Government affair service process optimization method and system based on artificial intelligence

Abstract

The application discloses a government affair service process optimization method and system based on artificial intelligence, relates to the field of government affair service digitization, and comprises the following steps: constructing a government affair knowledge graph for storing static knowledge in the field of government affairs; constructing a government affair reason graph of an event causal chain based on the government affair knowledge graph, integrating the government affair reason graph into a process digital twin, and forming a government affair reason graph enhanced digital twin; based on historical successful government affair process cases, pre-training a reinforcement learning agent by using the government affair reason graph to obtain an initial process optimization strategy; and in the government affair reason graph enhanced digital twin, the reinforcement learning agent guides an exploration process according to the causal correlation relationship revealed by the government affair reason graph based on the initial process optimization strategy. The application deploys the verified strategy to a real environment for optimization, and dynamically updates the enhanced digital twin based on running feedback, so that closed-loop optimization and continuous self-evolution of the government affair service process are realized.

Description

Technical Field

[0001] This invention relates to the field of digital government services, and in particular to a method and system for optimizing government service processes based on artificial intelligence. Background Technology

[0002] In the field of digital government services, AI-based process optimization technology has become a research hotspot. Existing technical solutions typically employ a combination of reinforcement learning and digital twins. By establishing process models in a virtual environment, intelligent agents can explore optimization strategies through trial and error. To improve efficiency, some advanced solutions introduce imitation learning, using historical successful case data to pre-train the intelligent agent in order to quickly obtain high-performance process optimization strategies. This data-driven approach has shown potential in improving process efficiency.

[0003] Existing technologies mainly rely on learning statistical correlations from historical data, lacking an understanding and modeling of the inherent causal logic between process steps. While the resulting strategies may perform well in terms of efficiency indicators, the decision-making logic may deviate from business norms, posing compliance risks. The agent may learn to shorten the time by skipping certain non-critical verifications. Although no problems are recorded in historical data, this violates rigid policy constraints. The optimization objective function is difficult to incorporate the causal common sense of the domain, and the compliance of the strategy cannot be guaranteed. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an artificial intelligence-based method for optimizing government service processes, which solves the core problem that existing technologies may lead to compliance risks in decision-making due to a lack of causal understanding.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for optimizing government service processes based on artificial intelligence, which includes constructing a government knowledge graph that stores static knowledge in the government domain; Based on the aforementioned government knowledge graph, a government affairs logic graph of event causal chains is constructed, and the government affairs logic graph is integrated into the process digital twin to form a government affairs logic graph enhanced digital twin; Based on successful historical government process cases, the government process graph is used to pre-train the reinforcement learning agent to obtain the initial process optimization strategy. In the digital twin augmented with a government affairs logic graph, the reinforcement learning agent, based on the initial process optimization strategy, guides the exploration process according to the causal relationships revealed by the government affairs logic graph, and generates the target process optimization strategy through simulation iteration. Using the political affairs graph, we can provide a causal explanation for the decision-making logic of the target process optimization strategy and trace the root causes of process anomalies monitored in the real political affairs operation environment. The target process optimization strategy, verified by causal explanation, is deployed to the real government operation environment to optimize the process. Based on the performance monitoring data and root cause tracing results, the digital twin of the government affairs logic graph is updated and fed back.

[0007] As a preferred embodiment of the AI-based government service process optimization method of the present invention, the following steps are included in constructing a government knowledge graph that stores static knowledge in the government domain: The system obtains heterogeneous data from multiple sources from government data sources and cleans it. Based on the cleaned heterogeneous data, it uses natural language processing technology to identify and extract entities in the government domain. Among the identified entities in the government domain, natural language processing technology is used to identify and extract the static relationships between them. The identified entities in the government domain and the static relationships between them are stored in a graph database in the form of nodes and edges, forming a government knowledge graph that stores static knowledge in the government domain.

[0008] As a preferred embodiment of the AI-based government service process optimization method of the present invention, the following steps are included: Constructing a government affairs logic graph based on the government knowledge graph to establish an event causal chain: The government knowledge graph is analyzed to identify entities and relationships that serve as dynamic event elements, and these are instantiated as event nodes. Based on historical process logs and business rules, we mine the causal logic relationships between event nodes, combine the causal logic relationships with event nodes, and construct a government affairs logic graph of event causal chains in the form of event nodes and causal logic relationship edges.

[0009] As a preferred embodiment of the AI-based government service process optimization method of the present invention, the method involves integrating a government affairs process graph into a process digital twin to form a government affairs process graph-enhanced digital twin, comprising the following steps: A digital twin of processes is constructed based on historical government process data, and data mapping rules are established between the government affairs logic graph and the digital twin of processes. The causal reasoning logic of the government affairs logic graph is embedded into the process digital twin, and the process digital twin is enhanced into an enhanced digital twin of the government affairs logic graph.

[0010] As a preferred embodiment of the AI-based government service process optimization method of the present invention, the method includes the following steps: based on historical successful government process cases, a reinforcement learning agent is pre-trained using a government affairs logic graph to obtain an initial process optimization strategy. Successful historical government process cases are selected from the historical records of government processes. Based on the government affairs graph, the corresponding event sequences are analyzed from the successful historical government process cases to form event causal chains. The event causal chain is transformed into connection weights of the reinforcement learning agent policy network through imitation learning algorithm, and an initial weight matrix is ​​generated. The initial weight matrix is ​​used to configure the parameters of the reinforcement learning agent's policy network to complete the initial training of the policy network and obtain the initial optimized policy defined by the initial weight matrix.

[0011] As a preferred embodiment of the AI-based government service process optimization method of the present invention, the following steps are included: In the government affairs logic graph-enhanced digital twin, the reinforcement learning agent, based on the initial process optimization strategy, guides the exploration process according to the causal relationships revealed by the government affairs logic graph, and generates the target process optimization strategy through simulation iteration: The initial weight matrix corresponding to the initial optimization strategy is loaded into the policy network of the reinforcement learning agent. In the simulation environment of the digital twin with enhanced event graph, the reinforcement learning agent dynamically adjusts its action exploration space based on the causal relationships revealed by the event graph. Based on the dynamically adjusted action exploration space, the reinforcement learning agent performs multi-round simulation iteration optimization strategies in the event graph-enhanced digital twin.

[0012] As a preferred embodiment of the AI-based government service process optimization method of the present invention, the method includes: using a government affairs logic graph to perform causal explanation of the decision logic of the target process optimization strategy, and tracing the root causes of process anomalies monitored in the real government operation environment, comprising the following steps: Based on the updated policy network weight matrix, the key decision sequence of the target process optimization strategy is simulated and reproduced in the event graph-enhanced digital twin. By utilizing key decision sequences, generate causal explanation reports describing the basis for decision-making; Map process anomalies detected in real government operations to corresponding abnormal event nodes in the government affairs logic graph; Based on the event causal chain of the government affairs graph, reverse causal reasoning is performed on the abnormal event nodes identified by the mapping of abnormal events in the process to locate the root cause event that caused the abnormal event.

[0013] As a preferred embodiment of the AI-based government service process optimization method of the present invention, the following steps are included: Deploying the target process optimization strategy, verified by causal explanation, to the real government operation environment to perform process optimization: The updated strategy network weight matrix corresponding to the target process optimization strategy verified by causal explanation is converted into a workflow rule executable in the real government operation environment; Workflow rules are deployed to the process execution engine in the real government operation environment. The process execution engine in the real government operation environment automatically processes newly arriving government process instances and performs process optimization based on the deployed workflow rules. During the process optimization, real-time data on the operational status of government process instances is collected to monitor the optimization effect.

[0014] As a preferred embodiment of the AI-based government service process optimization method described in this invention, the following steps are included: updating and providing feedback on the enhanced digital twin of the government affairs process graph based on performance monitoring data and root cause analysis results: By analyzing the performance monitoring data and root cause analysis results obtained from the real government operation environment, deviations were identified from the predictions of the enhanced digital twin of the government affairs graph, and the event causal chain in the government affairs graph was calibrated. Based on the new patterns reflected in the calibrated administrative affairs graph and performance monitoring data, update the simulation parameters of the process digital twin; The calibrated administrative affairs graph is re-integrated with the simulation parameters of the updated process digital twin to form an updated administrative affairs graph enhanced digital twin.

[0015] Secondly, the present invention provides an artificial intelligence-based government service process optimization system, including a government knowledge graph construction module for constructing a government knowledge graph that stores static knowledge in the government domain; The construction module, based on the aforementioned government knowledge graph, constructs a government affairs logic graph of event causal chains, and integrates the government affairs logic graph into the process digital twin to form a government affairs logic graph enhanced digital twin; The training module, based on historical successful government process cases, uses a government affairs logic graph to pre-train the reinforcement learning agent on the strategy to obtain the initial process optimization strategy. In the optimization module, within the enhanced digital twin of the government affairs logic graph, the reinforcement learning agent, based on the initial process optimization strategy, guides the exploration process according to the causal relationships revealed by the government affairs logic graph, and generates the target process optimization strategy through simulation iteration. The anomaly diagnosis module uses the government affairs logic graph to explain the causal logic of the decision-making logic of the target process optimization strategy and to trace the root causes of process anomalies monitored in the real government affairs operation environment. The update module deploys the target process optimization strategy, verified by causal explanation, to the real government operation environment to optimize the process. Based on the performance monitoring data and root cause tracing results, it updates and provides feedback to the enhanced digital twin of the government affairs logic graph.

[0016] The beneficial effects of this invention are as follows: By constructing a government affairs knowledge graph and a government affairs event logic graph, the causal logic of events is integrated into the process digital twin to form an enhanced digital twin. The government affairs event logic graph is used to pre-train the reinforcement learning agent to generate an initial optimization strategy. In the enhanced twin, the exploration process is guided according to causal relationships. The target optimization strategy is generated through simulation iteration. The government affairs event logic graph is used to explain the causal logic of the strategy decision and to trace the root cause of process anomalies. The verified strategy is deployed to the real environment for optimization. The enhanced digital twin is dynamically updated based on operational feedback, thereby realizing the closed-loop optimization and continuous self-evolution of the government service process. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a method for optimizing government service processes based on artificial intelligence.

[0019] Figure 2 This is a schematic diagram of an AI-based government service process optimization system.

[0020] Figure 3 Build a flowchart for government knowledge graph.

[0021] Figure 4 Generate a flowchart for the initial process optimization strategy. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figures 1-4 This is one embodiment of the present invention, which provides a method for optimizing government service processes based on artificial intelligence, including the following steps: S1. Construct a government knowledge graph that stores static knowledge in the government affairs domain.

[0026] S1.1 Obtain multi-source heterogeneous data from government data sources and clean it. Based on the cleaned multi-source heterogeneous data, use natural language processing technology to identify and extract entities in the government domain.

[0027] Furthermore, the multi-source heterogeneous data obtained from government data sources is cleaned and then processed using named entity recognition methods in natural language processing. This process is not simply keyword matching, but relies on a language model pre-trained on government text corpora. This model can deeply understand the specific semantic context in government documents, regulations, and service guides. By analyzing the contextual context, grammatical structure, and specific expression habits in the government domain, it accurately identifies and extracts entities with clear government semantics.

[0028] S1.2. Among the identified government affairs entities, natural language processing technology is used to identify and extract the static relationships between them.

[0029] Furthermore, based on the identified entities in the government domain, relation extraction methods from natural language processing are used to analyze sentence fragments containing two or more target government domain entities. Through techniques such as semantic role labeling and dependency parsing, specific static relationship categories between entities are determined. For example, when processing the text that a company needs to submit its articles of association to apply for a business license, the relation extraction model needs to accurately identify the relationship between the company and the business license as an application, and the relationship between the business license and the articles of association as required materials. This relation extraction based on deep semantic analysis goes beyond simple co-occurrence statistics, accurately capturing the inherent static associations between government domain entities that are clearly defined by laws or regulations, thereby ensuring that the constructed government knowledge graph can truly reflect the business rules and knowledge structure of the government domain.

[0030] S1.3. The identified government domain entities and the static relationships between them are stored in a graph database in the form of nodes and edges, forming a government domain knowledge graph that stores static knowledge in the government domain.

[0031] Furthermore, each identified government domain entity is instantiated as an independent node in the graph database, and each node is labeled with key attribute information extracted from the original data. The static relationships between the identified government domain entities are instantiated as directed edges connecting the corresponding node entities, and each edge is labeled with a clear relationship type. This transformation process integrates the static knowledge of the government domain, which was originally scattered in multi-source heterogeneous data, into an interconnected semantic network, namely the government knowledge graph. Stored in the graph database, it can efficiently support complex reasoning operations such as multi-hop relationship queries and path discovery. Thus, unstructured text information is transformed into structured knowledge that can be understood and computed by machines, providing strong knowledge support for upper-layer applications. The result is a government knowledge graph that stores static knowledge of the government domain in the form of nodes and edges in the graph database.

[0032] S2. Based on the aforementioned government knowledge graph, construct a government affairs logic graph of event causal chains.

[0033] S2.1 Analyze the government knowledge graph to identify entities and relationships that are dynamic event elements in the government knowledge graph, and instantiate them as event nodes.

[0034] Furthermore, semantic analysis of the government knowledge graph aims to identify dynamic and actionable entities and relationships within the graph and elevate them to event nodes. While the government knowledge graph stores a large number of static facts, not all entities are suitable for conversion into events and semantic filtering. For example, administrative licensing matters, processing departments, and application materials in the government knowledge graph are static concepts, but when they are associated with dynamic relationships representing actions or status changes such as submission, review, and approval, they constitute potential event elements.

[0035] Specifically, identifying these key dynamic relationships and instantiating the participating entities in the relationships into specific event nodes, and instantiating the relationship between the enterprise and the submission and establishment registration application into the enterprise submitting establishment registration application event, the approach of extracting dynamic event elements from static knowledge realizes the foundation for the transformation from a static knowledge network describing what is to a dynamic event sequence describing what happens.

[0036] S2.2 Based on historical process logs and business rules, mine the causal logic relationships between event nodes, combine the causal logic relationships with event nodes, and construct a government affairs logic graph of event causal chains in the form of event nodes and causal logic relationship edges.

[0037] Furthermore, discrete event nodes are connected into chains through causal logic relationships. By comprehensively utilizing the temporal information in historical process logs and the logical constraints contained in business rules, the order and conditional probability of events between event nodes can be statistically analyzed through historical process logs, thereby uncovering potential causal or sequential relationships. For example, event B is highly likely to follow event A. Business rules provide strict logical constraints, such as explicitly stipulating that a substantive review can only be triggered after the material pre-review event has occurred. By integrating these two types of information, edges with causal semantics can be established between event nodes. For example, defining incomplete event materials as the reason for the event pre-review failure, and combining event nodes with these causal logic relationship edges, an event causal chain that can express dynamic logic such as "because A occurred, B occurred" is constructed. The generated government affairs logic graph not only contains event sequences, but more importantly, reveals the inherent causal driving mechanism between events, giving the graph logical reasoning capabilities. The resulting government affairs logic graph is a causal chain of events constructed in the form of event nodes and causal logic relationship edges.

[0038] S3. Integrate the administrative affairs graph into the process digital twin to form an enhanced digital twin of the administrative affairs graph.

[0039] S3.1 Construct a digital twin of processes based on historical government process data, and establish data mapping rules between the government affairs logic map and the digital twin of processes.

[0040] Furthermore, the construction of the process digital twin is based on historical government process data, recording the state changes, resource consumption, and time consumption of process instances at each stage. The completed process digital twin can simulate the dynamic execution process of government processes, such as simulating random behaviors like user arrival and fluctuations in stage processing time. It establishes data mapping rules between the government affairs graph and the process digital twin, defining how event nodes and causal logic relationships in the government affairs graph correspond to the state variables and transition conditions within the process digital twin.

[0041] The event of material pre-review failure in the administrative affairs graph needs to be mapped to the review result state variable of the material pre-review stage in the process digital twin, which is set to failure, and triggers the corresponding state transition logic. The establishment of this mapping rule realizes the semantic alignment between static, declarative domain knowledge (administrative affairs graph) and dynamic, executable process model (process digital twin), ensuring that causal reasoning can be applied to the specific state of process simulation.

[0042] S3.2 Embed the causal reasoning logic of the government affairs logic graph into the process digital twin, and the process digital twin is enhanced into an enhanced digital twin of the government affairs logic graph.

[0043] Furthermore, the causal reasoning logic inherent in the government affairs logic graph is transformed into decision rules or state transition constraints within the process digital twin. When the process digital twin is in a certain state during simulation, it queries the government affairs logic graph in real time. Based on the currently occurring event nodes, it infers the set of events that may or should occur at the next moment by traversing the causal logic relationship edges. This inference result is then transformed into guidance or constraints for the subsequent simulation path of the process digital twin. When an incomplete materials event is triggered in the simulation process, the causal logic of the government affairs logic graph indicates that this is highly likely to lead to a pre-approval failure event. After receiving this inference result, the process digital twin can adjust the probability distribution of its internal state transitions or directly trigger the corresponding state change. This transforms the process digital twin from merely a random simulator based on historical statistical patterns into an intelligent simulation environment capable of causal inference and understanding business consequences. The process digital twin is thus enhanced into a government affairs logic graph-enhanced digital twin. S4. Based on successful historical government process cases, use the government affairs logic graph to pre-train the reinforcement learning agent to obtain the initial process optimization strategy.

[0044] S4.1 Select successful historical government process cases from the government process history records. Based on the government process logic graph, analyze the corresponding event sequences from the successful historical government process cases to form an event causal chain.

[0045] Furthermore, by filtering historical records of government processes and selecting successful cases of efficient and compliant completion as positive samples, and based on the government affairs process graph, the key to analyzing these cases lies in mapping the records in the cases to the corresponding event nodes in the government affairs process graph. Based on the actual execution order of the operations in the cases, the triggering chain between the event nodes is reconstructed, thereby forming a causal chain of events verified by successful experience.

[0046] Specifically, a successful business registration case is analyzed as a series of event nodes in the government affairs process graph, including the submission of application materials, the passing of preliminary review of materials, the completion of qualification verification, and the issuance of business license. The original success sequence and causal relationship are maintained, and the specific and instanced successful experience is abstracted into an event sequence pattern with universal guiding significance and conforming to the logic of the domain.

[0047] S4.2. Transform the event causal chain into connection weights of the reinforcement learning agent policy network through the imitation learning algorithm, and generate the initial weight matrix.

[0048] Furthermore, an imitation learning algorithm incorporating causal constraints is employed, with its objective function being a composite loss function consisting of two parts: the first part is the standard imitation learning loss, which minimizes the negative log-likelihood, thereby improving the policy network of the reinforcement learning agent. The state experienced by historical success cases Next, select the successful action. probability Maximize, thereby driving the behavioral distribution of the reinforcement learning agent's policy network. Approximating the reference behavior distribution The second part is the innovative causal consistency loss term, which calculates the loss of the reinforcement learning agent's policy network. One's own behavior relative to the political and administrative logic map The expected causal violation degree, derived from the causal violation degree function. Quantify it.

[0049] Specifically, through regularization coefficients By weighting and summing the two losses, the optimization process, while pursuing behavioral imitation, incorporates constraints on the causal rationality of the behavior. This guides the reinforcement learning agent's policy network not only to learn the appearance of historical successful cases, but also to deeply understand the causal logic behind their behavior. This generates an initial weight matrix that has both successful experience and inherently conforms to domain common sense, providing a high-starting-point, low-risk initial optimization strategy for subsequent reinforcement learning.

[0050] The expression for the composite loss function is: ; in, The initial weight matrix, In a given state Time-reinforcement learning agent policy network Select the action to perform. The conditional probability, As a reference to the mathematical expectation of the behavior distribution, For reference behavior distribution, For action, Given a state, To enhance the learning agent policy network, For the political affairs and rationale diagram, For the degree of causality violation, The regularization coefficient is used. S4.3. Use the initial weight matrix to configure the parameters of the reinforcement learning agent policy network to complete the initialization training of the policy network and obtain the initial optimization policy defined by the initial weight matrix.

[0051] Furthermore, the initial weight matrix is ​​assigned to all connection weights of the reinforcement learning agent's policy network to complete the initial training of the policy network. The knowledge obtained through joint optimization of imitation learning and causal constraints is solidified into the parameters of the reinforcement learning agent's policy network. Obtaining the initial optimization policy defined by the initial weight matrix means that before the reinforcement learning agent begins to interact with the environment and learn through trial and error, it already has a baseline policy that tends to choose historically successful actions and whose decision-making logic conforms to the causal relationship of the political and administrative logic graph. This greatly improves the starting performance of the reinforcement learning agent and effectively avoids the problems of low initial performance and blind exploration when reinforcement learning starts from scratch, because the initial policy has embedded causal logic.

[0052] S5. In the digital twin of the government affairs graph, the reinforcement learning agent guides the exploration process based on the initial process optimization strategy and the causal relationship revealed by the government affairs graph, and generates the target process optimization strategy through simulation iteration.

[0053] S5.1 Load the initial weight matrix corresponding to the initial optimization strategy into the policy network of the reinforcement learning agent. In the simulation environment of the digital twin with enhanced event graph, the reinforcement learning agent dynamically adjusts its action exploration space based on the causal relationships revealed by the event graph.

[0054] Furthermore, the initial weight matrix corresponding to the initial optimization strategy is loaded into the reinforcement learning agent's policy network, enabling the reinforcement learning agent to have preliminary decision-making capabilities based on historical successful experiences. In the simulation environment of the digital twin augmented with the government affairs graph, the reinforcement learning agent dynamically adjusts its action exploration space according to the causal relationships revealed by the government affairs graph. When the simulation process reaches a certain state, the reinforcement learning agent will query the government affairs graph in real time and predict the set of reasonable events that may occur in the next stage based on the currently occurring event nodes and causal logical relationships. The reinforcement learning agent's policy network will amplify the exploration probability of actions that may lead to favorable results and suppress the exploration probability of actions that are known to violate causal logic or lead to adverse consequences.

[0055] Specifically, when the preliminary review of the administrative affairs graph indicates that the on-site inspection is a necessary condition for the on-site inspection to be initiated, the reinforcement learning agent will reduce the probability of choosing to initiate the on-site inspection action if the preliminary review of the materials has not been passed. This transforms the exploration process of the reinforcement learning agent from a purely random trial and error to a directional intelligent search guided by domain knowledge, which greatly improves the exploration efficiency and effectively avoids the waste of resources on ineffective or high-risk actions.

[0056] S5.2 Based on the dynamically adjusted action exploration space, the reinforcement learning agent performs multi-round simulation iteration optimization strategies in the event graph-enhanced digital twin.

[0057] Furthermore, based on the dynamically adjusted action exploration space, the reinforcement learning agent performs multiple rounds of simulation iterations to optimize its strategy within the government affairs graph augmented digital twin. Each iteration includes four stages: policy execution, reward acquisition, policy evaluation, and policy update. The reinforcement learning agent selects actions within the causal logic-constrained exploration space, interacts with the government affairs graph augmented digital twin, and receives reward signals based on process efficiency indicators. Reinforcement learning algorithms such as policy gradients are used to update the weight parameters of the reinforcement learning agent's policy network based on cumulative rewards. The government affairs graph augmented digital twin provides a safe, controllable, and semantically rich simulation environment, enabling the reinforcement learning agent to undergo extensive trial-and-error learning within it, which would be costly or risky in the real world.

[0058] Specifically, through multiple iterations, the reinforcement learning agent policy network not only learns how to obtain high rewards, but also internalizes the causal laws contained in the government affairs logic graph. Finally, it converges to a target process optimization strategy that is both efficient and in line with business logic. By combining causal knowledge guidance with data-driven optimization, a high-quality target process optimization strategy that surpasses the initial strategy and simple historical experience is generated.

[0059] S6. Use the government affairs logic graph to explain the causal logic of the decision-making logic of the target process optimization strategy, and trace the root cause of process anomalies monitored in the real government affairs operation environment.

[0060] S6.1 Based on the updated policy network weight matrix, simulate and reproduce the key decision sequence of the target process optimization strategy in the event graph-enhanced digital twin.

[0061] Furthermore, the updated policy network weight matrix corresponding to the trained target process optimization strategy is loaded into the reinforcement learning agent policy network, and the simulation is run in the government affairs logic graph augmented digital twin. By inputting the initial state, the reinforcement learning agent policy network is driven to make decisions, thereby completely reproducing the complete execution path of the target process optimization strategy in the virtual environment and recording the key decision sequence from the initial state to the final state.

[0062] Specifically, the enhanced digital twin of the political affairs graph provides a high-fidelity, repeatable simulation environment that ensures consistent conditions for each simulation run, thereby stably acquiring the decision trajectory of the strategy. It transforms the neural network weight matrix, which is difficult to understand intuitively, into a specific and traceable state-action sequence, providing a clear analytical object for subsequent causal explanation and realizing the transformation from a parameterized model to an interpretable behavioral sequence.

[0063] S6.2 Utilize the key decision sequence to generate a causal explanation report describing the basis for the decisions.

[0064] Furthermore, for key decision sequences, causal relationship analysis is conducted using a political affairs graph. For each decision point in the sequence, i.e. the transition from state to action, the event nodes and causal logic edges related to that state and action are queried in the political affairs graph, thereby explaining the causal rationality of the chosen action.

[0065] Specifically, when the key decision sequence shows that the parallel approval action has been selected in the material pre-review stage, the analysis process will associate the logic that the material completeness event in the government affairs logic graph is the reason for initiating the parallel approval event. This will generate a natural language description that parallel approval is initiated to improve efficiency because the material pre-review has been passed, and summarize it into a causal explanation report. This will link the decision logic of the strategy with the generally accepted causal common sense in the government affairs logic graph, thereby enhancing the credibility and acceptability of the optimization strategy.

[0066] S6.3 Map process anomalies detected in the real government affairs operation environment to corresponding abnormal event nodes in the government affairs logic graph.

[0067] Furthermore, feature extraction and semantic analysis are performed on process anomalies detected in real-world government operations. These anomalies are then matched with predefined event node types in the government affairs graph. Process anomalies typically manifest as a surge in processing time, failure of processing results, or abnormal resource consumption. These anomaly features need to be mapped to anomaly event nodes with clear semantics in the government affairs graph, such as timeout events in the preliminary review of materials, final approval rejection events, and server resource exhaustion events. The established database of correspondences between anomaly features and event node types elevates anomalies in numerical indicators observed in the real world to anomalies with business semantics that the government affairs graph can understand.

[0068] S6.4 Based on the event causal chain of the government affairs graph, reverse causal reasoning is performed on the abnormal event nodes identified by the process abnormal event mapping to locate the root cause event that caused the abnormal event.

[0069] Furthermore, starting from the abnormal event nodes identified by the mapping, a reverse traversal and reasoning is performed on the event causal chain of the political affairs graph. Along the reverse direction of the causal logical relationship edge, upstream cause event nodes that may have caused the abnormal event are investigated one by one, and the probability or conditional probability of each cause event node is evaluated until the most fundamental, usually the most upstream triggering factor, i.e. the root cause event, is located.

[0070] Specifically, for example, in the event of an abnormal event being rejected in the final approval process, reverse reasoning may reveal that the direct cause was the failure of the on-site inspection, and the root cause of the failure of the on-site inspection can be traced back to the applicant's insufficient preparation or misunderstanding of the inspection standards. It utilizes the structured causal knowledge of the political affairs graph to achieve automated and systematic tracing from the surface abnormality to the deep root cause, avoiding governance that only stays at the level of surface symptoms, thus providing a clear target for precise optimization and rectification.

[0071] S7. Deploy the target process optimization strategy, which has been verified by causal explanation, to the real government operation environment to perform process optimization.

[0072] S7.1. Convert the updated strategy network weight matrix corresponding to the target process optimization strategy verified by causal explanation into a workflow rule executable in a real government operation environment.

[0073] Furthermore, the decision logic contained in the updated strategy network weight matrix corresponding to the target process optimization strategy verified by causal explanation is transformed into executable workflow rules in the real government operation environment. The input-output mapping relationship of the strategy network weight matrix is ​​analyzed, and the decision mode of the strategy network to select high-probability actions in a specific state is mapped into business rules with clear condition judgments and execution paths.

[0074] Specifically, for example, the strategy network weight matrix learns that when the application materials are complete and the industry is in the encouraged category, the green channel action is preferred. This decision-making pattern is transformed into a workflow rule. The condition part is defined as the application materials being complete and the industry type belonging to the encouraged category. The execution path is defined as routing the process instance to the green channel processing queue. It realizes a smooth transition from a data-driven, parameterized intelligent decision-making model to a rule-based workflow definition that can be directly interpreted and executed by existing government information systems, ensuring that the optimization strategy can be implemented in a low-risk and highly compatible manner.

[0075] S7.2 Deploy workflow rules to the process execution engine of the real government affairs operation environment. The process execution engine in the real government affairs operation environment automatically processes newly arrived government affairs process instances and performs process optimization based on the deployed workflow rules.

[0076] Furthermore, the workflow rules obtained through transformation are deployed to the process execution engine of the real government operation environment. After loading the new workflow rules, the process execution engine will automatically route and process the newly arriving government process instances according to the condition logic defined in the rules. When a new enterprise establishment registration application enters the system, the process execution engine will automatically verify its material status and industry type. If it meets the rule conditions, it will automatically allocate it to the green channel, thereby realizing automated process optimization without human intervention. It seamlessly embeds the optimization strategy verified by simulation and causal interpretation into the actual business operating system, so that the optimization results can directly and in real time affect a large number of government process instances, improve the overall processing efficiency and consistency, and reduce the error risk caused by improper human operation.

[0077] S7.3 During the process optimization, real-time data collection of the operational status of government process instances is used to monitor the optimization effect.

[0078] Furthermore, after the workflow rules take effect, the entire lifecycle of government process instances processed in the real government operation environment is monitored, and their operational status data is collected in real time. This data includes key performance indicators such as the time consumed at each stage, the workflow path, resource consumption, and the final result. The collected data is continuously aggregated and used to calculate various performance indicators to objectively evaluate the actual optimization effect brought about by the workflow rule changes. It establishes a feedback channel based on real operational data, enabling the actual value of the optimization strategy to be quantitatively evaluated, and ensuring that the entire optimization method can form a complete closed loop from virtual simulation to real deployment and then to effect verification.

[0079] S8. Update and provide feedback on the enhanced digital twin of the government affairs management graph based on performance monitoring data and root cause analysis results.

[0080] S8.1 Analyze the performance monitoring data and root cause analysis results obtained from the real government affairs operation environment, identify the deviations from the predictions of the enhanced digital twin of the government affairs process graph, and calibrate the event causal chain in the government affairs process graph.

[0081] Furthermore, a comparative analysis was conducted on the performance monitoring data obtained from the real government operation environment and the root cause analysis results. The focus was on identifying links or event sequences where there were significant differences between the predictions of the enhanced digital twin of the government affairs graph and the actual operation results. The performance monitoring data showed that the actual average time spent in the material pre-review stage was much higher than the simulation prediction value of the enhanced digital twin of the government affairs graph. At the same time, the root cause analysis results indicated that the delay in the material pre-review stage was mainly due to the inconsistent interpretation of new policies, a factor that was not depicted in the original government affairs graph.

[0082] Specifically, based on this deviation analysis, the causal chain of events in the administrative affairs graph is calibrated. A new causal logical relationship edge is added between the material pre-review event node and the pre-review duration extension event node. The event of inconsistent interpretation of new policies is used as the trigger condition for this causal relationship. It uses real-world operational feedback to continuously verify and correct the completeness and accuracy of the causal chain of events in the administrative affairs graph, enabling the domain knowledge model to dynamically adapt to policy changes and business evolution, thereby maintaining its reliability as a basis for simulation and decision-making.

[0083] S8.2. Update the simulation parameters of the process digital twin based on the new patterns reflected in the calibrated administrative affairs graph and performance monitoring data.

[0084] Furthermore, based on the new patterns reflected in the calibrated administrative affairs graph and performance monitoring data, the simulation parameters of the process digital twin are updated. The calibrated administrative affairs graph clarifies that inconsistencies in the interpretation of new policies will lead to extended processing time in the material pre-review stage. Combined with the statistical distribution of the actual time spent in this stage in the performance monitoring data, the parameters of the service time probability distribution model corresponding to the material pre-review activity in the process digital twin need to be refitted. The parameters of the simulated resource scheduling strategy in the process digital twin may also need to be adjusted according to the new resource consumption pattern. This ensures that the dynamic behavior model of the process digital twin can accurately reflect the latest characteristics of the real environment, keeping the simulation environment and the real environment synchronized in statistical behavior, and providing a high-fidelity foundation for subsequent simulation-based analysis, prediction, and optimization.

[0085] S8.3. Reintegrate the simulation parameters of the calibrated administrative affairs graph with the updated process digital twin to form an updated administrative affairs graph enhanced digital twin.

[0086] Furthermore, the calibrated administrative affairs graph is re-integrated with the simulation parameters of the updated process digital twin to form an enhanced digital twin of the updated administrative affairs graph. The mapping and driving relationship between the updated event causal logic in the calibrated administrative affairs graph and the adjusted simulation components in the updated process digital twin is re-established. Newly added policy interpretation inconsistencies and their causal relationships need to be mapped to rules that trigger specific delay logic in the process digital twin. This completes the closed loop from real-world observation feedback to virtual model iterative optimization, making the enhanced digital twin of the administrative affairs graph a living entity that can continuously evolve with changes in the business environment. The simulation prediction and decision support capabilities are continuously enhanced, providing a dynamically adaptive intelligent foundation for the continuous optimization of administrative processes.

[0087] This embodiment also provides an artificial intelligence-based government service process optimization system, including: a government knowledge graph construction module, which constructs a government knowledge graph that stores static knowledge in the government domain; The construction module, based on the aforementioned government knowledge graph, constructs a government affairs logic graph of event causal chains, and integrates the government affairs logic graph into the process digital twin to form a government affairs logic graph enhanced digital twin; The training module, based on historical successful government process cases, uses a government affairs logic graph to pre-train the reinforcement learning agent to obtain initial process optimization strategies. In the optimization module, within the enhanced digital twin of the government affairs logic graph, the reinforcement learning agent, based on the initial process optimization strategy, guides the exploration process according to the causal relationships revealed by the government affairs logic graph, and generates the target process optimization strategy through simulation iteration. The anomaly diagnosis module uses the government affairs logic graph to explain the causal logic of the decision-making logic of the target process optimization strategy and to trace the root causes of process anomalies monitored in the real government affairs operation environment. The update module deploys the target process optimization strategy, verified by causal explanation, to the real government operation environment to optimize the process. Based on the performance monitoring data and root cause tracing results, it updates and provides feedback to the enhanced digital twin of the government affairs logic graph.

[0088] This embodiment also provides a computer device applicable to the situation of the government service process optimization method based on artificial intelligence, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the government service process optimization method based on artificial intelligence as proposed in the above embodiment.

[0089] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0090] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the method for optimizing government service processes based on artificial intelligence as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0091] In summary, this invention constructs a government affairs knowledge graph and a government affairs event logic graph, integrating the causal logic of events into a process digital twin to form an enhanced digital twin. The government affairs event logic graph is used to pre-train the reinforcement learning agent to generate initial optimization strategies. Within the enhanced twin, the exploration process is guided by causal relationships, and target optimization strategies are generated through simulation iteration. The government affairs event logic graph is used to provide causal explanations for strategy decision-making logic and to trace the root causes of process anomalies. The validated strategies are deployed to a real environment for optimization, and the enhanced digital twin is dynamically updated based on operational feedback. This achieves closed-loop optimization and continuous self-evolution of government service processes.

[0092] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for optimizing government service processes based on artificial intelligence, characterized by: This includes constructing a government knowledge graph that stores static knowledge in the government affairs domain; Based on the aforementioned government knowledge graph, a government affairs logic graph of event causal chains is constructed, and the government affairs logic graph is integrated into the process digital twin to form a government affairs logic graph enhanced digital twin; The government knowledge graph is analyzed to identify entities and relationships that serve as dynamic event elements, and these are instantiated as event nodes. Based on historical process logs and business rules, we mine the causal logic relationships between event nodes, combine the causal logic relationships with event nodes, and construct a government affairs logic graph of event causal chains in the form of event nodes and causal logic relationship edges. A digital twin of processes is constructed based on historical government affairs process data, and data mapping rules are established between the government affairs logic graph and the digital twin of processes. The causal reasoning logic of the government affairs logic graph is embedded into the process digital twin, and the process digital twin is enhanced into a government affairs logic graph enhanced digital twin. Based on successful historical government process cases, the government process graph is used to pre-train the reinforcement learning agent to obtain the initial process optimization strategy. Successful historical government process cases are selected from the historical records of government processes. Based on the government affairs graph, the corresponding event sequences are analyzed from the successful historical government process cases to form event causal chains. The event causal chain is transformed into connection weights of the reinforcement learning agent policy network through imitation learning algorithm, and an initial weight matrix is ​​generated. The initial weight matrix is ​​used to configure the parameters of the reinforcement learning agent's policy network to complete the initial training of the policy network and obtain the initial optimization policy defined by the initial weight matrix. In the digital twin augmented with a government affairs logic graph, the reinforcement learning agent, based on the initial process optimization strategy, guides the exploration process according to the causal relationships revealed by the government affairs logic graph, and generates the target process optimization strategy through simulation iteration. The initial weight matrix corresponding to the initial optimization strategy is loaded into the policy network of the reinforcement learning agent. In the simulation environment of the digital twin with enhanced event graph, the reinforcement learning agent dynamically adjusts its action exploration space based on the causal relationships revealed by the event graph. Based on the dynamically adjusted action exploration space, the reinforcement learning agent performs multi-round simulation iteration optimization strategies in the event graph-enhanced digital twin. Using the political affairs graph, we can provide a causal explanation for the decision-making logic of the target process optimization strategy and trace the root causes of process anomalies monitored in the real political affairs operation environment. The target process optimization strategy, verified by causal explanation, is deployed to the real government operation environment to optimize the process. Based on the performance monitoring data and root cause tracing results, the digital twin of the government affairs logic graph is updated and fed back.

2. The method for optimizing government service processes based on artificial intelligence as described in claim 1, characterized in that: Constructing a government knowledge graph that stores static knowledge in the government domain includes the following steps: The system obtains heterogeneous data from multiple sources from government data sources and cleans it. Based on the cleaned heterogeneous data, it uses natural language processing technology to identify and extract entities in the government domain. Among the identified entities in the government domain, natural language processing technology is used to identify and extract the static relationships between them. The identified entities in the government domain and the static relationships between them are stored in a graph database in the form of nodes and edges, forming a government knowledge graph that stores static knowledge in the government domain.

3. The method for optimizing government service processes based on artificial intelligence as described in claim 1, characterized in that: Using a government affairs process graph, the decision-making logic of the target process optimization strategy is explained causally, and the root causes of process anomalies monitored in the real government affairs operation environment are traced, including the following steps: Based on the updated policy network weight matrix, the key decision sequence of the target process optimization strategy is simulated and reproduced in the event graph-enhanced digital twin. By utilizing key decision sequences, generate causal explanation reports describing the basis for decision-making; Map process anomalies detected in real government operations to corresponding abnormal event nodes in the government affairs logic graph; Based on the event causal chain of the government affairs graph, reverse causal reasoning is performed on the abnormal event nodes identified by the mapping of abnormal events in the process to locate the root cause event that caused the abnormal event.

4. The method for optimizing government service processes based on artificial intelligence as described in claim 1, characterized in that: Deploying the target process optimization strategy, validated by causal explanation, to the real-world government operational environment to perform process optimization includes the following steps: The updated strategy network weight matrix corresponding to the target process optimization strategy verified by causal explanation is converted into a workflow rule executable in the real government operation environment; Workflow rules are deployed to the process execution engine in the real government operation environment. The process execution engine in the real government operation environment automatically processes newly arriving government process instances and performs process optimization based on the deployed workflow rules. During the process optimization, real-time data on the operational status of government process instances is collected to monitor the optimization effect.

5. The method for optimizing government service processes based on artificial intelligence as described in claim 1, characterized in that: The digital twin of the government affairs management graph is updated and fed back based on performance monitoring data and root cause analysis results, including the following steps: By analyzing the performance monitoring data and root cause analysis results obtained from the real government operation environment, deviations were identified from the predictions of the enhanced digital twin of the government affairs graph, and the event causal chain in the government affairs graph was calibrated. Based on the new patterns reflected in the calibrated administrative affairs graph and performance monitoring data, update the simulation parameters of the process digital twin; The calibrated administrative affairs graph is re-integrated with the simulation parameters of the updated process digital twin to form an updated administrative affairs graph enhanced digital twin.

6. An AI-based government service process optimization system, based on the AI-based government service process optimization method according to any one of claims 1 to 5, characterized in that: This includes a government affairs knowledge graph construction module, which constructs a government affairs knowledge graph that stores static knowledge in the government affairs field; The construction module, based on the aforementioned government knowledge graph, constructs a government affairs logic graph of event causal chains, and integrates the government affairs logic graph into the process digital twin to form a government affairs logic graph enhanced digital twin; The training module, based on historical successful government process cases, uses a government affairs logic graph to pre-train the reinforcement learning agent on the strategy to obtain the initial process optimization strategy. In the optimization module, within the enhanced digital twin of the government affairs logic graph, the reinforcement learning agent, based on the initial process optimization strategy, guides the exploration process according to the causal relationships revealed by the government affairs logic graph, and generates the target process optimization strategy through simulation iteration. The anomaly diagnosis module uses the government affairs logic graph to explain the causal logic of the decision-making logic of the target process optimization strategy and to trace the root causes of process anomalies monitored in the real government affairs operation environment. The update module deploys the target process optimization strategy, verified by causal explanation, to the real government operation environment to optimize the process. Based on the performance monitoring data and root cause tracing results, it updates and provides feedback to the enhanced digital twin of the government affairs logic graph.

Images