project tasks and the historical project tasks, and adjust the corresponding reference coping strategies based on them to obtain a corresponding set of alternative coping strategies.

[0012] Furthermore, the process of obtaining the optimal coping strategy includes: Based on the dynamic subgraph, obtain the dependency relationships between the corresponding risk nodes and other data nodes, and based on the influence intensity of other data nodes on the corresponding risk nodes under their corresponding risk types; Obtain the influence intensity of other data nodes on the corresponding risk nodes, and combine the obtained risk types and the set of alternative coping strategies to construct a corresponding project selection optimization model; and define the objective function of the corresponding project selection optimization model based on the preset constraint conditions; Obtain the task execution scenario, and construct a number of expected task execution scenarios based on the uncertain factors in the task execution scenario; obtain the strategy stability indicators corresponding to the alternative coping strategies based on the expected task execution scenarios; Obtain the management preferences of the corresponding staff for the projects to be managed and the strategy stability indicators in each expected task execution scenario, and solve the corresponding objective function by combining the particle swarm optimization algorithm to obtain the corresponding optimal coping strategy.

[0013] Furthermore, the process of monitoring the execution process of the corresponding optimal coping strategy in real time and providing information feedback based on the monitoring results includes: Construct a real-time monitoring mechanism to dynamically track the process of the corresponding optimal coping strategy, collect the key indicators of the corresponding project tasks during the execution of the project optimal coping strategy in real time, and compare them with the preset expected goals. If the expected goals are not met, collect the project management information of the corresponding managed project in real time and perform data preprocessing, and update the dynamic subgraph based on it. Optimize the currently executed optimal coping strategy based on the updated dynamic subgraph; at the same time, feedback the strategy execution results to the project knowledge graph to update the corresponding data nodes.

[0014] The technical effects and advantages of a dynamic optimization method for project management big data according to the present invention: 1. By obtaining the dynamic subgraph through real-time tasks, clearly display the real-time status of the project and the task associations; enable managers to grasp the project from a global perspective, make communication and collaboration between team members smoother, effectively avoid task connection problems, achieve reasonable resource allocation, improve project execution efficiency, and ensure the orderly progress of the project; 2. By performing image inference on the dynamic subgraph, accurately predicting task metrics and potential risks, and structurally organizing detailed project prediction information; by calculating the project risk score and comparing it with the threshold, timely and accurately determining the risk status and issuing early warnings; enabling managers to plan risk response strategies in advance, changing from passive response to proactive prevention and control, significantly reducing the probability of risk occurrence, reducing the losses caused by risks, and ensuring the achievement of project goals. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic diagram of a method for dynamically optimizing project management big data according to the present invention; Figure 2 It is a schematic diagram of a system for dynamically optimizing project management big data according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0017] Embodiment 1 Please refer to Figure 1 As shown, a method for dynamically optimizing project management big data in this embodiment includes: Step 1: Obtain project management information, and perform data preprocessing on the corresponding project management information based on a pre-constructed ontology database to obtain corresponding project optimization information; Step 2: Construct a project knowledge graph, and perform graph traversal on the project knowledge graph based on the real-time task set of the project to be managed to obtain a corresponding dynamic subgraph; Step 3: Perform image inference based on the obtained dynamic subgraph to obtain corresponding project expected information, and determine whether there are project risks in the project to be managed based on it. If there are risks, generate corresponding risk warning information; Step 4: Construct a set of alternative response strategies based on the risk warning information, perform strategy selection optimization based on the dependency relationship between each data node in the dynamic subgraph to obtain corresponding optimal response strategies, monitor the execution process of the corresponding optimal response strategies in real time, and perform information feedback based on the monitoring results.

[0018] It should be further noted that, in the specific implementation process, the process of obtaining project management information includes: A data acquisition unit is set up to perform multi-source data acquisition on the target management project based on the data acquisition unit, and obtain corresponding project management information. The project management information includes internal information and external information. The internal information includes project parameters such as project risks, response strategies, costs, schedules, and quality. The external data includes information such as market environment, industry trends, and policies and regulations. Among them, the data acquisition unit is a distributed data acquisition terminal, which is connected to multiple data sources through an API interface and obtains information based on the API interface. It should be further noted that in the specific implementation process, the process of obtaining project optimization information includes: Extract data entities such as tasks, risks, resources, and personnel from the obtained project management information to obtain corresponding source data entities. The source data entities include entity names and word vectors. Construct an ontology database, which includes a project domain ontology, a project management entity ontology, and a project risk ontology, and is used to define the concept system and entity relationship rules in the project management field. Among them, the construction process of the ontology database is prior art, and the present invention will not elaborate too much. Compare the obtained source data entities with the project management entity ontology in the corresponding ontology database to obtain the corresponding entity similarity metric ; where e1 and e2 respectively represent the source data entity and the project entity ontology in the ontology database, and both represent entity names; is used to calculate the edit distance between e1 and e2. The edit distance refers to the minimum number of edit operations required to convert one string to another, including inserting, deleting, and replacing characters; and respectively represent the word vectors of the entities; The matching score based on domain rules, where the domain rules are preset by the staff in this field based on industry standards; , and represent weight parameters, and ; Compare the obtained entity similarity metric with a preset similarity threshold. If the entity similarity metric is not less than the similarity threshold, then classify the corresponding source data entity and the corresponding project entity ontology into the same category to obtain the corresponding entity combination; Furthermore, based on the obtained entity combinations, map the project management information under multi-source data collection to the semantic and data formats unified with the ontology database; and synchronously extract entity attribute information and relationship information; for example, if the source data entity and the project management entity ontology in the entity combination are "project delay" and "schedule delay" respectively, then map them uniformly to "schedule delay".

[0019] After the mapping is completed, construct a corresponding project Gantt chart based on the mapped project management information. The project Gantt chart includes the time information of each project task in the corresponding project set and the dependency relationships between tasks; Taking a certain project task as an example, obtain the dependency relationship corresponding to the project task at the corresponding time node based on the project Gantt chart, and determine the number of dependency edges based on it; construct a corresponding dynamic window based on the number of dependency edges. The formula for constructing the dynamic window is: ; where, 、 and represent trainable parameters obtained by regression fitting of historical project data; represents the maximum window limit, and represent the number of dependency edges and the total number of tasks of a certain task respectively; and represent the resource volatility of the corresponding task and the historical number of anomalies of the corresponding task respectively; Furthermore, obtain the local project management information corresponding to the corresponding dynamic window, and input it into a pre-selected variational autoencoder to obtain the corresponding reconstructed project information; and perform anomaly evaluation on the task data of the project task at the corresponding time node based on it to obtain the corresponding data score ; where, represents the weight parameter, and by default ; x and x^ represent the local project management data and the reconstructed project data of the dynamic window respectively; represents the KL divergence operation, which is used to measure the difference between two probability distributions, represents the standard normal distribution, The probability distribution of the latent variable z under the original data; among them, the working principle of the variational autoencoder is to map the input local project optimization data to the latent space through the built-in encoder to obtain the corresponding latent variable and probability distribution; furthermore, reconstruct the latent variable through the decoder to obtain the corresponding reconstructed project information; Set an anomaly threshold. If the anomaly score is less than the anomaly threshold, no other operations are performed. If the anomaly score is not less than the anomaly threshold, it indicates that the task data of the project task at the corresponding time point is abnormal data. Then, optimize the abnormal data points based on the preset business constraints, and replace the original abnormal data points with the optimized abnormal data points. Among them, the optimized abnormal data points need to meet the business constraints and the anomaly score is less than the anomaly threshold. Among them, the business constraints include resource constraints, business cost constraints, task time constraints, etc. The optimization process of the abnormal data points refers to searching for the best adaptation value of the corresponding abnormal data points within the corresponding latent variables based on the business constraints. Repeat the above process until all abnormal data points are corrected, and obtain the corresponding project optimization data.

[0020] It should be further noted that in the specific implementation process, the construction process of the project knowledge graph includes: Based on the project optimization data corresponding to the corresponding project set, divide the corresponding project set into several sub-projects according to functional modules, business domains or organizational structures. For example, a software development project can be divided into sub-projects such as requirements analysis, design, coding, testing, and deployment. Among them, each sub-project can correspond to multiple project tasks. Obtain the data entities corresponding to each sub-project and record them as project entities. Based on the project optimization information, obtain the communication records corresponding to each sub-project. The communication records reflect the collaboration relationship and task allocation situation among personnel. Construct data nodes, map the project entities into the corresponding data nodes, and at the same time, obtain the entity attribute information corresponding to the project entities based on the ontology database and store it in the corresponding data nodes. Furthermore, based on the task relationships and communication records in the obtained project Gantt chart, connect the data nodes. Obtain project document data, extract text features from it to obtain the corresponding text feature vectors, synchronously convert the time information in the corresponding project Gantt chart into time vectors, and store the corresponding text feature vectors and time vectors in the corresponding data nodes to obtain the corresponding project sub-graphs to enhance the information representation ability of the nodes. Combine the project sub-graphs corresponding to different sub-projects to obtain the corresponding project knowledge graph. The project knowledge graph consists of several data nodes and the edges connecting the data nodes. It should be further noted that in the specific implementation process, the process of obtaining the dynamic sub-graph includes: Obtain real-time project optimization information, and based on it, obtain the project phase and the set of real-time tasks being executed by the project to be managed at the current moment. Furthermore, based on the project stage and the real-time task set, corresponding data nodes are matched within the corresponding project knowledge graph, and used as the path starting point, and all other data nodes and edges directly or indirectly related to the project tasks are recursively extracted based on it to obtain the corresponding connected subgraph; the time attributes corresponding to each project task in the real-time task set are obtained, and based on them, the extracted data nodes are arranged in time sequence, converting the static project subgraph into a dynamic subgraph with time sequence characteristics; the time sequence arrangement process uses a topological sorting algorithm.

[0021] It should be further noted that in the specific implementation process, the process of performing image inference based on the obtained dynamic subgraph includes: Input the obtained dynamic subgraph into a pre-constructed graph neural network inference model to obtain the corresponding project expected information, where the project expected information includes task index prediction data and potential node risk information; Furthermore, based on the project knowledge graph, the corresponding project expected information is structurally sorted to obtain the corresponding project prediction information, where the project prediction information includes the predicted values, confidence intervals, risk point descriptions, risk levels, influence ranges, and corresponding time information of each key indicator; Among them, the construction process of the graph neural network inference model includes: The backbone network of the graph neural network inference model is a hierarchical neural network, which is composed of a graph neural network and a time sequence neural network. The basic framework of the hierarchical neural network includes an input layer, a convolutional layer, a prediction layer, a detection layer, and an output layer; The input layer is used to receive the input dynamic subgraph and perform image preprocessing on it to meet the input requirements of the hierarchical neural network; The convolutional layer is used to perform graph convolution operations on the dynamically preprocessed subgraph to obtain the corresponding node feature matrix. Among them, the mathematical formula for the graph convolution operation is: ; in the formula, and respectively represent the node feature matrices corresponding to the -th convolutional layer and the -th convolutional layer. The node feature matrix is used to represent the feature information of the corresponding data nodes at the corresponding level; D represents a diagonal matrix, where the matrix elements on the main diagonal of the diagonal matrix are the number of edges connected to the corresponding data nodes; represents a self-loop matrix. The self-loop matrix means that in addition to the original node connection information, a connection of the node itself (i.e., a self-loop) is added on the main diagonal, which helps to retain the information of each node itself and prevent the over-smoothing problem to a certain extent; and softmax represent the weight matrix and the activation function respectively; The prediction layer is used to read the historical task parameter sequences of each data node in the dynamic subgraph, perform data prediction based on the historical task parameter sequences and the node feature matrix, obtain the predicted states corresponding to the data nodes at several future time nodes, perform state analysis on the corresponding predicted states to obtain corresponding feature vectors, and obtain the predicted values of each key indicator of the project task corresponding to the corresponding data node based on the feature vectors; meanwhile, the Monte Carlo dropout method is used to predict the confidence intervals corresponding to the corresponding key indicators;

[0022] The detection layer respectively performs anomaly scoring on the corresponding data node and the node edge corresponding to the node according to the feature vectors in the prediction layer, and determines whether the corresponding data node is an abnormal node based on it. If it is an abnormal node, the risk level is divided based on the anomaly score to obtain the corresponding node risk level; among them, the formula for anomaly scoring of the data node is: ; in the formula, represents the anomaly score of the i-th data node in the dynamic subgraph; and respectively represent the actual feature vector and the predicted feature vector of the corresponding data node; The formula for anomaly scoring of the node edge corresponding to the corresponding data node is: ; in the formula, represents the anomaly score of the j-th node edge corresponding to the i-th data node, represents the vector splicing operation; represents the actual feature vector of the data node connected by the j-th edge; represents a pre-selected feature mapping function; The output layer is used to receive the predicted values of the key indicators, the anomaly scores and risk levels of each data node, and map them into a pre-set data space to obtain the corresponding project prediction information; Define the loss function of the graph neural network inference model, construct a training data set based on the historical project optimization information, and perform iterative training on the corresponding network inference model based on the training data set until the corresponding loss function gradually converges; that is, the model training is completed; It should be further noted that in the specific implementation process, the process of determining whether there is an abnormal risk in the project to be managed based on the project prediction information includes: Based on the obtained project prediction information, calculate the deviation between the key indicator parameters corresponding to each project task in the corresponding project prediction information and the corresponding confidence interval, obtain the parameter deviation degree of each project task, and perform a comprehensive anomaly assessment on the project to be managed based on it to obtain the corresponding project risk score ; in the formula, and represent risk weights; , respectively represent the index weight and parameter deviation degree of the key index corresponding to the i-th data node, represents the index weight of the j-th edge connected to the i-th data node, represents the risk level corresponding to the data node; I0 and i respectively represent the total number and index of data nodes in the dynamic subgraph; Set a risk threshold, compare the obtained project risk score with the corresponding risk threshold, and determine whether the project risk is in a controllable state based on it. If it is in a controllable state, continue to monitor the project; if it is not in a controllable state, generate corresponding risk warning information.

[0023] It should be further noted that in the specific implementation process, the process of constructing the set of alternative response strategies includes: When a risk warning information is identified, obtain the data nodes with risks in the dynamic subgraph and mark them as risk nodes; Based on the risk point description, risk level, and influence scope in the project prediction information, obtain the risk types corresponding to the corresponding risk nodes and the triggering probabilities of each risk type; Furthermore, based on the project prediction information, obtain the risk feature vector of the project task corresponding to the corresponding risk node, and combine the task parameters and risk feature vector of the project task to obtain several similar historical project task cases from the historical project database, and use the risk response measures of the corresponding historical project task cases as reference response strategies; among them, the historical project database stores project management information, risk feature vectors, and risk response measures; Furthermore, obtain the task differences between the current project task and the historical project task, and adjust the corresponding reference response strategies based on them to obtain the corresponding set of alternative response strategies; among them, the adjustment process of the reference response strategy is: model the adjustment process of the corresponding reference response strategy as a Markov decision process, use the corresponding task differences as the state space, and the strategy adjustment decision as the action space; optimize the corresponding reference response strategy based on the deep deterministic policy gradient algorithm to obtain the corresponding set of alternative response strategies; It should be further noted that in the specific implementation process, the process of obtaining the optimal response strategy includes: Taking a certain risk node as an example, based on the dynamic subgraph, obtain the dependency relationship between the corresponding risk node and other data nodes, and the influence intensity of other data nodes on the corresponding risk node under its corresponding risk type; among them, the formula for obtaining the influence intensity is: ; in the formula, represents the influence intensity of risk type a of the project task corresponding to the risk on other projects with a dependency relationship during its corresponding project cycle; It is indicated that the corresponding project cycle is divided into T project sub - cycles, and t0 represents the time index within the project sub - cycle; It represents the project tasks that are being executed at time t0 except for the project tasks corresponding to the risk nodes; k represents the index of the number of the other project tasks that are being executed; It represents the dependence strength between the project tasks corresponding to the risk nodes and other project tasks; it is determined by the direct and indirect dependence relationships between projects; and respectively represent the success probability and failure probability of the other executed project task k; and respectively represent whether there is an impact when project task k fails or succeeds. If there is an impact, the value is 1; if there is no impact, the value is 0; among them, , , and are all set by the corresponding staff based on the actual situation of the project to be managed; Furthermore, obtain the influence strength of other data nodes on the corresponding risk nodes, and construct the corresponding project selection optimization model by combining the obtained risk types and the set of alternative coping strategies; and define the objective function of the corresponding project selection optimization model based on the pre - set constraint conditions; the constraint conditions include dependence relationships, resource constraints, time constraints, and cost constraints; Furthermore, obtain the task execution scenarios of the project tasks corresponding to the corresponding risk nodes, and obtain the uncertain factors under the corresponding task execution scenarios. The uncertain factors include task duration, resource availability, etc.; and construct several possible task execution scenarios based on them, and assign occurrence probability weights to each task execution scenario; furthermore, simulate and run the corresponding alternative coping strategies under the corresponding task execution scenarios, and obtain the strategy stability indicators corresponding to the corresponding alternative coping strategies. The strategy stability indicators are used to measure the fluctuation degree of the corresponding alternative coping strategies under different task execution scenarios. The higher the fluctuation degree, the more sensitive the alternative coping strategy is to the corresponding task execution scenario; among them, the simulation software of digital twin technology is used in the corresponding simulation operation process; Obtain the management preferences of the corresponding staff for the project to be managed and the strategy stability indicators under each task execution scenario, and solve the corresponding objective function by combining the particle swarm optimization algorithm to obtain the corresponding optimal coping strategy; and apply it to the project to be managed; Build a real-time monitoring mechanism to dynamically track the process of the corresponding optimal response strategy, collect the key indicators of the corresponding project tasks during the execution of the project's optimal response strategy in real time, and compare them with the preset expected goals. If the expected goals are not met, collect the project management information of the corresponding management project in real time, perform data preprocessing, and feedback it to update the dynamic subgraph within the steps. Based on this, optimize the currently executed optimal response strategy; at the same time, feedback the strategy execution results to the project knowledge graph, update the attribute values and relationship strengths of relevant nodes, and add effective experience to the historical project experience database to form a closed-loop optimization mechanism for continuously optimizing project management methods and improving the management efficiency of future projects.

[0024] The present invention obtains project management information through a distributed data collection unit, preprocesses it through an ontology database to obtain project optimization information; then constructs a project knowledge graph and obtains a dynamic subgraph based on real-time tasks; then uses a graph neural network inference model to perform image inference on the dynamic subgraph, determine whether there are risks in the project, and generate early warnings; then construct a set of alternative response strategies based on the early warnings, optimize them to obtain the optimal response strategy and monitor it in real time; it can predict risks in advance and provide early warning and prevention and control; optimize strategy selection to improve the effectiveness of response plans; form a closed-loop optimization through real-time monitoring feedback to continuously improve project management methods and management efficiency.

[0025] Embodiment 2 Please refer to Figure 2 As shown, for the parts not described in detail in this embodiment, refer to the description content of Embodiment 1. Provide a project management big data dynamic optimization system, including: An information acquisition module, which acquires project management information and performs data preprocessing on the corresponding project management information based on a pre-constructed ontology database to obtain the corresponding project optimization information; A graph construction module, which is used to construct a project knowledge graph and perform graph traversal on the project knowledge graph based on the real-time task set of the project to be managed to obtain the corresponding dynamic subgraph; A risk assessment module, which is used to perform image inference based on the obtained dynamic subgraph to obtain the corresponding project expected information, and based on this, determine whether there are project risks in the project to be managed. If there are risks, generate the corresponding risk warning information; A strategy construction module, which constructs a set of alternative response strategies based on the risk warning information, and performs strategy selection optimization based on the dependency relationships between various data nodes in the dynamic subgraph to obtain the corresponding optimal response strategy, A monitoring and feedback module, which is used to monitor the execution process of the corresponding optimal response strategy in real time and perform information feedback based on the monitoring results.

[0026] Each module is connected by wired and / or wireless means to realize data transmission between modules.

[0027] Example 3 This embodiment discloses and provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the operation mode of the above-provided method for dynamically optimizing project management big data.

[0028] Since the electronic device introduced in this embodiment is the electronic device adopted for implementing a method for dynamically optimizing project management big data in an embodiment of the present application, based on the method for dynamically optimizing project management big data introduced in the embodiment of the present application, those skilled in the art can understand the specific implementation manners and various variations of the electronic device in this embodiment. Therefore, the specific implementation of how this electronic device implements the method in the embodiment of the present application will not be described in detail here. As long as those skilled in the art implement the electronic device adopted for implementing a method for dynamically optimizing project management big data in the embodiment of the present application, it falls within the scope of protection of the present application.

[0029] The above formulas are all dimensionless and take their numerical values for calculation. The formula is obtained by collecting a large amount of data and performing software simulation to obtain a formula closest to the actual situation. The preset parameters and threshold selection in the formula are set by those skilled in the art according to the actual situation.

[0030] The above is only the preferred implementation manner of the present invention. The protection scope of the present invention is not limited to the above embodiments. Any technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for ordinary technical users in the technical field, several improvements and refinements made without departing from the principle of the present invention should also be regarded as within the protection scope of the present invention.

Claims

1. A method for dynamically optimizing project management big data, characterized in that, Including: Step 1: Obtain project management information, and perform data preprocessing on the corresponding project management information based on a pre-constructed ontology database to obtain corresponding project optimization information; Step 2: Construct a project knowledge graph, and perform graph traversal on the project knowledge graph based on the real-time task set of the project to be managed to obtain a corresponding dynamic subgraph; Step 3: Perform image reasoning based on the obtained dynamic subgraph to obtain corresponding project expected information, and determine whether there are project risks in the project to be managed based on it. If there are risks, generate corresponding risk warning information; Step 4: Construct a set of alternative coping strategies based on the risk warning information, optimize the strategy selection based on the dependency relationships between various data nodes in the dynamic subgraph, obtain corresponding optimal coping strategies, monitor the execution process of the corresponding optimal coping strategies in real time, and perform information feedback based on the monitoring results.

2. The method for optimizing project management big data according to claim 1, wherein The process of obtaining project optimization information includes: Extract data entities from the obtained project management information to obtain corresponding source data entities, where the source data entities include entity names and word vectors; Construct an ontology database, compare the obtained source data entities with the project management entity ontology in the corresponding ontology database to obtain corresponding entity similarity metrics; Compare the obtained entity similarity metrics with a pre-set similarity threshold, and construct entity combinations based on the comparison results; Map the project management information to the same semantic and data format as the ontology database based on the obtained entity combinations; and synchronously extract entity attribute information and relationship information; After the mapping is completed, construct a corresponding project Gantt chart based on the mapped project management information; Obtain the dependency relationships corresponding to project tasks at corresponding time nodes based on the project Gantt chart, and determine the number of dependency edges based on it; construct a corresponding dynamic window based on the number of dependency edges; Obtain the local project management information corresponding to the corresponding dynamic window, and input it into a pre-selected variational autoencoder to obtain corresponding reconstructed project information; and perform anomaly assessment on the task data of project tasks at corresponding time nodes based on it to obtain corresponding data scores; evaluate whether there are anomalies in the task data at corresponding time nodes based on the data scores. If there are, correct the data; Repeat the above process until the data correction of all abnormal data points is completed to obtain corresponding project optimization data.

3. The method for optimizing project management big data according to claim 2, wherein The process of constructing a project knowledge graph includes: Based on the project optimization data corresponding to the corresponding project set, divide the corresponding project set into several sub-projects according to functional modules, business domains or organizational structures; obtain the data entities corresponding to each sub-project and record them as project entities; obtain the communication records corresponding to each sub-project based on the project optimization information; Construct data nodes, map the project entities to the corresponding data nodes, and store the entity attribute information of the corresponding project entities in the corresponding data nodes; Connect the data nodes based on the task relationships and communication records in the obtained project Gantt chart; Obtain the project document data, extract the text features therefrom to obtain the corresponding text feature vectors, synchronously convert the time information in the corresponding project Gantt chart into time vectors, store the corresponding text feature vectors and time vectors in the corresponding data nodes, and obtain the corresponding project sub-graphs; Combine the project sub-graphs corresponding to different sub-projects to obtain the corresponding project knowledge graph.

4. A method for optimizing project management big data according to claim 3, characterized in that The process of obtaining the dynamic sub-graph includes: Obtain the real-time project optimization information, and based on it, obtain the project stage where the required management project is at the current moment and the set of real-time tasks being executed; Based on the project stage and the set of real-time tasks, match the corresponding data nodes in the corresponding project knowledge graph, and use them as the path starting points, and recursively extract all other data nodes and edges directly or indirectly related to the project tasks based on them to obtain the corresponding connected sub-graphs; Obtain the time attributes corresponding to each project task in the set of real-time tasks, and perform chronological arrangement on the extracted data nodes based on them, and convert the static project sub-graph into a dynamic sub-graph with chronological characteristics.

5. A method for optimizing project management big data according to claim 4, characterized in that The process of performing image reasoning based on the obtained dynamic sub-graph includes: Input the obtained dynamic sub-graph into a pre-constructed graph neural network reasoning model to obtain the corresponding project expected information, where the project expected information includes task index prediction data and potential node risk information; Furthermore, based on the project knowledge graph, structurally organize the corresponding project expected information to obtain the corresponding project prediction information, where the project prediction information includes the predicted values of each key index, confidence intervals, risk point descriptions, risk levels, influence ranges, and corresponding time information.

6. A method for optimizing project management big data according to claim 5, characterized in that, The backbone network of the graph neural network reasoning model is a hierarchical neural network, which is composed of a graph neural network and a temporal neural network. The basic framework of the hierarchical neural network includes an input layer, a convolutional layer, a prediction layer, a detection layer, and an output layer; The input layer is used to receive the input dynamic sub-graph and perform image preprocessing on it; The convolutional layer is used to perform graph convolution operations on the dynamically sub-graphed image after preprocessing to obtain the corresponding node feature matrix; The prediction layer is used to read the historical task parameter sequences of each data node in the dynamic sub-graph, perform data prediction based on the historical task parameter sequences and the node feature matrix, obtain the predicted states corresponding to the data nodes at several future time nodes, perform state analysis on the corresponding predicted states to obtain the corresponding feature vectors, and obtain the predicted values of each key index of the project tasks corresponding to the corresponding data nodes based on the feature vectors; At the same time, use the Monte Carlo dropout method to predict the confidence intervals corresponding to the corresponding key indicators; The detection layer respectively performs anomaly scoring on the corresponding data nodes and the edges corresponding to the nodes based on the feature vectors in the prediction layer, and determines whether the corresponding data nodes are abnormal nodes based on them. If they are abnormal nodes, perform risk level classification based on the anomaly scores to obtain the corresponding node risk levels; The output layer is used to receive the predicted values of key indicators, the anomaly scores and risk levels of each data node, map them into a pre-set data space, and obtain corresponding project prediction information.

7. A method for optimizing project management big data according to claim 6, characterized in that, The process of determining whether there is an abnormal risk in the project to be managed based on the project prediction information includes: Based on the obtained project prediction information, calculate the deviation between the key indicator parameters corresponding to each project task in the corresponding project prediction information and the corresponding confidence interval, obtain the parameter deviation degree of each project task, and conduct a comprehensive anomaly assessment of the project to be managed based on it to obtain the corresponding project risk score; Set a risk threshold, compare the obtained project risk score with the corresponding risk threshold, and determine whether the project risk is in a controllable state based on it. If it is in a controllable state, continue to monitor the project; if it is not in a controllable state, generate corresponding risk warning information.

8. A project management big data optimization method according to claim 7, characterized in that, The process of the set of alternative coping strategies includes: When a risk warning information is identified, obtain the data nodes with risks in the dynamic subgraph and mark them as risk nodes; Based on the project prediction information, obtain the risk types corresponding to the corresponding risk nodes and the triggering probabilities of each risk type; Based on the project prediction information, obtain the risk feature vectors corresponding to the project tasks of the corresponding risk nodes, and combine the task parameters and risk feature vectors of the project tasks to obtain several similar historical project task cases from the historical project database, and use the risk coping measures of the corresponding historical project task cases as reference coping strategies; Obtain the task differences between the current project task and the historical project tasks, and adjust the corresponding reference coping strategies based on them to obtain the corresponding set of alternative coping strategies.

9. A method for optimizing project management big data according to claim 8, characterized in that, The process of obtaining the optimal coping strategy includes: Based on the dynamic subgraph, obtain the dependency relationship between the corresponding risk nodes and other data nodes, and the influence intensity of other data nodes on the corresponding risk nodes under the corresponding risk types; Obtain the influence intensity of other data nodes on the corresponding risk nodes, and combine the obtained risk types and the set of alternative coping strategies to construct a corresponding project selection optimization model; and define the objective function of the corresponding project selection optimization model based on the pre-set constraint conditions; Obtain the task execution scenario, and construct several expected task execution scenarios based on the uncertain factors in the task execution scenario; obtain the strategy stability indicators corresponding to the alternative coping strategies based on the expected task execution scenarios; Obtain the management preferences of the corresponding staff for the project to be managed and the strategy stability indicators in each expected task execution scenario, and combine the particle swarm optimization algorithm to solve the corresponding objective function to obtain the corresponding optimal coping strategy.

10. A project management big data optimization method according to claim 9, characterized in that The process of real-time monitoring of the execution process of the corresponding optimal coping strategy and information feedback based on the monitoring results includes: Build a real-time monitoring mechanism to dynamically track the process of the corresponding optimal response strategy, collect in real time the key indicators of the corresponding project tasks during the execution of the project's optimal response strategy, and compare them with the pre-set expected goals. If the expected goals are not met, collect and preprocess the project management information of the corresponding management project in real time, update the dynamic subgraph based on it, and optimize the currently executing optimal response strategy based on the updated dynamic subgraph; at the same time, feedback the strategy execution results to the project knowledge graph to update the corresponding data nodes.

Images