A policy information recommendation method and system based on big data

Abstract

This invention relates to the field of big data processing and discloses a policy information recommendation method and system based on big data. One method involves breaking down clauses concerning applicable conditions, application requirements, and implementation restrictions into independently identifiable policy elements. Based on basic policy data, the method compares the correspondence between policy elements and business process nodes item by item. For the recorded policy elements, it performs cross-textual organization and rearrangement of relevant policy content according to policy effective time, business process progression sequence, and the support and dependency relationships between different policies. Based on the rearranged policy content, it performs a completeness check on the currently presented policy information. After supplementing policy entries into the presentation scope, it adjusts the policy presentation order and display scope to form a policy information presentation result that matches the business process stage. This invention has the advantage of improving the completeness of policy acquisition.

Description

Technical Field

[0001] This invention relates to the field of big data processing, specifically to a method and system for recommending policy information based on big data. Background Technology

[0002] In application scenarios such as government service platforms, industry support information disclosure systems, and enterprise service windows, policy documents are usually released in batches annually, by competent authorities, or by policy category, and provided to users through keyword search or manual screening. However, in actual use, users often only have a vague background of policy needs, and it is difficult for enterprises to accurately describe the corresponding policy names or key points of clauses based on their industry, development stage, or current business operations. This results in a low hit rate for traditional search methods based on keywords or directory structures.

[0003] For example, when small and medium-sized enterprises apply for a certain period of special subsidies, they often need to understand multiple supporting policies related to their business scale, industry attributes and application time window. However, the existing system is difficult to effectively integrate and recommend policies based on the large-scale policy data stored in a scattered manner when the user does not explicitly specify all the query conditions, which leads to insufficient policy access or missed application opportunities.

[0004] Therefore, it is necessary to design a big data-based policy information recommendation method and system to improve the completeness of policy access. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a policy information recommendation method and system based on big data, which has the advantage of improving the completeness of policy acquisition and solves the problems mentioned in the background technology.

[0006] To achieve the aforementioned goal of improving the completeness of policy access, this invention provides the following technical solution: a policy information recommendation method based on big data, comprising the following steps: The clauses involving applicable conditions, application requirements and implementation restrictions are broken down into policy elements that can be independently identified, and the reference relationships, preconditions and exceptions between the clauses are marked simultaneously. The applicable premises that were originally scattered in the main text are organized into identifiable policy basic data. Based on the policy data, the process node information corresponding to the user's current handling matters is introduced. The correspondence between policy elements and business process nodes is compared item by item. By verifying the preconditions required by the policy clauses and the current status of the business, policy elements that are not explicitly retrieved but constrain the handling of business are recorded. For the recorded policy elements, the relevant policy content is organized and rearranged across texts according to the policy effective time, the order of business process advancement, and the support and dependency relationships between different policies. The originally scattered constraint clauses are aligned and marked with the corresponding business process nodes one by one. Based on the rearranged policy content, the completeness of the currently presented policy information is checked, policy content that has supporting relationships but is not included in the presentation scope is marked, and corresponding supplementary policy entries are generated. After supplementing policy items into the presentation scope, the order and scope of policy presentation are adjusted based on changes in user operations during business processing and policy review, resulting in policy information presentation outcomes that match the stages of the business process.

[0007] Preferably, the process of organizing the applicable premises that are originally scattered throughout the main text into identifiable policy basis data is as follows: The original policy text is parsed at the paragraph and syntactic levels to identify key semantic segments that represent conditional triggering, eligibility restrictions, time constraints, and exclusive application. Using clause number, issuing entity, and effective date as indexes, the identified semantic fragments are decomposed into elements to generate policy element units with unique identifiers; Structured annotations are used to identify references, prerequisite relationships, and exceptions between policy element units, thus constructing explicit descriptions of the relationships between policy elements. The decomposed policy element units and their associated descriptions are uniformly stored in the policy basic data set to form a structured policy data foundation that can be retrieved, compared, and reused.

[0008] The preferred method is to compare the correspondence between policy elements and business process nodes item by item as follows: Obtain the business type, processing stage, and completed node information corresponding to the user's current task, and construct a set of business process node status descriptions; The system matches the set of business process node status descriptions with the policy element application conditions in the policy basic data set item by item to identify the applicability and limitations of policy elements to the current process node. Cases where conditions are missing, states are not met, or sequences are inconsistent during the matching process are marked, forming a mapping table of correspondence between policy elements and business process nodes.

[0009] Preferably, the process of recording policy elements that are not explicitly retrieved but constrain business operations is as follows: Based on the mapping table, the status of each precondition marked on each policy element is checked to determine whether it has been met in the current business process node. When a precondition is detected that is not met but is not explicitly presented in the current policy search results, the corresponding policy element will be marked as an implicit constraint element. Record the triggering conditions, impact process nodes, and potential consequences of implicit constraint elements to form an implicit policy constraint record set.

[0010] The preferred method for cross-textual organization and rearrangement of relevant policy content is as follows: The identified policy elements are sorted in chronological order according to the policy release time, effective date, and expiration conditions. Based on the sequence of business process nodes, the applicable positions of policy elements at different processing stages are redefined; Based on the support and dependency relationships marked between policy elements, elements from different policy texts are logically aggregated to generate a multi-dimensional policy element sequence rearranged according to time, process, and dependency relationships.

[0011] Preferably, the process of aligning and annotating the originally scattered constraint clauses with their corresponding business process nodes item by item is as follows: Based on the rearranged policy element sequence, extract constraint-type elements with conditional restrictions, eligibility constraints, and operational prohibitions. Map and match constraint elements with business process node states to determine the effective nodes and effective ranges; Multiple constraint elements corresponding to the same process node are merged and labeled, and the types of mandatory constraints and additional constraints are distinguished to generate policy constraints and process node aligned labeling results.

[0012] The preferred process for verifying the completeness of currently presented policy information is as follows: Obtain the set of policy information currently displayed to the user and parse the coverage of the corresponding policy elements; By comparing the policy information set with the rearranged policy element sequence, missing supporting policy elements and unpresented constraint elements can be identified. Missing elements are categorized and labeled according to their impact on business processing results, forming a policy presentation integrity verification result.

[0013] Preferably, the process of generating corresponding supplementary policy entries is as follows: Based on the completeness check results, filter out unpresented policy content that has a supporting relationship with the current business process nodes; Extract applicable conditions, impact points, and prompts from the selected policy content, and generate supplementary policy entry descriptions.

[0014] The preferred process for generating policy information presentation results that match the stages of the business process is as follows: Collect user operation records in the business processing interface and policy viewing interface, and analyze the currently focused process nodes and policy types; Based on changes in user actions, the priority of policy items in the candidate set for policy information presentation is adjusted. While ensuring that key constraint policies and supplementary policy items are effectively highlighted, the display order and scope of policy information are dynamically controlled to output policy information presentation results that are adapted to the current business process stage.

[0015] This invention also discloses another technical solution: a policy information recommendation system based on big data, comprising: Element annotation module: Annotates the reference relationships, preconditions and exceptions between clauses, and organizes them into identifiable policy basic data; Process comparison module: It introduces the process node information of the user's current handling matter, compares the policy elements with the business process nodes one by one, and records the policy elements that are not explicitly retrieved but constrain the handling of the business. Relationship rearrangement module: Based on the policy effective time, the order of business process implementation, and the support and dependency relationships between different policies, the constraint clauses are aligned and marked with the corresponding process nodes; Complete verification module: Based on the rearranged policy content, it performs a completeness verification of the currently presented policy information and generates supplementary policy entries; Dynamic adjustment module: After adding policy items to the presentation scope, the order and scope of policy presentation are adjusted.

[0016] Compared with existing technologies, the present invention provides a policy information recommendation method and system based on big data, which has the following beneficial effects: This invention breaks down the scattered application conditions, application requirements, and enforcement restrictions in policy texts into independently identifiable policy elements. These elements are then structured and annotated with citations between clauses, preconditions, and exceptions, making the policy application prerequisites explicit and calculable. Furthermore, by incorporating business process node information of the user's current transaction, the invention verifies each policy element and business status item by item. This not only accurately identifies directly applicable policy content at the current stage but also uncovers implicit policy elements that, while not explicitly retrieved, substantially constrain business transactions. This avoids repeated processes or failures due to incomplete policy understanding. By organizing relevant content in different policy texts chronologically… The system reorganizes and rearranges the effective order and interdependencies across texts, aligning and marking binding clauses with specific business process nodes. This ensures that the presentation order of policy content aligns with the actual processing logic, significantly improving the readability and operability of policy information. By verifying the completeness of presented policy information and automatically generating supplementary policy entries, the system guarantees comprehensive coverage of policy display results in terms of support and constraints. Furthermore, the system dynamically adjusts the display order and scope of policy information based on user actions during processing and browsing, enabling policy recommendations to adaptively update as business process stages change. This reduces the cost of policy comprehension for users while improving policy matching accuracy and business processing success rates. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the method of the present invention; Figure 2 This is a schematic diagram of the structure of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1: Please refer to Figure 1 As shown in the figure, a policy information recommendation method based on big data in an embodiment of the present invention includes the following steps: S1: The clauses involving applicable conditions, application requirements and implementation restrictions are broken down into policy elements that can be independently identified, and the reference relationships, preconditions and exceptions between the clauses are marked simultaneously, and the applicable premises that were originally scattered in the main text are organized into identifiable policy basic data.

[0020] The process in S1 of organizing the applicable premises, which were originally scattered throughout the main text, into identifiable policy basis data is as follows: The original policy text is parsed at the paragraph and syntactic levels to identify key semantic segments indicating conditional triggering, eligibility restrictions, time constraints, and exclusive application. The obtained policy text undergoes structural preprocessing, including title identification, clause segmentation, and punctuation standardization, to form a standardized text structure suitable for parsing. Constrained by natural language processing rules and a domain semantic lexicon, the text is segmented at the paragraph level, and further syntactic component analysis is performed within each paragraph to identify sentence structures containing semantic features such as "under the condition of…", "meets the requirements of…", "from the date of…", and "not applicable to…". For the identified syntactic segments, semantic role labeling is used to distinguish triggering conditions, applicable objects, time constraints, and exclusionary descriptions, thereby extracting the application premises originally implicit in natural language into locatable key semantic segments. Using clause number, issuing entity, and effective date as indexes, the identified semantic fragments are decomposed into elements to generate policy element units with unique identifiers. After identifying key semantic fragments, a master index is established according to the corresponding clause number in the policy text. A multi-dimensional index system is constructed by combining information on the policy issuing authority, policy level, and effective or expired date. Under this index system, each semantic fragment is decomposed into elements, breaking it down into basic fields such as "description of applicable conditions," "description of applicable objects," and "description of behavioral or result constraints." A unique policy element identifier code is assigned to each set of fields, so that different applicable premises in the same policy clause are separated into multiple independently referable policy element units, while preserving the correspondence between the source clause and the policy meta-information. The system provides structured annotations for references, prerequisites, and exceptions between policy element units, constructing explicit descriptions of their relationships. After generating policy element units, the system further analyzes the logical relationships between them. By scanning semantic fragments for referential expressions, cross-clause references, and exception statements, the system identifies dependencies between elements. For policy content containing expressions such as "must first meet…clause," "except in…cases," or "refer to…implementation," the system labels these relationships as prerequisites, exceptions, or references, respectively. The system then records these relationships in a structured manner using fields for association type, association direction, and effective conditions. This transforms the implicit logical relationships between policy elements into calculable and traceable explicit descriptions of their relationships. The decomposed policy element units and their associated descriptions are uniformly stored in the policy basic data set to form a structured policy data foundation that can be retrieved, compared, and reused. After the labeling of policy element units and their relationships is completed, all element data is stored and managed according to a unified data model. The data model includes at least element identifiers, source clause information, semantic field content, relationship descriptions, and validity period information. By establishing a storage structure based on element identifiers and index fields, the policy basic data set can support rapid retrieval based on policy source, applicable condition type, or business matching needs.

[0021] S2: Based on the policy data, introduce the process node information corresponding to the user's current handling matters, compare the correspondence between policy elements and business process nodes item by item, and record policy elements that are not explicitly retrieved but constrain business handling by verifying the preconditions required by the policy clauses and the current status of the business.

[0022] The process of comparing the correspondence between policy elements and business process nodes in S2 is as follows: The system obtains the business type, processing stage, and completed node information corresponding to the user's current processing item, and constructs a set of business process node status descriptions. By connecting to the business processing system or business process management system, it obtains the item identifier of the user's current processing item, and retrieves a predefined business process template based on the item identifier. It determines the business type and current processing stage of the item from the business process template, and reads the process execution log or node status record to identify completed, in progress, and untriggered process nodes. It combines the node identifier, node type, completion status, and time sequence information of each process node to form a set of business process node status descriptions reflecting the current processing progress. The system performs item-by-item matching between the set of business process node status descriptions and the policy element applicability conditions in the policy basic data set to identify the applicability and limitations of policy elements to the current process node. After constructing the set of business process node status descriptions, it extracts the corresponding applicability condition fields for each policy element unit stored in the policy basic data set, including information such as business type restrictions, processing stage requirements, and prerequisite node constraints. Using business type consistency, processing stage matching degree, and node completion status as comparison dimensions, it performs item-by-item matching between the set of business process node status descriptions and the policy element applicability conditions. By judging whether the current process node status meets the triggering conditions and sequence constraints required by the policy element, it determines the applicability status of the policy element under the current business process node and simultaneously identifies the restrictive conditions in the policy element that limit or block the progress of the business process. Cases involving missing conditions, unmet states, or inconsistent sequences during the matching process are marked to form a mapping table of correspondence between policy elements and business process nodes. After completing item-by-item matching, the matching results are checked for consistency and anomaly marking is performed. Cases where the necessary conditions required by the policy element are missing in the business process node state, the node state does not meet the policy triggering requirements, or the node execution order is inconsistent with the policy prerequisites are marked as missing conditions, unmet states, or inconsistent sequences, respectively. The marking results are associated with the corresponding policy element identifier, business process node identifier, and restriction reason, generating a mapping table of correspondence between policy elements and business process nodes. This table is used to clearly identify the applicable, restricted, or temporarily inapplicable status of each policy element in the current business process node.

[0023] The process in S2 that records policy elements that are not explicitly retrieved but constrain business transactions is as follows: Based on the mapping table, the status of each precondition marked by each policy element is checked to determine whether it has been met in the current business process node. The mapping table between the policy elements and business process nodes is read, and the description of the precondition associated with each policy element and its corresponding business process node requirements are extracted. Based on the set of business process node status descriptions, the status, completion result or stage attribute of the process node pointed to by each precondition is checked item by item. By comparing the execution status of the current node with the trigger threshold or completion requirement defined in the precondition of the policy element, it is determined whether the precondition has been met in the current business process node, and the check result is recorded in the form of met, not met or uncertain status. When a precondition is not met but is not explicitly presented in the current policy search results, the corresponding policy element is marked as an implicit constraint element. After completing the precondition status verification, the verification results are further cross-compared with the list of policy elements that have been retrieved and presented to the user during the current business process. When it is found that the precondition of a certain policy element is not met and the policy element does not appear as an explicit applicable condition in the current policy search or recommendation results, it is determined that the policy element belongs to the situation where it is not explicitly retrieved but actually constrains the business process. For such policy elements, an implicit constraint identifier field is set to mark them so that they are recorded and managed in the system in a way that is different from explicit policy elements. For implicit constraint elements, record the triggering conditions, affected process nodes, and potential consequences to form an implicit policy constraint record set. After identifying the implicit constraint element, extract its triggering condition information from the applicable condition field and the relationship description corresponding to the policy element, and combine it with the business process node status description set to identify the process nodes directly or indirectly affected by it. Based on the restrictive descriptions or exception clauses defined in the policy element, organize its potential consequences on the business process, including process delays, node rollbacks, or processing failures. The identification information, triggering conditions, affected process nodes, and potential consequences information of the implicit constraint element are structured and stored to form an implicit policy constraint record set.

[0024] S3: For the recorded policy elements, organize and rearrange the relevant policy content across texts according to the policy effective time, the order of business process advancement, and the support and dependency relationships between different policies, and align and annotate the originally scattered constraint clauses with the corresponding business process nodes item by item.

[0025] The process of cross-textual organization and rearrangement of relevant policy content in S3 is as follows: The identified policy elements are sorted by time series according to policy release time, effective date, and expiration conditions. The identified policy element units are retrieved from the policy basic data set, and their corresponding policy meta-information fields are read, including policy release time, effective date, clearly defined expiration date, and attached expiration trigger conditions. For policy elements with multiple time attributes, their time attributes are uniformly processed according to the principle of prioritizing effective date, followed by release time, and then correcting for expiration conditions. After the time attribute uniformity is completed, the policy elements are sorted based on the time axis, and the status of elements that have exceeded the expiration conditions or are in an ineffective state is marked. By combining the order of business process nodes, the applicable positions of policy elements at different processing stages are repositioned. After completing the time series sorting, a business process model is introduced to read the order of business process nodes corresponding to the current processing item. Based on the applicable stage, preconditions, and restrictive descriptions marked in the policy elements, the policy elements are mapped to business process nodes to determine whether they should be applied to different processing stages such as process initiation, material preparation, review and verification, or result confirmation. When the description order in the original source text of the policy element is inconsistent with the order of business process progression, the policy element is adjusted to the business process node where it actually has a binding effect through repositioning, so that the policy content is consistent with the actual processing order in the process dimension. Based on the support and dependency relationships marked between policy elements, elements from different policy texts are logically aggregated to generate a policy element sequence that is rearranged in multiple dimensions according to time, process, and dependency. After completing the rearrangement in the time and process dimensions, the policy elements from different policy texts are logically aggregated using the support, prerequisite dependency, and exception exclusion relationships marked between policy elements. By taking the supported element as the core, the superior elements, supplementary explanatory elements, and restrictive elements that it depends on are associated and combined to form an element set with complete constraint semantics. The element set is then sorted in multiple dimensions according to time order, business process node order, and element dependency relationship to generate a policy element sequence that is cross-text, cross-time, and conforms to the actual handling logic.

[0026] In S3, the process of aligning and annotating the originally scattered constraint clauses with their corresponding business process nodes item by item is as follows: Based on the rearranged policy element sequence, constraint elements with conditional restrictions, eligibility constraints, and operational prohibitions are extracted. From the formed policy element sequence, the semantic fields and attribute identifiers of each policy element are read one by one. By parsing the conditional descriptions, subject eligibility requirements, and prohibitive or negative operational statements contained in the elements, the policy elements are classified by attributes. For policy elements containing constraint semantic features such as "should meet", "limited to", "shall not be implemented", and "not accepted", they are identified as constraint elements. Their constraint objects, constraint content, and triggering conditions are further distinguished. The constraint clauses scattered in different policy texts are uniformly extracted into a structured set of constraint elements. The constraint elements are mapped and matched with the business process node states to determine the effective nodes and effective intervals. After identifying the constraint elements, the applicable stage, preconditions, and time constraint information marked for each constraint element are analyzed in conjunction with the business process node state description set. By comparing the triggering conditions of the constraint elements with the state attributes, execution order, and time points of the business process nodes, the actual business process nodes in which the constraint element applies are determined. Based on the effective time and failure conditions of the constraint elements, their effective interval in the business process is calculated, and the effective nodes and effective intervals are recorded as the node mapping results of the constraint elements. Multiple constraint elements corresponding to the same process node are merged and labeled, and mandatory constraints and additional constraints are distinguished to generate policy constraints and process node aligned labeling results. After completing the mapping of constraint elements to process nodes, multiple constraint elements associated with the same business process node are merged according to their constraint strength, trigger necessity and impact on process progress. Constraint elements that, if not satisfied, directly cause the process to be unable to continue or fail are marked as mandatory constraints, and constraint elements that only take effect under specific conditions or only affect some situations are marked as additional constraints. The merged constraint labeling results are associated and stored with the corresponding business process node identifiers to generate labeling results that align policy constraints with business process nodes item by item.

[0027] S4: Based on the rearranged policy content, perform a completeness check on the currently presented policy information, identify policy content that has supporting relationships but is not included in the presentation scope, and generate corresponding supplementary policy entries.

[0028] The process of performing a completeness check on the currently presented policy information in S4 is as follows: Obtain the set of policy information currently displayed to the user and parse the corresponding policy element coverage; obtain the set of policy information that has been displayed to the user, which includes policy article summaries, applicable prompts or policy recommendation results; for the set of policy information, parse the policy element identifiers that it references or is associated with, extract its corresponding applicable conditions, constraint descriptions and supporting relationship fields; by mapping the presented policy information to its source policy elements, determine the policy element coverage of the currently displayed content. The policy information set is compared with the rearranged policy element sequence to identify missing supporting policy elements and unpresented constraint elements. After parsing the presented policy element set, the generated rearranged policy element sequence is introduced as the complete benchmark of the policy content. By comparing the presented policy element set with the rearranged policy element sequence, each item in the currently displayed content is checked to see if there are other policy elements that have a supporting relationship, a prerequisite dependency relationship, or a restrictive relationship with the presented policy elements. When it is found that a policy element is not directly displayed, but it has a real impact on the current business process as a supporting element or a constraint element, the policy element is identified as an unpresented missing element and its corresponding missing type is recorded. Missing elements are categorized and labeled according to their impact on the business processing outcome, forming a policy presentation integrity verification result. After identifying missing elements, the potential impact of the missing elements on the business processing outcome is assessed by combining the business process node status description set and the policy constraint labeling result. Missing elements that directly affect whether the process can continue or lead to processing failure are marked as high-impact level, missing elements that may cause process correction, delay, or duplicate processing are marked as medium-impact level, and other missing elements that have a small impact on the processing outcome but have informational value are marked as low-impact level. The missing element labels, impact levels, and corresponding business process node information are recorded in a structured manner to form a policy presentation integrity verification result.

[0029] The process of generating corresponding supplementary policy entries in S4 is as follows: Based on the integrity check results, unpresented policy content that supports the current business process nodes is screened; the policy presentation integrity check results are read, and policy elements marked as unpresented are extracted. These elements are then preliminarily screened based on their impact level and relationship type. For unpresented policy elements, the support relationships, prerequisite dependencies, or constraints between them and the current business process nodes are analyzed to determine whether they provide actual support for the current processing stage. For policy elements confirmed to have direct or indirect support relationships with the current business process nodes, their corresponding policy content is compiled as candidate supplementary policy content to avoid irrelevant policy information being presented repeatedly. The applicable conditions, impact points, and prompts are extracted from the selected policy content to generate supplementary policy item descriptions. After determining the candidate supplementary policy content, the applicable condition descriptions, constraint triggering conditions, and impact business process node information are extracted from the corresponding policy element fields. Combined with the potential consequences and impact levels of the policy elements, the key information that needs to be prompted to users in the current handling scenario is organized to generate supplementary policy items that include explanations of applicable prerequisites, prompts of action points, and descriptions of precautions. The supplementary policy items are encapsulated in a structured description form so that they can be used for supplementary display, risk reminders, or handling guidance without interfering with the existing policy display structure.

[0030] S5: After supplementing policy items into the presentation scope, adjust the policy presentation order and display scope based on changes in user operations during business processing and policy review, forming a policy information presentation result that matches the business process stage.

[0031] The process in S5 to generate policy information presentation results that match the business process stages is as follows: Collect user operation records in the business processing interface and policy viewing interface, and analyze the process nodes and policy types currently of interest; by deploying an operation record collection mechanism in the business processing interface and policy viewing interface, record users' page access, node clicks, policy expansion, scrolling and return operations in real time; based on the collected operation records, combined with the current business process node status, analyze the process node information that users stay for a long time, interact frequently or view repeatedly; by analyzing users' operation behavior of expanding, collapsing or switching policy items, identify the policy types that users are currently focusing on. Based on changes in user actions, the priority of policy items in the candidate set of policy information presentation is adjusted. After completing the user attention analysis, for the aforementioned candidate set of policy information presentation, the display priority of each policy item is dynamically adjusted in combination with changes in the degree of attention reflected in the user's operation records. Policy items that are highly relevant to the user's current focus process node and have been frequently viewed or operated recently are given higher display priority, while policy items with low relevance or that the user has clearly ignored are appropriately de-prioritized. Through the priority adjustment mechanism, the order of policy information presentation can dynamically respond to changes in user operations, thereby enhancing the relevance and readability of policy presentation. While ensuring that key constraint policies and supplementary policy items are effectively highlighted, the system dynamically controls the display order and scope of policy information to output policy information presentation results that are appropriate for the current business process stage. When implementing display control, it ensures that policy elements marked as key constraints and supplementary policy items are always in a visible or quickly expandable display position. Based on the current business process stage and policy item priority, the system dynamically controls the display order of policy information and restricts the default expansion level of non-key policy items. Through coordinated control of display order, expansion scope, and prompting methods, the system outputs policy information presentation results that match the current business process stage, enabling users to obtain the most influential policy content for the current processing stage in a timely manner without being overwhelmed by excessive information.

[0032] Example 2: As Figure 2 As shown, a policy information recommendation system based on big data includes: Element annotation module: Annotates the reference relationships, preconditions and exceptions between clauses, and organizes them into identifiable policy basic data; Process comparison module: It introduces the process node information of the user's current handling matter, compares the policy elements with the business process nodes one by one, and records the policy elements that are not explicitly retrieved but constrain the handling of the business. Relationship rearrangement module: Based on the policy effective time, the order of business process implementation, and the support and dependency relationships between different policies, the constraint clauses are aligned and marked with the corresponding process nodes; Complete verification module: Based on the rearranged policy content, it performs a completeness verification of the currently presented policy information and generates supplementary policy entries; Dynamic adjustment module: After adding policy items to the presentation scope, the order and scope of policy presentation are adjusted.

[0033] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0034] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A policy information recommendation method based on big data, characterized in that, Includes the following steps: The clauses involving applicable conditions, application requirements and implementation restrictions are broken down into policy elements that can be independently identified, and the reference relationships, preconditions and exceptions between the clauses are marked simultaneously. The applicable premises that were originally scattered in the main text are organized into identifiable policy basic data. Based on the policy data, the process node information corresponding to the user's current handling matters is introduced. The correspondence between policy elements and business process nodes is compared item by item. By verifying the preconditions required by the policy clauses and the current status of the business, policy elements that are not explicitly retrieved but constrain the handling of business are recorded. For the recorded policy elements, the relevant policy content is organized and rearranged across texts according to the policy effective time, the order of business process advancement, and the support and dependency relationships between different policies. The originally scattered constraint clauses are aligned and marked with the corresponding business process nodes one by one. Based on the rearranged policy content, the completeness of the currently presented policy information is checked, policy content that has supporting relationships but is not included in the presentation scope is marked, and corresponding supplementary policy entries are generated. After supplementing policy items into the presentation scope, the order and scope of policy presentation are adjusted based on changes in user operations during business processing and policy review, resulting in policy information presentation outcomes that match the stages of the business process.

2. The policy information recommendation method based on big data according to claim 1, characterized in that, The process of organizing the applicable premises, which were originally scattered throughout the main text, into identifiable policy-based data is as follows: The original policy text is parsed at the paragraph and syntactic levels to identify key semantic segments that represent conditional triggering, eligibility restrictions, time constraints, and exclusive application. Using clause number, issuing entity, and effective date as indexes, the identified semantic fragments are decomposed into elements to generate policy element units with unique identifiers; Structured annotations are used to identify references, prerequisite relationships, and exceptions between policy element units, thus constructing explicit descriptions of the relationships between policy elements. The decomposed policy element units and their associated descriptions are uniformly stored in the policy basic data set to form a structured policy data foundation that can be retrieved, compared, and reused.

3. The policy information recommendation method based on big data according to claim 2, characterized in that, The process of comparing the correspondence between policy elements and business process nodes item by item is as follows: Obtain the business type, processing stage, and completed node information corresponding to the user's current task, and construct a set of business process node status descriptions; The system matches the set of business process node status descriptions with the policy element application conditions in the policy basic data set item by item to identify the applicability and limitations of policy elements to the current process node. Cases where conditions are missing, states are not met, or sequences are inconsistent during the matching process are marked, forming a mapping table of correspondence between policy elements and business process nodes.

4. The policy information recommendation method based on big data according to claim 3, characterized in that, The process of recording policy elements that are not explicitly retrieved but constrain business operations is as follows: Based on the mapping table, the status of each precondition marked on each policy element is checked to determine whether it has been met in the current business process node. When a precondition is detected that is not met but is not explicitly presented in the current policy search results, the corresponding policy element will be marked as an implicit constraint element. Record the triggering conditions, impact process nodes, and potential consequences of implicit constraint elements to form an implicit policy constraint record set.

5. The policy information recommendation method based on big data according to claim 4, characterized in that, The process of cross-textual organization and rearrangement of relevant policy content is as follows: The identified policy elements are sorted in chronological order according to the policy release time, effective date, and expiration conditions. Based on the sequence of business process nodes, the applicable positions of policy elements at different processing stages are redefined; Based on the support and dependency relationships marked between policy elements, elements from different policy texts are logically aggregated to generate a multi-dimensional policy element sequence rearranged according to time, process, and dependency relationships.

6. The policy information recommendation method based on big data according to claim 5, characterized in that, The process of aligning and annotating the originally scattered constraint clauses with their corresponding business process nodes item by item is as follows: Based on the rearranged policy element sequence, extract constraint-type elements with conditional restrictions, eligibility constraints, and operational prohibitions. Map and match constraint elements with business process node states to determine the effective nodes and effective ranges; Multiple constraint elements corresponding to the same process node are merged and labeled, and the types of mandatory constraints and additional constraints are distinguished to generate policy constraints and process node aligned labeling results.

7. The policy information recommendation method based on big data according to claim 6, characterized in that, The process of verifying the completeness of the currently presented policy information is as follows: Obtain the set of policy information currently displayed to the user and parse the coverage of the corresponding policy elements; By comparing the policy information set with the rearranged policy element sequence, missing supporting policy elements and unpresented constraint elements can be identified. Missing elements are categorized and labeled according to their impact on business processing results, forming a policy presentation integrity verification result.

8. The policy information recommendation method based on big data according to claim 7, characterized in that, The process of generating the corresponding supplementary policy entries is as follows: Based on the completeness check results, filter out unpresented policy content that has a supporting relationship with the current business process nodes; Extract applicable conditions, impact points, and prompts from the selected policy content, and generate supplementary policy entry descriptions.

9. The policy information recommendation method based on big data according to claim 8, characterized in that, The process of creating policy information presentation results that match the stages of the business process is as follows: Collect user operation records in the business processing interface and policy viewing interface, and analyze the currently focused process nodes and policy types; Based on changes in user actions, the priority of policy items in the candidate set for policy information presentation is adjusted. While ensuring that key constraint policies and supplementary policy items are effectively highlighted, the display order and scope of policy information are dynamically controlled to output policy information presentation results that are adapted to the current business process stage.

10. A policy information recommendation system based on big data, applied to the policy information recommendation method based on big data as described in any one of claims 1-9, characterized in that, include: Element annotation module: Annotates the reference relationships, preconditions and exceptions between clauses, and organizes them into identifiable policy basic data; Process comparison module: It introduces the process node information of the user's current handling matter, compares the policy elements with the business process nodes one by one, and records the policy elements that are not explicitly retrieved but constrain the handling of the business. Relationship rearrangement module: Based on the policy effective time, the order of business process implementation, and the support and dependency relationships between different policies, the constraint clauses are aligned and marked with the corresponding process nodes; Complete verification module: Based on the rearranged policy content, it performs a completeness verification of the currently presented policy information and generates supplementary policy entries; Dynamic adjustment module: After adding policy items to the presentation scope, the order and scope of policy presentation are adjusted.

Images