A knowledge graph-based intelligent analysis system for trade union policies and its control method

Abstract

The present application discloses a knowledge graph-based intelligent analysis system for trade union policies and a control method thereof, which relate to the field of trade union management. The system comprises: a current trade union policy collection module, which is used to collect information on implemented trade union policy documents within a target enterprise; a member information collection module, which is used to collect information on members in attendance and member inquiry information within the target enterprise; a relevant regulations collection module, which is used to collect relevant regulations information related to the trade union; an intelligent analysis module, which is used to generate an implemented policy knowledge graph based on information on implemented trade union policy documents, and also to generate a relevant regulations knowledge graph based on relevant regulations information; a future trade union policy prediction module, which is used to generate future trade union policy prediction information based on member inquiry information, information on members in attendance, implemented policy knowledge graph and relevant regulations knowledge graph; and an intelligent reply module, which is used to generate intelligent reply information based on member inquiry information, information on members in attendance and implemented policy knowledge graph.

Description

Technical Field

[0001] This specification relates to the field of trade union management. More specifically, this application relates to a knowledge graph-based intelligent analysis system for trade union policies and a control method thereof. Background Art

[0002] In traditional union policy management and information systems, information collection, processing, and analysis are often decentralized, relying on manual operations and unstructured data processing methods. This results in slow information updates and difficulty in quickly responding to member needs and policy changes. Furthermore, traditional systems often lack the ability to effectively integrate and visualize complex policy content and related regulations, resulting in inefficiencies and inaccuracies in policy analysis, forecasting, and responding to member inquiries.

[0003] Therefore, it is necessary to build a knowledge graph-based intelligent analysis system for union policies and its control method that is more automated and capable of automatic optimization. Summary of the Invention

[0004] The Summary of the Invention introduces a series of simplified concepts that will be further described in the Detailed Description of the Invention. The Summary of the Invention of this application is not intended to limit the key features and essential technical features of the claimed technical solution, nor is it intended to determine the scope of protection of the claimed technical solution.

[0005] First, this application proposes a knowledge graph-based intelligent analysis system for union policies, which includes:

[0006] The current union policy collection module is used to collect information on union policy documents that have been implemented in the target enterprise;

[0007] Member information collection module, used to collect information on current members and member enquiry information within the above-mentioned target enterprise;

[0008] Relevant regulations collection module, used to collect relevant regulations information related to the above-mentioned trade unions;

[0009] An intelligent parsing module is used to generate a knowledge graph of implemented policies based on the information of the implemented union policy documents, and is also used to generate a knowledge graph of relevant regulations based on the information of relevant regulations;

[0010] A future union policy prediction module is used to generate future union policy prediction information based on the member inquiry information, the current member information, the implemented policy knowledge graph, and the relevant regulations knowledge graph;

[0011] The intelligent reply module is used to generate intelligent reply information based on the above-mentioned member inquiry information, the above-mentioned member information, and the above-mentioned implemented policy knowledge graph.

[0012] In a second aspect, embodiments of the present application further provide a control method for the above-mentioned knowledge graph-based intelligent analysis system for union policies in the first aspect, comprising:

[0013] Collect information on the aforementioned implemented union policies and documents, information on current union members, and information on member inquiries within the aforementioned target enterprises;

[0014] Generate a knowledge graph of implemented policies based on the above-mentioned implemented union policy documents;

[0015] Generate a knowledge graph of relevant regulations based on relevant regulations information;

[0016] Generate the above-mentioned future union policy forecast information based on the above-mentioned member inquiry information, the above-mentioned member information, the above-mentioned implemented policy knowledge graph, and the above-mentioned relevant regulations knowledge graph;

[0017] The above-mentioned intelligent reply information is generated based on the above-mentioned member inquiry information, the above-mentioned member information, and the above-mentioned implemented policy knowledge graph.

[0018] In a feasible implementation, the above-mentioned generation of the implemented policy knowledge graph based on the implemented union policy document information includes:

[0019] Clean and standardize the above-mentioned implemented union policy document information to obtain union policy standard data, where the above-mentioned union policy document information includes union document information, union database information, and union meeting minutes information;

[0020] Applying natural language processing to the union policy data to obtain first key entity information, wherein the first key entity information includes policy name information, implementation date information, involved department information, and associated employee group information;

[0021] Performing a relation extraction operation on the above-mentioned union policy data using a relation extraction model to obtain first entity relationship information corresponding to the above-mentioned first key entity information;

[0022] The first node information is constructed based on the first key entity information, the first side information is constructed based on the first entity relationship information, and the implemented policy knowledge graph is generated based on the first node information and the first side information.

[0023] In a feasible implementation manner, the above method further includes:

[0024] Calculate the policy initial weight information based on the above-mentioned department information, the above-mentioned related employee group information and the query frequency information;

[0025] Identify the main group information of the above-mentioned currently implemented policy knowledge graph through the Louvain algorithm;

[0026] Conduct centrality analysis on the above-mentioned currently implemented policy knowledge graph to obtain core department information;

[0027] Based on the above main group information and the above core department information, the above initial weight information is modified to obtain the target weight information;

[0028] Based on the above target weight information, the above implemented policy knowledge graph is optimized to obtain an optimized implemented policy knowledge graph.

[0029] In a feasible implementation, the above-mentioned generation of the relevant regulations knowledge graph based on the relevant regulations information includes:

[0030] Clean and standardize the above-mentioned relevant regulatory information to obtain standard data of relevant regulatory information, where the above-mentioned relevant regulatory information includes legal and regulatory information, industry guidance information, and official announcement information;

[0031] Apply NER operation to the above-mentioned relevant regulation information standard data to obtain the second key entity information, wherein the above-mentioned second key entity information includes legal article number information, regulation name information, scope of application information, relevant department information and defined term information;

[0032] Performing a relationship extraction operation on the above-mentioned union policy data using a rule-based method to obtain second entity relationship information corresponding to the above-mentioned second key entity information;

[0033] The second node information is constructed based on the above-mentioned second key entity information, the second side information is constructed based on the above-mentioned second entity relationship information, and the above-mentioned relevant regulations knowledge graph is generated based on the above-mentioned second node information and the above-mentioned second side information, wherein the above-mentioned relevant regulations knowledge graph has different graph levels based on the entity types corresponding to the relevant regulations.

[0034] In a feasible implementation, the above-mentioned constructing the second node information according to the above-mentioned second key entity information, constructing the second side information according to the above-mentioned second entity relationship information, and generating the above-mentioned relevant prescribed knowledge graph according to the above-mentioned second node information and the above-mentioned second side information include:

[0035] Determining a node hierarchy according to the type corresponding to the second key entity information to construct the second node information, wherein the type corresponding to the second key entity information includes laws, regulations, guidelines, and clauses;

[0036] Constructing second side information based on the type corresponding to the second entity relationship information, wherein the types corresponding to the second entity relationship information include inclusion relationship, authorization relationship, application relationship, revision relationship, and reference relationship, and the second entity relationship information includes a top-down hierarchical relationship and a bottom-up reverse hierarchical relationship;

[0037] The above-mentioned relevant regulations knowledge graph is generated based on the above-mentioned second node information and the above-mentioned second edge information.

[0038] In a feasible implementation manner, the above-mentioned future union policy forecast information includes existing union policy abolition forecast information and new union policy forecast information;

[0039] The above-mentioned future union policy forecast information is generated based on the above-mentioned member inquiry information, the above-mentioned member information, the above-mentioned implemented policy knowledge graph and the above-mentioned relevant regulations knowledge graph, including:

[0040] Determine policy conflict information and policy missing information based on the above-mentioned implemented policy knowledge graph and the above-mentioned relevant regulations knowledge graph;

[0041] Determine the above-mentioned prediction information on the abolition of existing union policies based on the above-mentioned member inquiry information, the above-mentioned information on current members, and the above-mentioned policy conflict information;

[0042] The above-mentioned future union policy forecast information is determined based on the above-mentioned member inquiry information, the above-mentioned information on current members and the above-mentioned policy missing information.

[0043] In a feasible implementation, the above-mentioned determination of policy conflict information and policy missing information based on the above-mentioned implemented policy knowledge graph and the above-mentioned relevant regulations knowledge graph includes:

[0044] Traverse the above-mentioned implemented policy knowledge graph and the above-mentioned relevant regulations knowledge graph to extract the node information of all implemented policies and relevant regulations;

[0045] Determine the policy conflict information of the node information of the above-mentioned implementation policy and relevant regulations through natural language analysis based on preset rules;

[0046] The above-mentioned policy missing information is determined based on the difference analysis operation, and the node information of the above-mentioned implementation policy and related regulations is determined.

[0047] In a feasible implementation, the determining of the existing union policy abolition prediction information based on the member inquiry information, the current member information, and the policy conflict information includes:

[0048] Converting the member inquiry information, the member information at the meeting, and the policy conflict information into a first eigenvector;

[0049] Input the first eigenvector into a decision tree classifier, and use information gain and Gini impurity to determine the decision tree split point;

[0050] Based on the above decision tree split points, the prediction information of the abolition of existing union policies is determined.

[0051] In a feasible implementation manner, the above-mentioned determination of the above-mentioned future union policy forecast information based on the above-mentioned member inquiry information, the above-mentioned current member information and the above-mentioned policy missing information includes:

[0052] Classify the member inquiry information and the member information based on cluster analysis to obtain classified question and demand information;

[0053] Determine the correlation between member needs and policy gaps based on the Apriori algorithm according to the classified problem and demand information and the above policy gap information;

[0054] The above-mentioned union policy forecast information is generated based on the policy deficiency information that the correlation between member needs and policy deficiency is greater than the preset correlation.

[0055] In summary, traditional union policy management systems rely on manual operations for information collection and processing, which can easily lead to slow information updates and delayed responses. In the system proposed in the embodiments of this application, through automated modules such as the current union policy collection module and the member information collection module, information can be updated and processed in real time, greatly improving the efficiency and timeliness of information processing. Automated processing reduces human error and ensures data accuracy and update speed. The integration and visualization capabilities of policy content and related regulations in related technologies are limited, which affects the parsing and understanding of policies. This application uses an intelligent parsing module to generate a knowledge graph of implemented policies and a knowledge graph of related regulations. This not only helps union managers and members more clearly understand the policy content and their interrelationships, but also improves the transparency and accessibility of policies. Utilizing a future union policy prediction module, this system can predict future policy trends and areas that require adjustment based on real-time collected data and knowledge graphs. This prediction is not only based on historical data, but also incorporates current member needs and market changes, making policy adjustments more proactive and in line with actual needs. The intelligent reply module ensures more accurate and timely responses to member inquiries. By intelligently analyzing member inquiries and policy content, the system provides specific and clear answers, improving the member service experience and increasing member satisfaction and engagement with union work. By integrating and analyzing complex data, the system provides powerful decision-making support, helping union managers identify potential risks and opportunities and make more informed decisions. This approach reduces the risk of decisions based on intuition or incomplete information.

[0056] The present application proposes an intelligent analysis system for trade union policies based on knowledge graphs. Other advantages, objectives and features of the present application will be reflected in part through the following description, and will also be understood by technical personnel in this field through research and practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present description. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0058] Figure 1 A structural diagram of a knowledge graph-based intelligent analysis system for trade union policies provided in an embodiment of the present application;

[0059] Figure 2 A flow chart of a control method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0060] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the embodiments described are only some embodiments of the present application, not all embodiments.

[0061] See also Figure 1 , which is a structural diagram of a knowledge graph-based intelligent analysis system 10 for trade union policies provided in an embodiment of the present application, and may specifically include:

[0062] The current union policy collection module 101 is used to collect information on union policy documents that have been implemented in the target enterprise;

[0063] Member information collection module 102, used to collect information on current members and member inquiry information within the target enterprise;

[0064] The relevant regulations collection module 103 is used to collect relevant regulations information related to the above-mentioned trade union;

[0065] Intelligent parsing module 104, used to generate an implemented policy knowledge graph based on the implemented union policy document information, and also used to generate a relevant regulations knowledge graph based on relevant regulations information;

[0066] The future union policy prediction module 105 is used to generate future union policy prediction information based on the member inquiry information, the current member information, the implemented policy knowledge graph, and the relevant regulations knowledge graph;

[0067] The intelligent reply module 106 is used to generate intelligent reply information based on the above-mentioned member inquiry information, the above-mentioned member information, and the above-mentioned implemented policy knowledge graph.

[0068] For example, the main function of the current union policy collection module 101 is to collect information on union policy documents that have been implemented in the target enterprise. This information includes but is not limited to union regulations, policy documents, implementation details, etc., which are the specific content of union policy implementation.

[0069] The member information collection module 102 is responsible for collecting information on current members and member inquiry information within the target enterprise. Current member information refers to members' basic information, membership status, etc., while member inquiry information involves members' questions about union policies or specific content that needs to be answered.

[0070] The relevant regulations collection module 103 is used to collect regulations information related to the trade union, including other relevant policies, laws and regulations of the trade union, etc. These are important bases that need to be referred to when formulating or adjusting trade union policies.

[0071] Intelligent parsing module 104 has two main functions: first, generating a knowledge graph of implemented policies based on information from implemented union policy documents; and second, generating a knowledge graph of relevant regulations based on information about relevant regulations. A knowledge graph is a method for representing knowledge entities and their relationships in a graphical form, which helps the system deeply understand and analyze policy content and their interrelationships.

[0072] The future union policy prediction module 105 is used to generate future union policy prediction information. It analyzes member inquiry information, current member information, implemented policy knowledge graphs, and relevant regulations knowledge graphs to predict new policies or policy adjustments that the union may adopt.

[0073] The intelligent response module 106 generates intelligent responses based on member inquiries, information about current members, and the knowledge graph of implemented policies. This process uses intelligent analysis to provide accurate responses to member inquiries, aiming to answer members' questions and provide necessary policy explanations and guidance.

[0074] In summary, traditional union policy management systems rely on manual operations for information collection and processing, which can easily lead to slow information updates and delayed responses. In the system proposed in the embodiments of this application, through automated modules such as the current union policy collection module and the member information collection module, information can be updated and processed in real time, greatly improving the efficiency and timeliness of information processing. Automated processing reduces human error and ensures data accuracy and update speed. The integration and visualization capabilities of policy content and related regulations in related technologies are limited, which affects the parsing and understanding of policies. This application uses an intelligent parsing module to generate a knowledge graph of implemented policies and a knowledge graph of related regulations. This not only helps union managers and members more clearly understand the policy content and their interrelationships, but also improves the transparency and accessibility of policies. Utilizing a future union policy prediction module, this system can predict future policy trends and areas that require adjustment based on real-time collected data and knowledge graphs. This prediction is not only based on historical data, but also incorporates current member needs and market changes, making policy adjustments more proactive and in line with actual needs. The intelligent reply module ensures more accurate and timely responses to member inquiries. By intelligently analyzing member inquiries and policy content, the system provides specific and clear answers, improving the member service experience and increasing member satisfaction and engagement with union work. By integrating and analyzing complex data, the system provides powerful decision-making support, helping union managers identify potential risks and opportunities and make more informed decisions. This approach reduces the risk of decisions based on intuition or incomplete information.

[0075] Second, please refer to Figure 2 , this application also proposes a process diagram of a control method for the above-mentioned knowledge graph-based intelligent analysis system for union policies in the first aspect, including:

[0076] S110, collecting the information on the implemented trade union policy documents, the information on the current members, and the information on the members' inquiries within the target enterprise;

[0077] S120. Generate a knowledge graph of implemented policies based on the implemented union policy document information;

[0078] S130. Generate a knowledge graph of relevant regulations based on relevant regulations information;

[0079] S140. Generate the forecast information of future union policies based on the member inquiry information, the information of current members, the knowledge graph of implemented policies, and the knowledge graph of relevant regulations;

[0080] S150. Generate the intelligent reply information based on the member inquiry information, the member information in attendance, and the implemented policy knowledge graph.

[0081] For example, the current union policy collection module is responsible for collecting information on union policy documents implemented in the target enterprise, including various union policies, regulations, and other relevant documents that have taken effect and are being implemented in the enterprise.

[0082] The member information collection module is responsible for collecting information on current union members and member inquiries within the target enterprise. Current union member information refers to the identity information and activity participation of union members, while member inquiries are questions and concerns raised by members regarding current union policies.

[0083] The relevant regulations collection module is responsible for collecting regulations and information related to union policies. These regulations may come from higher-level union organizations or laws and regulations, and provide a basis for formulating or adjusting union policies.

[0084] The intelligent parsing module performs two main functions: generating a knowledge graph of implemented policies based on the information in implemented union policy documents, and converting the content in the policy documents into a graph format to clearly display the relationships and hierarchical structures between the various policies.

[0085] A knowledge graph of relevant regulations is generated based on the relevant regulations information, and the structure and connections of the relevant regulations are also presented in the form of a graph, which facilitates the system to understand and refer to these regulations for policy analysis and prediction.

[0086] The Future Union Policy Prediction module uses information from member inquiries, current member information, a knowledge graph of implemented policies, and a knowledge graph of relevant regulations. By integrating this data and graph information, it predicts potential future union policies or adjustments to existing policies. This forecast information helps union leaders plan their strategies in advance.

[0087] The Smart Response module utilizes collected member inquiry information, information about current members, and a knowledge graph of implemented policies to generate intelligent responses to specific member questions. This module analyzes member inquiries and current policies to intelligently generate answers, aiming to provide accurate and useful responses to inquiring members.

[0088] In some examples, the generated implemented policy knowledge graph based on the implemented union policy document information includes:

[0089] Clean and standardize the above-mentioned implemented union policy document information to obtain union policy standard data, where the above-mentioned union policy document information includes union document information, union database information, and union meeting minutes information;

[0090] Applying natural language processing to the union policy data to obtain first key entity information, wherein the first key entity information includes policy name information, implementation date information, involved department information, and associated employee group information;

[0091] Performing a relation extraction operation on the above-mentioned union policy data using a relation extraction model to obtain first entity relationship information corresponding to the above-mentioned first key entity information;

[0092] The first node information is constructed based on the first key entity information, the first side information is constructed based on the first entity relationship information, and the implemented policy knowledge graph is generated based on the first node information and the first side information.

[0093] For example, first, data from implemented union policy documents is cleansed and standardized. This process includes eliminating data errors, removing redundant information, and standardizing data formats. Implemented union policy documents include union documents, union databases, and union meeting minutes. These steps ensure that the data extracted from various policy documents is accurate and standardized, referred to as standardized union policy data.

[0094] Natural language processing was used to process the union policy standard data to extract key entity information. This entity information includes the policy name, implementation date, involved departments, and associated employee groups. These key entities are fundamental to understanding the policy's content and structure.

[0095] Next, a relation extraction model was used to perform relation extraction on the union policy data. This operation aims to identify the relationships between different entities, namely the first entity relationship information. Relationship extraction can clarify how the various policy entities are interconnected, such as which policies are implemented by specific departments or which policies specifically focus on a certain employee group.

[0096] Based on the first key entity information, first node information is constructed. Nodes represent policy entities in the graph, such as specific policy documents or policy implementation details. Based on the first entity relationship information, first edge information is constructed. Edges represent connecting lines between entities in the knowledge graph, demonstrating the relationships between them.

[0097] Finally, the first node information and the first edge information are used to generate a knowledge graph of implemented policies. This knowledge graph graphically displays the entire picture of union policies, clearly showing the connections and impacts between policies.

[0098] This detailed and systematic approach allows us to effectively extract key information from complex policy documents and construct a knowledge graph that helps us understand and analyze the structure and implementation logic of union policies. This not only improves the transparency of policy management but also provides a powerful support tool for policy formulation and implementation.

[0099] In some examples, the method further includes:

[0100] Calculate the policy initial weight information based on the above-mentioned department information, the above-mentioned related employee group information and the query frequency information;

[0101] Identify the main group information of the above-mentioned currently implemented policy knowledge graph through the Louvain algorithm;

[0102] Conduct centrality analysis on the above-mentioned currently implemented policy knowledge graph to obtain core department information;

[0103] Based on the above main group information and the above core department information, the above initial weight information is modified to obtain the target weight information;

[0104] Based on the above target weight information, the above implemented policy knowledge graph is optimized to obtain an optimized implemented policy knowledge graph.

[0105] For example, the initial weight of each policy is calculated based on the relevant departments, associated employee groups, and query frequency. This calculation takes into account the policy's scope (e.g., the departments and employee groups involved) and the frequency of queries or citations in actual applications, thereby determining the initial importance of each policy within the overall policy system.

[0106] We use the Louvain algorithm to perform community detection on the currently implemented policy knowledge graph to identify key groups within the graph. The Louvain algorithm is a community discovery algorithm based on modularity optimization. It effectively identifies key groups or communities within the policy knowledge graph. These groups are typically composed of closely interconnected policy nodes.

[0107] Perform centrality analysis on the currently implemented policy knowledge graph to help identify core department information. Centrality analysis evaluates the influence and core position of each department in the policy network by calculating the centrality of nodes (such as degree centrality and betweenness centrality).

[0108] The initial weights are revised based on information on key groups and core departments. This revision is based on an understanding of the core departments and key groups in policy implementation, ensuring that the weights reflect the actual importance and influence of departments and groups in policy implementation.

[0109] Finally, the implemented policy knowledge graph is optimized based on the target weight information. This optimization process involves adjusting the connections between policies and strengthening the display of core nodes, so that the optimized policy knowledge graph more accurately reflects the structure of actual policies and the priority of implementation.

[0110] Through these detailed analysis and optimization steps, the policy knowledge map can more scientifically reflect the actual impact and importance of policies, providing a powerful support tool for the formulation, adjustment and implementation of union policies.

[0111] In some examples, the above-mentioned generation of the relevant regulations knowledge graph based on the relevant regulations information includes:

[0112] Clean and standardize the above-mentioned relevant regulatory information to obtain standard data of relevant regulatory information, where the above-mentioned relevant regulatory information includes legal and regulatory information, industry guidance information, and official announcement information;

[0113] Apply NER operation to the above-mentioned relevant regulation information standard data to obtain the second key entity information, wherein the above-mentioned second key entity information includes legal article number information, regulation name information, scope of application information, relevant department information and defined term information;

[0114] Performing a relationship extraction operation on the above-mentioned union policy data using a rule-based method to obtain second entity relationship information corresponding to the above-mentioned second key entity information;

[0115] The second node information is constructed based on the above-mentioned second key entity information, the second side information is constructed based on the above-mentioned second entity relationship information, and the above-mentioned relevant regulations knowledge graph is generated based on the above-mentioned second node information and the above-mentioned second side information, wherein the above-mentioned relevant regulations knowledge graph has different graph levels based on the entity types corresponding to the relevant regulations.

[0116] For example, data cleansing and standardization of relevant regulatory information is performed to obtain standardized data. This process includes removing invalid and redundant data, unifying data formats, and other operations to ensure data quality and consistency. Relevant regulatory information primarily includes legal and regulatory information, industry guidance information, and official announcements.

[0117] Use named entity recognition (NER) to process the relevant regulatory information standard data to obtain the second key entity information. This includes the legal article number, regulation name, scope of application, relevant departments, and defined terms. Named entity recognition is a technique in natural language processing used to identify specific entities in text, such as names of people, places, and legal articles.

[0118] Relationship extraction is performed using a rule-based approach on the relevant standard data. This step aims to clarify the relationships between the second key entity information, namely the second entity relationship information. Relationship extraction refers to determining the logical or functional connections between entities in the text, such as "applicable to" and "defined as".

[0119] Second node information is constructed based on the second key entity information. Each node represents an entity, such as a specific article or scope of application of a law or regulation. Second edge information is constructed based on the second entity relationship information. Edges represent connections between nodes in the knowledge graph and display relationships between entities.

[0120] The second node information and the second edge information are used to generate a knowledge graph of relevant regulations. The knowledge graph has different graph levels based on the entity types corresponding to the relevant regulations, thereby reflecting different types of regulations and the relationship levels between them.

[0121] Through this detailed processing and analysis, the knowledge graph of relevant regulations can clearly demonstrate the relationships and hierarchies between various laws, regulations, industry guidance, and official announcements, providing strong regulatory support and guidance for the formulation and adjustment of union policies. This helps ensure the legitimacy and adaptability of policies, while also enhancing the transparency and credibility of policymaking.

[0122] In some examples, constructing the second node information based on the second key entity information, constructing the second side information based on the second entity relationship information, and generating the relevant prescribed knowledge graph based on the second node information and the second side information include:

[0123] Determining a node hierarchy according to the type corresponding to the second key entity information to construct the second node information, wherein the type corresponding to the second key entity information includes laws, regulations, guidelines, and clauses;

[0124] Constructing second side information based on the type corresponding to the second entity relationship information, wherein the types corresponding to the second entity relationship information include inclusion relationship, authorization relationship, application relationship, revision relationship, and reference relationship, and the second entity relationship information includes a top-down hierarchical relationship and a bottom-up reverse hierarchical relationship;

[0125] The above-mentioned relevant regulations knowledge graph is generated based on the above-mentioned second node information and the above-mentioned second edge information.

[0126] Exemplarily, the node hierarchy is determined based on the type corresponding to the second key entity information to construct the second node information. This step involves identifying various entities such as laws, regulations, guidelines, and clauses and assigning them appropriate hierarchies. For example, laws may be at the top level, followed by relevant regulations, and then specific guidelines and clauses. This hierarchical structure helps clearly represent the authority and applicability relationships between entities.

[0127] The second side information is constructed based on the type of the second entity relationship information. Relationship types include inclusion, authorization, application, revision, and reference. These relationship types describe the functional and logical connections between different legal entities.

[0128] The second entity relationship information also includes top-down hierarchical relationships and bottom-up reverse hierarchical relationships. Top-down hierarchical relationships may indicate how a higher-level legal document affects or includes regulations or clauses at a lower level, while bottom-up reverse hierarchical relationships involve the support or dependency of a lower-level entity on a higher-level entity.

[0129] The second node information and second edge information determined above are used to generate a knowledge graph of relevant regulations. This knowledge graph visually represents different legal documents and their interrelationships, providing users with an intuitive way to understand the complex legal and regulatory framework and its operational logic.

[0130] Through this method, the knowledge graph of relevant regulations can accurately reflect the structural relationships and logical connections between documents such as laws and regulations, providing a powerful tool for trade unions and relevant decision makers to better understand and apply these regulations.

[0131] In some examples, the future union policy forecast information includes forecast information on the abolition of existing union policies and forecast information on new union policies;

[0132] The above-mentioned future union policy forecast information is generated based on the above-mentioned member inquiry information, the above-mentioned member information, the above-mentioned implemented policy knowledge graph and the above-mentioned relevant regulations knowledge graph, including:

[0133] Determine policy conflict information and policy missing information based on the above-mentioned implemented policy knowledge graph and the above-mentioned relevant regulations knowledge graph;

[0134] Determine the above-mentioned prediction information on the abolition of existing union policies based on the above-mentioned member inquiry information, the above-mentioned information on current members, and the above-mentioned policy conflict information;

[0135] The above-mentioned future union policy forecast information is determined based on the above-mentioned member inquiry information, the above-mentioned information on current members and the above-mentioned policy missing information.

[0136] For example, the knowledge graph of implemented policies and relevant regulations is used to systematically analyze and identify policy conflicts and policy gaps. Policy conflicts refer to situations where there are inconsistencies or contradictions between two sets of policies or regulations, while policy gaps refer to areas that are not covered or not fully regulated in the current policy system.

[0137] Determine predictions for the repeal of existing union policies based on information from member inquiries, current members, and policy conflicts. This step involves analyzing member inquiries about existing policies, the specific circumstances of current members, and identified policy conflicts to predict which existing policies may require repeal or significant revision due to conflicts.

[0138] Determine forecasts for future union policies based on information from member inquiries, information on current members, and information on policy gaps. This includes analyzing member needs and issues, considering the specific circumstances of current members, and predicting potential new policies or policy areas based on identified policy gaps.

[0139] Through the above steps, future union policy forecasts can effectively help union leaders and decision-makers understand the shortcomings of the current policy system and the actual needs of members, allowing them to make more targeted policy adjustments and formulations. The generation of this forecast information not only improves the timeliness and accuracy of policy responses, but also optimizes the quality and efficiency of union services to members.

[0140] In some examples, the policy conflict information and policy missing information determined based on the implemented policy knowledge graph and the relevant regulations knowledge graph include:

[0141] Traverse the above-mentioned implemented policy knowledge graph and the above-mentioned relevant regulations knowledge graph to extract the node information of all implemented policies and relevant regulations;

[0142] Determine the policy conflict information of the node information of the above-mentioned implementation policy and relevant regulations through natural language analysis based on preset rules;

[0143] The above-mentioned policy missing information is determined based on the difference analysis operation, and the node information of the above-mentioned implementation policy and related regulations is determined.

[0144] For example, we traverse the implemented policy knowledge graph and the relevant regulations knowledge graph, systematically examining both knowledge graphs to extract node information for all implemented policies and relevant regulations. These nodes represent each policy or regulation in the graph, including its detailed attributes and parameters. The traversal aims to comprehensively collect data from all nodes in the graph, providing foundational information for subsequent analysis.

[0145] Based on pre-set rules, natural language analysis is used to identify policy conflicts within the node information for implementing policies and related regulations. Natural language analysis primarily parses text data and identifies the logic and structure within the language. Pre-set rules include checks for logical consistency in policy implementation, conflicting objectives, or inappropriate resource allocation. This step aims to identify and document policy conflicts that could lead to implementation or legal issues.

[0146] Gap analysis is used to identify policy gaps in the node information for implementing policies and related regulations. Gap analysis involves comparing the coverage and level of detail across different policies or regulations to identify areas or issues not covered by existing policies. This step helps reveal gaps in policy that require new or detailed additions.

[0147] Through these steps, the system can identify and address conflicts and gaps in policies, providing unions and decision-makers with important information for policy adjustments and updates. This analysis not only enhances the effectiveness of policy implementation but also ensures the integrity and adaptability of the policy system. By identifying conflicts and gaps, unions can more effectively respond to member needs and changes in the external environment, thereby improving the effectiveness and fairness of policies.

[0148] In some examples, determining the prediction information of the abolition of existing union policies based on the member inquiry information, the current member information, and the policy conflict information includes:

[0149] Converting the member inquiry information, the member information at the meeting, and the policy conflict information into a first eigenvector;

[0150] Input the first eigenvector into a decision tree classifier, and use information gain and Gini impurity to determine the decision tree split point;

[0151] Based on the above decision tree split points, the prediction information of the abolition of existing union policies is determined.

[0152] For example, first, member inquiry information, current member information, and policy conflict information are integrated and converted into a first feature vector. This step involves quantifying these different types of data (such as text information, member statistics, and conflict records) into numerical feature vectors, making them readable and processable by the decision tree model. Feature vectorization converts actual data into a format acceptable to machine learning models for effective data analysis and prediction.

[0153] This first feature vector is fed into a decision tree classifier. Information gain and Gini impurity are used to determine the split point for the decision tree. Information gain is used to select the features that are most effective at classifying the data, while Gini impurity is a measure used to determine the purity or impurity of the data after splitting it at the current node.

[0154] Based on the cut-off points determined by the decision tree, the system predicts which existing union policies may need to be eliminated or revised. This prediction is based on input data such as member inquiries, member information, and policy conflicts. The model analyzes patterns and trends in this data to determine whether certain policies are problematic and may need to be eliminated.

[0155] This approach combines data science techniques and machine learning algorithms to provide a systematic and data-driven way to predict policy changes, helping unions to efficiently and accurately understand and respond to member needs and potential problems between policies, thereby optimizing the policy system.

[0156] In some examples, determining the future union policy forecast information based on the member inquiry information, the current member information, and the policy missing information includes:

[0157] Classify the member inquiry information and the member information based on cluster analysis to obtain classified question and demand information;

[0158] Determine the correlation between member needs and policy gaps based on the Apriori algorithm according to the classified problem and demand information and the above policy gap information;

[0159] The above-mentioned union policy forecast information is generated based on the policy deficiency information that the correlation between member needs and policy deficiency is greater than the preset correlation.

[0160] For example, first, member inquiry information and in-person member information are processed using cluster analysis to categorize members' questions and needs. Cluster analysis is a statistical method used to group objects in a dataset that share similar characteristics into the same group (i.e., a "cluster"), while objects that differ in these characteristics are grouped into different groups. The purpose of this step is to group various member questions and needs based on their similarities, thereby obtaining categorized question and need information.

[0161] Based on the classified questions and needs information and policy gap information, the Apriori algorithm is used to determine the correlation between member needs and policy gaps. The Apriori algorithm is an algorithm for frequent item set mining and association rule learning, and is widely used in transaction databases. This algorithm can identify strong correlations between member needs and missing components of the current policy system.

[0162] Based on the policy gap information identified by the Apriori algorithm, where the correlation between member needs and policy gaps exceeds a preset correlation threshold, union policy forecasts are generated. This step aims to identify areas that are highly relevant to member needs and not currently covered by policy, and to predict the potential need for future policy development or revisions in these areas.

[0163] This approach effectively leverages existing member feedback and policy analysis, using a data-driven approach to predict and guide the future direction of union policy, ensuring that policies better align with members' actual needs and expectations while filling gaps in the existing policy framework. This approach helps enhance the adaptability and foresight of union policy, fostering positive interaction and support between unions and their members.

[0164] It should be noted that, in the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0165] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0166] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram and the combination of processes and / or blocks in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0167] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0168] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0169] An embodiment of the present application also provides a computer program product, which includes computer software instructions. When the computer software instructions are executed on a processing device, the processing device executes the process of predicting the decomposition rate in alumina production in the corresponding embodiment.

[0170] A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, they fully or partially produce the processes or functions according to the embodiments of the present application. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be stored by a computer, or a data storage device such as a server or data center that integrates one or more available media. Available media can include magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state drives (SSDs)).

[0171] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0172] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0173] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0174] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0175] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program code.

[0176] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A control method, a union policy intelligent analysis system based on knowledge graph, characterized in that: The system comprises: The current union policy collection module is used to collect information on union policy documents that have been implemented in the target enterprise; A member information collection module is used to collect information on current members and member enquiry information within the target enterprise; A relevant regulations collection module is used to collect relevant regulations information related to the union; An intelligent parsing module, configured to generate a knowledge graph of implemented policies based on the implemented union policy document information, and further configured to generate a knowledge graph of relevant regulations based on the relevant regulations information; A future union policy prediction module, configured to generate future union policy prediction information based on the member inquiry information, the current member information, the implemented policy knowledge graph, and the relevant regulations knowledge graph; The intelligent reply module is used to generate intelligent reply information based on the member inquiry information, the member information at the meeting, and the implemented policy knowledge graph. The method comprises: Collecting information on the implemented union policy documents, information on current members, and information on member inquiries within the target enterprise; Generate a knowledge graph of implemented policies based on the information of the implemented union policy documents; Generate a knowledge graph of relevant regulations based on relevant regulations information; Generate the future union policy forecast information based on the member inquiry information, the current member information, the implemented policy knowledge graph, and the relevant regulations knowledge graph; Generate the intelligent reply information according to the member inquiry information, the member information at the meeting, and the implemented policy knowledge graph; The future union policy forecast information includes forecast information on the abolition of existing union policies and forecast information on new union policies; The generating of the future union policy forecast information based on the member inquiry information, the current member information, the implemented policy knowledge graph, and the relevant regulations knowledge graph includes: Determining policy conflict information and policy missing information based on the implemented policy knowledge graph and the relevant regulations knowledge graph; determining the existing union policy abolition prediction information based on the member inquiry information, the current member information, and the policy conflict information; Determining the future union policy forecast information based on the member inquiry information, the current member information, and the policy missing information; The determining of policy conflict information and policy missing information based on the implemented policy knowledge graph and the relevant regulations knowledge graph includes: Traversing the implemented policy knowledge graph and the relevant regulations knowledge graph to extract node information of all implemented policies and relevant regulations; Determining the policy conflict information of the node information of the implementation policy and related regulations through natural language analysis based on preset rules; Determining the policy missing information of the node information implementing the policy and related regulations based on the difference analysis operation; The determining of the existing union policy abolition prediction information based on the member inquiry information, the current member information, and the policy conflict information includes: Converting the member inquiry information, the member information at the meeting, and the policy conflict information into a first feature vector; Input the first feature vector into a decision tree classifier, and determine a decision tree split point using information gain and Gini impurity; The prediction information of the abolition of the existing union policy is determined based on the decision tree split point.

2. The control method according to claim 1, characterized in that: The generating of the implemented policy knowledge graph based on the implemented union policy document information includes: Cleaning and normalizing the implemented union policy document information to obtain union policy standard data, wherein the union policy document information includes union document information, union database information, and union meeting minutes information; Applying natural language processing to the union policy data to obtain first key entity information, wherein the first key entity information includes policy name information, implementation date information, involved department information, and associated employee group information; performing a relation extraction operation on the union policy data using a relation extraction model to obtain first entity relationship information corresponding to the first key entity information; First node information is constructed based on the first key entity information, first side information is constructed based on the first entity relationship information, and the implemented policy knowledge graph is generated based on the first node information and the first side information.

3. The control method according to claim 2, characterized in that: The method further comprises: Calculate the policy initial weight information based on the involved department information, the related employee group information and the query frequency information; Identify the main group information of the currently implemented policy knowledge graph through the Louvain algorithm; Perform centrality analysis on the currently implemented policy knowledge graph to obtain core department information; Modifying the initial weight information based on the main group information and the core department information to obtain target weight information; The implemented policy knowledge graph is optimized based on the target weight information to obtain an optimized implemented policy knowledge graph.

4. The control method according to claim 1, wherein: The generating of the relevant regulations knowledge graph according to the relevant regulations information includes: Cleaning and normalizing the relevant regulatory information to obtain standard data of the relevant regulatory information, wherein the relevant regulatory information includes legal and regulatory information, industry guidance information, and official announcement information; Applying a NER operation to the relevant regulation information standard data to obtain second key entity information, wherein the second key entity information includes legal article number information, regulation name information, applicable scope information, relevant department information, and defined term information; performing a relationship extraction operation on the union policy data using a rule-based method to obtain second entity relationship information corresponding to the second key entity information; Second node information is constructed based on the second key entity information, second side information is constructed based on the second entity relationship information, and the relevant regulations knowledge graph is generated based on the second node information and the second side information, wherein the relevant regulations knowledge graph has different graph levels based on the entity types corresponding to the relevant regulations.

5. The control method according to claim 4, characterized in that: The constructing second node information according to the second key entity information, constructing second side information according to the second entity relationship information, and generating the relevant prescribed knowledge graph according to the second node information and the second side information includes: Determining a node hierarchy according to a type corresponding to the second key entity information to construct the second node information, wherein the type corresponding to the second key entity information includes laws, regulations, guidelines, and clauses; Constructing second side information according to the type corresponding to the second entity relationship information, wherein the types corresponding to the second entity relationship information include inclusion relationship, authorization relationship, application relationship, revision relationship, and reference relationship, and the second entity relationship information includes a top-down hierarchical relationship and a bottom-up reverse hierarchical relationship; Generate the relevant prescribed knowledge graph based on the second node information and the second edge information.

6. The control method according to claim 5, characterized in that: The determining of the future union policy forecast information based on the member inquiry information, the current member information, and the policy missing information includes: Classify the member inquiry information and the member information based on the cluster analysis method to obtain classified question and demand information; Determine the correlation between member needs and policy gaps based on the Apriori algorithm according to the classified problem and demand information and the policy gap information; The union policy forecast information is generated based on the policy deficiency information in which the correlation between member demand and policy deficiency is greater than a preset correlation.

Images