Multi-modal semantic and dynamic adaptive industrial policy matching method and system

Abstract

The invention relates to a multi-modal semantic and dynamic adaptive industrial policy matching method and system, and relates to the field of information processing technologies, and the method comprises the steps: obtaining enterprise entity data and demand policy information; obtaining an enterprise type and an enterprise historical type based on the enterprise entity data; when the enterprise type is consistent with the enterprise historical type, calculating a policy matching degree based on the enterprise entity data and the demand policy information, and defining the policy matching degree as a final matching degree; when the enterprise type is inconsistent with the historical enterprise type, searching change policy information based on the enterprise type; calculating a change matching degree based on the enterprise entity data and the change policy information, and defining the change matching degree as a final matching degree; and analyzing policy suggestions through the final matching degree and outputting the policy suggestions. The method has the effect of improving the accuracy and reliability of the industrial policy matching result.

Description

Technical Field

[0001] This invention relates to the field of information processing technology, and in particular to a method and system for matching multimodal semantics and dynamic adaptation of industrial policies. Background Technology

[0002] In today's fields of industrial transformation and upgrading, precise policy implementation, high-quality enterprise development, and regional industrial planning, the precise matching of industrial policies with enterprises and industries is a key link to improve the effectiveness of policy implementation, optimize resource allocation, and promote coordinated industrial development.

[0003] Currently, specialized information processing devices for this type of matching have emerged in the industry. Their core is to automatically match industrial policies with enterprise information by integrating multi-source data acquisition modules, basic semantic matching modules, and policy database management modules. Combined with data update and result push modules, the process from information collection to matching result output is completed, replacing the traditional model of manually screening policies and manually comparing enterprises with policy conditions. This lays the foundation for improving policy matching efficiency and reducing policy implementation costs.

[0004] Regarding the aforementioned technologies, in the actual matching process, matching is based solely on basic semantics (i.e., keywords) without delving into actual industry data for verification. This leaves room for improvement in the reliability of the output suggestions and the accuracy of the final matching. Summary of the Invention

[0005] To improve the accuracy of matching results and the reliability of policy modification suggestions, this invention provides a multimodal semantic and dynamic adaptation method and system for matching industrial policies.

[0006] In a first aspect, the present invention provides a method for matching industrial policies with multimodal semantics and dynamic adaptation, employing the following technical solution: A multimodal semantic and dynamically adaptable method for matching industrial policies includes: Step 1: Obtain enterprise entity data and policy information; Step 2: Obtain the enterprise type and historical enterprise type based on the enterprise entity data; Step 3: When the enterprise type is consistent with the enterprise's historical type, calculate the policy matching degree based on the enterprise entity data and demand policy information and define it as the final matching degree; Step 4: When the enterprise type is inconsistent with the enterprise's historical type, search for policy information on changes based on the enterprise type; Step 5: Calculate the change matching degree based on enterprise entity data and change policy information, and define it as the final matching degree; Step 6: Analyze the final matching degree and provide policy recommendations.

[0007] By adopting the above technical solution, the system determines whether an enterprise is undergoing transformation by acquiring actual enterprise data. If the enterprise is not transforming, the system calculates the policy matching degree with demand policy information. If the enterprise is transforming, the system searches for new policy changes and calculates the change matching degree. Finally, policy recommendations are generated based on the obtained matching degree. This makes the matching of industrial policies not limited to keywords, reduces the possibility of unreliable matching results due to the failure to analyze the actual data of enterprises when matching industrial policies, and improves the accuracy and reliability of industrial policy matching.

[0008] Optionally, methods for calculating and defining the change matching degree based on enterprise entity data and change policy information include: Step 50: Define demand policy information as historical policy information; Step 51: Calculate the historical matching degree based on enterprise entity data and historical policy information; Step 52: When the historical matching degree is greater than the changed matching degree, define the historical matching degree as the final matching degree and output it; Step 53: When the historical matching degree is less than the changed matching degree, define the changed matching degree as the final matching degree and output it.

[0009] By adopting the above technical solution, when enterprises need to calculate the change matching degree based on the changed policy information during transformation, the historical matching degree is calculated simultaneously by combining the enterprise entity data and historical policy information. By comparing the numerical values ​​of the historical matching degree and the change matching degree, the matching degree with the higher value is selected as the final matching degree. This ensures that the matching of industrial policies after enterprise transformation does not solely refer to the new changed policy information, reducing the possibility of one-sided matching degree selection and unreasonable matching results caused by enterprises directly abandoning the historical policy adaptability analysis during transformation. This improves the accuracy and rationality of industrial policy matching.

[0010] Optionally, when the historical matching degree is greater than the changed matching degree, the method to define the historical matching degree as the final matching degree and output it includes: Step 520: Identify the improved and traditional categories by examining the company's historical categories; Step 521: When the enterprise type is an improved type, calculate the change matching degree based on the enterprise entity data and the change policy information; Step 522: Define the variation matching degree as the final matching degree and output it; Step 523: When the enterprise type is traditional, define the historical matching degree as the final matching degree and output it.

[0011] By adopting the above technical solution, the enterprise's historical categories are broken down into advanced and traditional categories, and the current category of the enterprise is determined. For enterprises in the advanced category, the matching degree is still calculated based on the change policy information and the change matching degree is used as the final matching degree. Only for enterprises in the traditional category, the historical matching degree is used. This reduces the situation where enterprises undergoing transformation are unilaterally bound to historical policy information, ignoring their adaptability to new policies, which may lead to matching results that do not match the enterprise's development trend. This improves the adaptability of the matching results to the enterprise's development.

[0012] Optionally, methods for analyzing and outputting policy recommendations based on the final match degree include: Step 60: Obtain the number of historical categories and the actual number of historical categories for the enterprise; Step 61: Calculate the difference in the number of transformations by comparing the number of historical types of enterprises with the actual number of historical types; Step 62: Find the corresponding conversion threshold based on the enterprise's historical categories; Step 63: When the difference in the number of transformations is less than the conversion threshold, policy recommendations are analyzed and output based on the final matching degree. Step 64: When the difference in the number of transformations exceeds the conversion threshold, analyze non-demand information through the enterprise's historical types and demand policy information; Step 65: Analyze and recommend policy changes based on non-demand information and output the results.

[0013] By adopting the above technical solution, the difference in the number of transformations is calculated by comparing the number of historical types of enterprises with the actual number of historical types. Then, based on the comparison results of the difference in the number of transformations and the corresponding conversion threshold, policy recommendations are output in a differentiated manner. During large-scale transformations, non-demand information is analyzed in a targeted manner and policy modification recommendations are output. This reduces the situation where, when enterprises undergo large-scale transformations, the conventional matching degree analysis logic is still applied to output policy recommendations that are out of touch with the actual development needs of enterprises, thereby further improving the adaptability and practicality of industrial policy recommendations.

[0014] Optionally, methods for analyzing and generating policy recommendations based on non-demand information include: Step 650: Locate the corresponding past policy information through the demand policy information; Step 651: Sort the past policy information according to the preset time rules to obtain the policy information sorting; Step 652: Output past policy information in order of sorting to obtain information on policy changes; Step 653: Analyze policy changes to obtain information on adjacent policy changes and historical policy changes; Step 654: When non-demand information falls into historical policy change scenarios, analyze the final matching degree to provide policy recommendations and output them. Step 655: When non-demand information falls into adjacent policy change scenarios, analyze the non-demand information to make policy change recommendations and output them.

[0015] By adopting the above technical solution, the past policy information corresponding to demand policies is sorted according to time rules and decomposed into different dimensions of policy changes. Combined with the different dimensions of policy changes to which non-demand information belongs, policy recommendations are output in a differentiated manner. This reduces the situation where policy recommendations are output in a single way when non-demand information corresponds to different policy change characteristics, which may lead to inconsistencies with policy development trends. This improves the adaptability of industrial policy recommendations to the laws of policy evolution.

[0016] Optional, also includes: Step 66: Extract enterprise planning information from enterprise entity data; Step 67: Analyze the company's future information based on the company's entity data and planning information; Step 68: Find corresponding future policy information based on the company's future information; Step 69: Calculate the company's future data using future company information; Step 70: Calculate the future matching degree based on future policy information and the company's future data; Step 71: Analyze future improvement suggestions based on future matching degree and output them.

[0017] By adopting the above technical solution, planning information is extracted from enterprise entity data and analyzed to obtain enterprise future information and data. Future policy information is then targeted and the future matching degree is calculated. Based on this, future improvement suggestions are output, reducing the situation where policy suggestions only focus on current needs and are out of touch with the future development direction of enterprises, thus improving the comprehensiveness of policy suggestions.

[0018] Optionally, methods for suggesting future improvements based on future matching scores include: Step 710: Determine the future matching type based on the future matching degree; Step 711: Determine future matching scenarios based on future matching types; Step 712: When the future matching situation is a preset beneficial situation, analyze future improvement suggestions based on the future matching degree; Step 713: When the future matching situation is not a beneficial situation, do not analyze future improvement suggestions based on the future matching degree.

[0019] By adopting the above technical solution, the matching type and matching situation are determined by the future matching degree. Future improvement suggestions are only analyzed and output when the matching situation is beneficial. This reduces the situation where suggestions are blindly output when the matching situation is poor, which may lead to a situation that does not meet the future development needs of the enterprise, and improves the pertinence and effectiveness of future improvement suggestions.

[0020] Optionally, methods for suggesting future improvements based on future matching scores also include: Step 714: Based on future policy information and future matching degree, find the corresponding number of similar enterprises using the same policy and the number of similar enterprises matching the same policy. Step 715: Calculate the policy usage ratio by combining the number of similar enterprises using the policy with the number of similar enterprises matching it; Step 716: When the policy usage ratio reaches the preset beneficial threshold, analyze future improvement suggestions based on future matching degree; Step 717: If the policy use ratio does not reach the beneficial threshold, repeat steps 66 to 71 to obtain recommendations for future improvement.

[0021] By adopting the above technical solution, the number of relevant enterprises can be found and counted by matching future policy information with future trends, and the policy usage ratio can be calculated. Based on the comparison results of the ratio with the beneficial threshold, it is decided whether to analyze and output future improvement suggestions or re-acquire relevant information. This reduces the situation where suggestions are blindly output when the policy usage ratio is too low or the policy is not practical enough, and improves the practicality of future improvement suggestions.

[0022] Optionally, methods for finding corresponding future policy information based on future enterprise information include: Step 680: Search for similar policy information based on the company's future information and pre-defined similarity rules; Step 681: Calculate the similarity matching degree based on the enterprise's future data and similar policy information; Step 682: When the similarity matching degree reaches the preset matching threshold, define the similar policy information as future policy information and output it; Step 683: When the similarity matching degree does not reach the matching threshold, combine the enterprise's future data with similar policy information to filter beneficial policy information and define it as future policy information output.

[0023] By adopting the above technical solution, similar policy information can be found through enterprise future information and similarity rules. The similarity matching degree can be calculated by combining enterprise future data. Future policy information can be screened and defined differently according to whether the matching degree meets the standard. This reduces the situation where the selection of future policy information is one-sided or inconsistent with the enterprise's future development, and improves the adaptability and practicality of future policy information.

[0024] Secondly, this invention provides a multimodal semantic and dynamically adaptable industrial policy matching system, employing the following technical solution: A multimodal semantic and dynamically adaptable industrial policy matching system, comprising: The acquisition module is used to acquire enterprise entity data, demand policy information, and the number of historical enterprise types and the actual number of historical types. The memory is used to store the program of the control method for the multimodal semantic and dynamic adaptation industrial policy matching method described above; The processor loads and executes programs from memory.

[0025] By adopting the above technical solution, the acquisition module realizes the unified acquisition of enterprise and policy-related data. Combined with the storage of the control program of the industrial policy matching method in the memory, the processor loads and executes the program to complete the policy matching and suggestion analysis of the whole process. This realizes the systematic implementation of the whole process of multimodal semantic and dynamic adaptation of industrial policy matching, and improves the system's adaptability to different scenarios such as enterprise transformation and development planning.

[0026] In summary, the present invention has at least one of the following beneficial technical effects: 1. By acquiring actual enterprise data, it is determined whether the enterprise is undergoing transformation. If it is not undergoing transformation, the policy matching degree is calculated with demand policy information. If it is undergoing transformation, new policy change information is searched and the change matching degree is calculated. Finally, policy recommendations are generated based on the obtained policy matching degree or change matching degree, so that enterprise transformation can match the corresponding policies in the first time. This reduces the simplification of industry policy matching by relying solely on keyword search and improves the accuracy and reliability of the final matching results. 2. By calculating the difference in the number of conversions among all enterprises in the industry, we can determine how to generate recommendations based on the policy, making the policy recommendations more aligned with actual needs, reducing the reliance on matching degree alone in recommendation generation, and improving the reliability and authenticity of the policy recommendations; 3. By predicting the company's future plans and estimating future data, and by finding relevant policy information and matching it to generate suggestions, policies become more relevant, the company's development direction becomes clearer, and the efficiency of the company's development is improved. Attached Figure Description

[0027] Figure 1 This is a flowchart of a multimodal semantic and dynamically adapted industrial policy matching method according to an embodiment of this application. Detailed Implementation

[0028] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0029] This invention discloses a method for matching industrial policies using multimodal semantics and dynamic adaptation. (Refer to...) Figure 1 A multimodal semantic and dynamically adaptable industrial policy matching method includes: Step 1: Obtain enterprise entity data and policy information.

[0030] Enterprise entity data refers to various multi-source actual data collected from the entire process of enterprise operation and management, reflecting the true operating status, development characteristics, and business attributes of the enterprise. This includes, but is not limited to, objective data related to the enterprise such as costs, profits, product R&D and production, industry categories, operating scale, and qualification certifications. The system acquires this enterprise entity data by connecting to enterprise management systems, industry databases, official platforms such as business registration and taxation, and enterprise self-reporting portals through multi-source data collection channels, achieving comprehensive collection and integration of various operational and development-related actual data. Demand policy information refers to various industry policy-related information suitable for the enterprise. This demand policy information is obtained by the system directly collecting relevant industry policy information that can be matched with the enterprise through the official websites of national, provincial, municipal, and district governments, industry regulatory databases, and third-party policy aggregation platforms.

[0031] Step 2: Obtain the enterprise type and historical enterprise type based on the enterprise entity data.

[0032] Enterprise type refers to the category identifier obtained by classifying and defining enterprises based on key information reflecting their core business attributes, industry affiliation, and business direction from enterprise entity data. The enterprise type is obtained by the system filtering and parsing the collected enterprise entity data, extracting core characteristic information such as industry affiliation, core business, and business model, and defining the current category of the enterprise. Enterprise historical type refers to the category identifier obtained by classifying and defining the historical operating status of enterprises based on historical key information reflecting their core business attributes, industry affiliation, and business direction from enterprise entity data during past operating stages. The historical enterprise type is obtained by the system filtering and parsing the collected enterprise entity data, extracting historical core characteristic information such as industry affiliation, core business, and business model during past operating stages, and defining the category of the historical operating status of the enterprise.

[0033] Step 3: When the enterprise type is consistent with the enterprise's historical type, calculate the policy matching degree based on the enterprise entity data and demand policy information and define it as the final matching degree.

[0034] Policy matching degree refers to the degree of fit between enterprise entity data and policy information, calculated quantitatively. The calculation method involves the system extracting core features from the enterprise entity data and key clauses such as applicable conditions and support scope from the policy information. A quantitative algorithm, such as a weighted scoring algorithm, is then used to calculate the matching score between the two. The final matching degree refers to the final determined quantitative score for policy matching.

[0035] When the enterprise type is consistent with the enterprise's historical type, it indicates that the enterprise has not transformed and is still developing in line with the original policy direction. Therefore, the policy matching degree is calculated based on the enterprise entity data and demand policy information and defined as the final matching degree.

[0036] Step 4: When the enterprise type is inconsistent with the enterprise's historical type, search for change policy information based on the enterprise type.

[0037] Policy change information refers to various industry policy-related information that is suitable for the current type of enterprise. The system searches for policy change information by combining the enterprise's current type with the official websites of national, provincial, municipal, and district governments, the databases of industry authorities, and third-party policy aggregation platforms to retrieve relevant industry policy information that matches the enterprise.

[0038] When the enterprise type is inconsistent with the enterprise's historical type, it indicates that the enterprise has undergone a transformation between the current statistical node and the historical statistical node. In order to make the results of subsequent calculation of the matching degree more accurate and reliable, the policy information on changes is searched based on the enterprise type.

[0039] Step 5: Calculate the change matching degree based on enterprise entity data and change policy information, and define it as the final matching degree.

[0040] Change matching degree refers to the fit score obtained by the system through quantitative calculation of the degree of fit between enterprise entity data and change policy information. The calculation method for change matching degree here is as follows: the system extracts the core features from the enterprise entity data and key clauses such as applicable conditions and support scope from the change policy information, and calculates the fit score between the two using a quantitative algorithm to obtain the change matching degree. When the system detects a discrepancy between the enterprise type and its historical type, it automatically calculates the change matching degree after finding the change policy information and considers it as the final matching degree.

[0041] Step 6: Analyze the final matching degree and provide policy recommendations.

[0042] Policy recommendations refer to action guidelines generated for enterprises regarding business development and policy application, as well as reference information provided to policymakers for subsequent policy promulgation and optimization. The analysis of policy recommendations involves the system assessing the suitability of policies and enterprises based on the final matching score, clarifying the enterprise's development alignment direction by combining enterprise entity data, and analyzing and deriving action suggestions for policy application and business adjustments. Simultaneously, based on the matching characteristics between enterprises and policies and industry development needs, the system analyzes and extracts reference suggestions for policymakers regarding the focus and scope of application optimization in subsequent policy promulgation. The output of policy recommendations involves the system organizing the action guidelines for enterprises and the promulgation reference suggestions for policymakers according to a specified format, and then pushing the policy recommendations to both the enterprise and policymaker ends through corresponding portals.

[0043] Among them, the methods for calculating the change matching degree based on enterprise entity data and changed policy information and defining it as the final matching degree include: Step 50: Define demand policy information as historical policy information.

[0044] Historical policy information refers to the policy information that was originally intended to be matched before the enterprise type changed. Here, historical policy information is defined as follows: when the enterprise type is inconsistent with its historical type, the system directly defines the acquired policy information as historical policy information.

[0045] Step 51: Calculate the historical matching degree based on enterprise entity data and historical policy information.

[0046] Historical matching degree refers to the degree of fit between the current entity data of an enterprise and the original demand policy information, which is defined as historical policy information, after the enterprise type has changed. The historical matching degree is calculated by the system extracting the core features from the enterprise entity data and key clauses such as applicable conditions and support scope from the historical policy information, and then using the same quantitative algorithm as that used to calculate the policy matching degree to obtain a matching score.

[0047] Step 52: When the historical matching degree is greater than the changed matching degree, define the historical matching degree as the final matching degree and output it.

[0048] When the historical matching degree is greater than the change matching degree, it means that the enterprise is still closer to the pre-transformation policy and should mainly rely on the pre-transformation policy. Therefore, the historical matching degree is defined as the final matching degree and output.

[0049] Step 53: When the historical matching degree is less than the changed matching degree, define the changed matching degree as the final matching degree and output it.

[0050] When the historical matching degree is less than the change matching degree, it indicates that the enterprise transformation is successful and needs to align itself with the policies after the transformation. Therefore, the change matching degree is defined as the final matching degree and output.

[0051] Among them, the methods for defining the historical matching degree as the final matching degree and outputting it when the historical matching degree is greater than the changed matching degree include: Step 520: Find the improved and traditional categories by enterprise history categories.

[0052] "Advanced categories" refer to enterprise categories that have achieved upgrades in technological content, product precision, and industry level based on their historical category. The search method for advanced categories involves pre-classifying and arranging categories from traditional to high-precision within each industry, as determined by professionals in the field. When the system receives a historical match rate greater than the change match rate, it automatically retrieves and matches the corresponding advanced category based on the enterprise's historical category. "Traditional categories" refer to enterprise categories within the same field that are at a lower level in terms of technological content, product precision, and industry level compared to their historical category. The search method for traditional categories involves pre-classifying and arranging categories from traditional to high-precision within each industry, as determined by professionals in the field. When the system receives a historical match rate greater than the change match rate, it automatically retrieves and matches the corresponding traditional category based on the enterprise's historical category. For example, if an enterprise's historical category is general metal parts processing, its advanced category would be high-precision mechanical parts processing, and its traditional category would be rough-machined metal blank production.

[0053] Step 521: When the enterprise type is an improved type, calculate the change matching degree based on the enterprise entity data and change policy information.

[0054] When the enterprise type is "refined", it means that the enterprise is developing towards the refined type. Therefore, the historical matching degree can no longer meet the needs of the enterprise. Thus, the change matching degree is calculated based on the enterprise entity data and the change policy information.

[0055] Step 522: Define the variation matching degree as the final matching degree and output it.

[0056] The final matching degree is output here by the system pushing the changed matching degree in the form of a quantitative value to the policy recommendation analysis module to complete the output, providing core basis for the subsequent generation of policy recommendations.

[0057] Step 523: When the enterprise type is traditional, define the historical matching degree as the final matching degree and output it.

[0058] When the enterprise type is traditional, it indicates that the enterprise's industry level has declined. In order to move the enterprise forward and develop towards a higher-precision type, the historical matching degree is defined as the final matching degree and output.

[0059] The methods for analyzing and outputting policy recommendations based on the final matching degree include: Step 60: Obtain the number of historical types of the enterprise and the actual number of historical types.

[0060] The "Historical Number of Categories" refers to the specific quantity value corresponding to the enterprise's historical categories in the previous statistical period, as calculated by the system. This historical number of categories is obtained by the system extracting category-related data from the enterprise entity data for the previous operating phase and performing specific quantity statistics for each historical category. The "Actual Number of Historical Categories" refers to the specific quantity value corresponding to the enterprise's historical categories at the current statistical node, as calculated by the system. This actual number of historical categories is also obtained by the system extracting category-related data from the enterprise entity data for the current statistical node and performing specific quantity statistics for each historical category.

[0061] Step 61: Calculate the difference in the number of transformations by comparing the number of historical types of the enterprise with the actual number of historical types.

[0062] The transformation quantity difference refers to the difference between the historical quantity of enterprise categories and the actual historical quantity of categories obtained by the system through numerical calculation. The transformation quantity difference is calculated by subtracting the actual historical quantity of categories from the historical quantity of categories as the absolute value of the difference.

[0063] Step 62: Find the corresponding conversion threshold based on the enterprise's historical categories.

[0064] The conversion threshold refers to the quantity difference calculated by taking the minimum guaranteed quantity set by the main business entities in each industry as a benchmark, and then combining it with the current actual statistical quantity of that category. The conversion threshold is found by the following method: each industry has a minimum guaranteed quantity, which is determined in advance by staff in this field through online research and entered into the system. The system then uses the company's historical categories as the search criteria to find the minimum guaranteed quantity for that category in the corresponding industry, and performs numerical calculations based on this, combined with the company's previous historical category quantity for that category, to obtain the conversion threshold.

[0065] Step 63: When the difference in the number of transformations is less than the conversion threshold, policy recommendations are analyzed and output based on the final matching degree.

[0066] When the difference in the number of transformations is less than the conversion threshold, it means that the remaining enterprises of this type can maintain the overall operation of development. There is no need to worry about a certain industry being uninvolved by any enterprise. Therefore, policy recommendations are made and output based on the final matching degree analysis.

[0067] Step 64: When the difference in the number of transformations exceeds the conversion threshold, analyze non-demand information through the enterprise's historical types and demand policy information.

[0068] Non-demand information refers to policy clauses and content in demand policy information that are not required for the development of the enterprise's historical categories or are unrelated to the enterprise's historical category's business development and policy adaptation needs. The analysis method for non-demand information here is that the system uses the development needs of the enterprise's historical categories as the core basis, matching and screening each clause and content of the demand policy information one by one, and extracting the content that is unrelated to the needs of the enterprise's historical categories.

[0069] Step 65: Analyze and recommend policy changes based on non-demand information and output the results.

[0070] Policy modification recommendations refer to targeted guidance suggestions generated for policymakers to adjust, modify, and optimize existing demand-related policy information. The analysis method for these recommendations involves the system systematically reviewing existing policy clauses corresponding to non-demand-related information. This analysis, combined with the characteristics of the enterprise after transformation, actual operational needs, and critical requirements for the number of industry categories, identifies the content in the existing policies that needs adjustment, forming specific policy modification directions and optimization details as policy modification recommendations. The output method involves the system pushing the analyzed policy modification recommendations to the corresponding receiving port of the policymaker.

[0071] The methods for analyzing and proposing policy changes based on non-demand information include: Step 650: Find the corresponding past policy information through the demand policy information.

[0072] Past policy information refers to a collection of existing policy information that corresponds to the current policy needs and is suitable for the company's historical categories. The method for finding past policy information here is as follows: the system uses the policy needs information as the search basis, searches for existing policy information that matches the policy needs information and is suitable for the company's historical categories on compliant policy information platforms such as the official websites of government policies at all levels and the databases of industry authorities, and integrates them into a single set to obtain the past policy information.

[0073] Step 651: Sort the past policy information according to the preset time rules to obtain the policy information sorting.

[0074] Time rules refer to the standards set by professionals in this field according to chronological order. Policy information sorting refers to the ordered sequence of policy information obtained by arranging past policy information in chronological order according to time rules. The policy information sorting here is generated by the system sequentially arranging the retrieved past policy information according to the chronological order standards set by professionals in this field.

[0075] Step 652: Output past policy information in order of policy information to obtain information on policy changes.

[0076] Policy changes refer to adjustments and changes in various aspects of this series of policies, including the content of the clauses, the scope of application, and the direction of support. The method for obtaining policy changes here is that the system outputs past policy information one by one in chronological order according to the policy information, compares and analyzes the differences and adjustments between the various versions of the policy, and summarizes and integrates the data to obtain the corresponding policy changes.

[0077] Step 653: Deconstruct the policy changes to obtain the changes in adjacent policies and historical policy changes.

[0078] Adjacent policy changes refer to the specific differences and adjustments in terms of clause content, scope of application, and support direction between two adjacent past policy documents. This is obtained by the system comparing each group of broken-down policy changes to extract the specific changes in adjacent past policy documents within the policy information sequence. Historical policy changes refer to the overall evolution, cumulative adjustments, and comprehensive changes in terms of clause content, scope of application, and support direction throughout the entire period from the first to the last past policy document in the policy information sequence. This is obtained by the system comprehensively analyzing the broken-down policy changes, integrating all policy adjustments within the entire period, and summarizing their evolutionary patterns.

[0079] Step 654: When non-demand information falls into historical policy change scenarios, policy recommendations are analyzed and output based on the final matching degree.

[0080] When non-demand information falls into historical policy changes, it indicates that these policies, which are detrimental to the industry's development, have already been formulated and the industry is no longer needed in the plan, requiring a comprehensive transformation. Therefore, policy recommendations are made and output through final matching degree analysis.

[0081] Step 655: When non-demand information falls into adjacent policy change scenarios, analyze the non-demand information to make policy change recommendations and output them.

[0082] When non-demand information falls into adjacent policy change scenarios, it indicates that the policy was recently issued and is intended to be piloted in the industry to see if it can help the industry's development, rather than to stop the industry's development. Therefore, policy modification suggestions are analyzed and output based on the non-demand information.

[0083] This also includes: Step 66: Extract enterprise planning information from enterprise entity data.

[0084] Enterprise planning information refers to information recorded in the enterprise entity data regarding various future development plans formulated by the enterprise for its own business development. The extraction method for enterprise planning information here involves the system filtering and extracting relevant content about the enterprise's future development plans from relevant data sections such as business planning, development strategy, and product category layout within the enterprise entity data.

[0085] Step 67: Analyze the company's future information based on the company's entity data and planning information.

[0086] Enterprise future information refers to the trend and directional predictions related to the enterprise's future entity data. The analysis method for this enterprise future information involves the system using the enterprise's existing entity data as a foundation, combined with the enterprise's future development plan information, and introducing trend prediction models, linear regression models, and category development quantitative models to conduct multi-dimensional comprehensive analysis and precise quantitative calculations. By fitting the model to the enterprise's business development patterns, the system deduces the business change trends after the plan is implemented, thus obtaining the corresponding enterprise future information.

[0087] Step 68: Find the corresponding future policy information based on the company's future information.

[0088] Future policy information refers to policies that have been promulgated and are applicable to the company's future business development cycle, matching its future information. This includes policy documents and clauses related to industry support, qualification certification, tax incentives, market access, and other aspects relevant to the company's business development. The method for finding future policy information involves the system matching the analyzed future information of the company with existing long-term effective or phased-implementation policies. This is combined with the company's future development direction, business layout, and planning goals to filter for relevant policy content, while simultaneously predicting the policy clauses applicable to the company in the future.

[0089] Step 69: Calculate the company's future data using the company's future information.

[0090] Enterprise future data refers to the specific numerical indicators of an enterprise's future entity data obtained through precise calculation and quantification, based on its future information. The calculation method for enterprise future data involves the system extracting core content such as trend predictions and development directions from the enterprise's future information, combining this with industry development benchmarks and enterprise operating coefficients to perform refined numerical calculations. This transforms trend-based future information into specific, quantifiable enterprise future entity data indicators, resulting in the corresponding enterprise future data.

[0091] Step 70: Calculate the future matching degree based on future policy information and the company's future data.

[0092] Future matching degree refers to a numerical indicator representing the degree of alignment between a company's future development plan and the corresponding policies, obtained by quantitatively analyzing the applicability requirements and support directions of future policy information with the company's future data development indicators and planning goals. The calculation method for future matching degree involves the system using the industry orientation, quantitative requirements, and support standards in future policy information as a benchmark, combined with various development indicators and planning values ​​in the company's future data, and performing multi-dimensional comparison calculations through a matching quantification algorithm.

[0093] Step 71: Analyze future improvement suggestions based on future matching degree and output them.

[0094] Future improvement suggestions refer to forward-looking action guidelines generated for enterprises' future business development and policy applications. They also serve as reference information for policymakers in the subsequent promulgation and optimization of policies. The analysis of these future improvement suggestions involves the system assessing the compatibility between future policy information and enterprise future data based on future matching scores. Combining this with future enterprise information, the system clarifies the policy adaptation direction for the enterprise's future development, and analyzes forward-looking action suggestions such as policy applications and business plan adjustments that the enterprise can undertake. Simultaneously, based on the matching characteristics between the enterprise's future development and future policies, and the future development needs of the industry, the system analyzes and extracts reference suggestions for policymakers regarding the focus and scope of future policy promulgation. The output of these future improvement suggestions involves the system organizing the action guidelines for enterprises' future development and the future policy optimization reference suggestions for policymakers into a specific format, and then pushing the future improvement suggestions to both the enterprise and policymakers through corresponding portals.

[0095] Among them, the methods for suggesting future improvements based on future matching degree analysis include: Step 710: Determine the future matching type based on the future matching degree.

[0096] Future matching categories refer to specific policy information categories that align with the company's future development. The method for determining future matching categories is as follows: the system first breaks down future policy information into multiple independent policy information categories based on content attributes, and then, based on the multi-dimensional adaptation analysis results of future matching degree, filters out policy information categories that fit the company's future data to obtain the future matching categories.

[0097] Step 711: Determine future matching scenarios based on future matching types.

[0098] Future matching status refers to the comprehensive situation of the actual value, support benefits, and development assistance that various matching policy information can bring to the enterprise's future operation and development. The future matching status is determined by the system based on the enterprise's entity data, combined with the enterprise's future development plan and future data. For the identified future matching types, the system analyzes the support content, applicable rights, and implementation value corresponding to each matching policy information, and integrates them to obtain the actual empowering effect and comprehensive matching benefits of various policies on the enterprise, thus obtaining the corresponding future matching status.

[0099] Step 712: When the future matching situation is a preset beneficial situation, analyze future improvement suggestions based on the future matching degree.

[0100] Beneficial situations refer to matching situations where the policy information corresponding to the future matching category can bring actual support value or positive development benefits to the enterprise's future operation and development. Beneficial situations are obtained by the system analyzing the specific content and applicable conditions of various matching policy information in future matching situations, and then combining this with the enterprise's entity data and future data to determine whether the policy information can bring actual support value and positive development benefits to the enterprise's future operation and development. A positive result indicates a beneficial situation.

[0101] When the future matching result is beneficial, it indicates that the policy can promote the development of the enterprise. Therefore, future improvement suggestions are analyzed based on the future matching degree.

[0102] Step 713: When the future matching situation is not a beneficial situation, do not analyze future improvement suggestions based on the future matching degree.

[0103] If the future matching situation is not beneficial, it means that the policy is not helpful to the company's development. Therefore, even if the matching degree is high, the policy is not needed, and therefore, future improvement suggestions are not analyzed based on the future matching degree.

[0104] The methods for suggesting future improvements based on future matching degree analysis also include: Step 714: Based on future policy information and future matching degree, find the number of similar enterprises using the same policy and the number of similar enterprises matching the same policy.

[0105] The "same type of enterprise usage" refers to the total number of enterprises that actually use or rely on the future policy information. The method for finding this "same type of enterprise usage" is that the system, based on the scope of application and attribute characteristics of the future policy information, searches and filters all enterprises that actually use or rely on the policy within the industry statistics platform and calculates the corresponding total number of enterprises. The "same type of enterprise matching" refers to the total number of enterprises that match the future policy information. The method for finding this "same type of enterprise matching" is that the system, based on the scope of application and attribute characteristics of the future policy information, combined with the criteria for determining the future matching degree, searches and filters all enterprises that match the policy information online and calculates the corresponding total number of enterprises.

[0106] Step 715: Calculate the policy usage ratio by combining the number of similar enterprises using the policy with the number of similar enterprises matching it.

[0107] The policy usage ratio refers to the percentage of enterprises that actually use or rely on the policy among those that match the information about the future policy. The policy usage ratio is calculated by the system using the number of similar enterprises using the policy as the dividend and the number of similar enterprises matching the policy as the divisor, and then performing a division operation to obtain the ratio of the two.

[0108] Step 716: When the policy usage ratio reaches the preset beneficial threshold, analyze future improvement suggestions based on the future matching degree.

[0109] The beneficial threshold refers to the critical value of the policy usage ratio used to determine whether future policy information has practical application value. The beneficial threshold is obtained by researchers in this field through multiple experiments to determine the ratio of the number of beneficial policies to the number of policy matches, and then inputting this value into the system.

[0110] When the policy usage ratio reaches the beneficial threshold, it indicates that the actual application rate of the policy is high. Therefore, future improvement suggestions can be made based on the future matching degree analysis.

[0111] Step 717: If the policy use ratio does not reach the beneficial threshold, repeat steps 66 to 71 to obtain recommendations for future improvement.

[0112] When the policy usage ratio does not reach the beneficial threshold, it indicates that the actual application rate of the policy is low. Therefore, the policy is abandoned and other policies are searched. Thus, steps 66 to 71 are repeated to obtain suggestions for future improvement.

[0113] The methods for finding corresponding future policy information based on future enterprise information include: Step 680: Search for similar policy information based on the company's future information and preset similarity rules.

[0114] Similarity rules refer to the criteria and matching standards for determining the similarity between different policy information and future corporate information. These similarity rules were determined experimentally by professionals in the field. Specifically, they are based on comparing identical or similar policy information between two policies with the total policy information to determine a similarity threshold. This comparison is then integrated to obtain the similarity rules, which are then input into the system. Similar policy information refers to policy information that, after comparison according to the similarity rules, possesses the same or similar characteristics in the core dimensions as the future corporate information. The method for finding similar policy information involves the system using the future corporate information as the core reference, combining the similarity rules to compare the features of policy information and the future corporate information in the core dimensions, and filtering out policy information that meets the similarity threshold requirements.

[0115] Step 681: Calculate the similarity matching degree based on the enterprise's future data and similar policy information.

[0116] Similarity matching degree refers to a numerical indicator of the degree to which a company's future development aligns with similar policy information. The similarity matching degree is calculated by the system using the applicability requirements and support orientation of similar policy information as a benchmark, combined with various development indicators and planning values ​​in the company's future data, and then performing multi-dimensional comparison calculations through an adaptability quantification algorithm.

[0117] Step 682: When the similarity matching degree reaches the preset matching threshold, define the similar policy information as future policy information and output it.

[0118] The matching threshold refers to the critical value used to determine whether a company's future data meets the required standard for compatibility with similar policy information. This matching threshold is obtained by researchers in the field through experiments, identifying the lowest critical value for policy and company matching, and then inputting this value into the system.

[0119] When the similarity matching degree reaches the matching threshold, it indicates that the policy may be consistent with the company's future data and can be used by the company. Therefore, the similar policy information is defined as future policy information and output.

[0120] Step 683: When the similarity matching degree does not reach the matching threshold, combine the enterprise's future data with similar policy information to filter beneficial policy information and define it as future policy information output.

[0121] Beneficial policy information refers to similar policy information that can bring actual positive value and development assistance to the company's future operations. The screening method for beneficial policy information here is that the system combines the company's future data and future development plans, analyzes the specific content, support rights and applicable scenarios of similar policy information that does not meet the matching threshold, determines whether it can bring actual positive benefits to the company's future development, and selects policy information that meets the judgment criteria as beneficial policy information.

[0122] When the similarity matching degree does not reach the matching threshold, it means that the enterprise may not be suitable for the policy and use it. However, in order to avoid blindly screening policy information, we combine the enterprise's future data with similar policy information to screen useful policy information and define it as future policy information output.

[0123] Based on the same inventive concept, embodiments of the present invention provide a multimodal semantic and dynamically adaptable industrial policy matching system.

[0124] One of them, a multimodal semantic and dynamically adaptable industrial policy matching system, includes: The acquisition module is used to acquire enterprise entity data, demand policy information, and the number of historical enterprise types and the actual number of historical types. The memory is used to store the program of a control method for a multimodal semantic and dynamically adapted industrial policy matching method; The processor loads and executes programs from memory.

[0125] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0126] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A multimodal semantic and dynamically adapted industrial policy matching method, characterized in that, include: Step 1: Obtain enterprise entity data and policy information; Step 2: Obtain the enterprise type and historical enterprise type based on the enterprise entity data; Step 3: When the enterprise type is consistent with the enterprise's historical type, calculate the policy matching degree based on the enterprise entity data and demand policy information and define it as the final matching degree; Step 4: When the enterprise type is inconsistent with the enterprise's historical type, search for policy information on changes based on the enterprise type; Step 5: Calculate the change matching degree based on enterprise entity data and change policy information, and define it as the final matching degree; Step 6: Analyze the final matching degree and provide policy recommendations.

2. The multimodal semantic and dynamic adaptation industrial policy matching method according to claim 1, characterized in that, Methods for calculating and defining the final matching degree based on enterprise entity data and information on changing policies include: Step 50: Define demand policy information as historical policy information; Step 51: Calculate the historical matching degree based on enterprise entity data and historical policy information; Step 52: When the historical matching degree is greater than the changed matching degree, define the historical matching degree as the final matching degree and output it; Step 53: When the historical matching degree is less than the changed matching degree, define the changed matching degree as the final matching degree and output it.

3. The multimodal semantic and dynamic adaptation method for matching industrial policies according to claim 2, characterized in that, When the historical matching degree is greater than the changed matching degree, the methods for defining the historical matching degree as the final matching degree and outputting it include: Step 520: Identify the improved and traditional categories by examining the company's historical categories; Step 521: When the enterprise type is an improved type, calculate the change matching degree based on the enterprise entity data and the change policy information; Step 522: Define the variation matching degree as the final matching degree and output it; Step 523: When the enterprise type is traditional, define the historical matching degree as the final matching degree and output it.

4. The multimodal semantic and dynamic adaptation method for matching industrial policies according to claim 1, characterized in that, Methods for analyzing and outputting policy recommendations based on final matching scores include: Step 60: Obtain the number of historical categories and the actual number of historical categories for the enterprise; Step 61: Calculate the difference in the number of transformations by comparing the number of historical types of enterprises with the actual number of historical types; Step 62: Find the corresponding conversion threshold based on the enterprise's historical categories; Step 63: When the difference in the number of transformations is less than the conversion threshold, policy recommendations are analyzed and output based on the final matching degree. Step 64: When the difference in the number of transformations exceeds the conversion threshold, analyze non-demand information through the enterprise's historical types and demand policy information; Step 65: Analyze and recommend policy changes based on non-demand information and output the results.

5. The multimodal semantic and dynamic adaptation method for matching industrial policies according to claim 4, characterized in that, Methods for analyzing and proposing policy changes based on non-demand information include: Step 650: Locate the corresponding past policy information through the demand policy information; Step 651: Sort the past policy information according to the preset time rules to obtain the policy information sorting; Step 652: Output past policy information in order of sorting to obtain information on policy changes; Step 653: Analyze policy changes to obtain information on adjacent policy changes and historical policy changes; Step 654: When non-demand information falls into historical policy change scenarios, analyze the final matching degree to provide policy recommendations and output them. Step 655: When non-demand information falls into adjacent policy change scenarios, analyze the non-demand information to make policy change recommendations and output them.

6. The multimodal semantic and dynamic adaptation method for matching industrial policies according to claim 5, characterized in that, Also includes: Step 66: Extract enterprise planning information from enterprise entity data; Step 67: Analyze the company's future information based on the company's entity data and planning information; Step 68: Find corresponding future policy information based on the company's future information; Step 69: Calculate the company's future data using future company information; Step 70: Calculate the future matching degree based on future policy information and the company's future data; Step 71: Analyze future improvement suggestions based on future matching degree and output them.

7. The multimodal semantic and dynamic adaptation method for matching industrial policies according to claim 6, characterized in that, Methods for suggesting future improvements based on future matching degree analysis include: Step 710: Determine the future matching type based on the future matching degree; Step 711: Determine future matching scenarios based on future matching types; Step 712: When the future matching situation is a preset beneficial situation, analyze future improvement suggestions based on the future matching degree; Step 713: When the future matching situation is not a beneficial situation, do not analyze future improvement suggestions based on the future matching degree.

8. The multimodal semantic and dynamic adaptation method for matching industrial policies according to claim 7, characterized in that, Methods for suggesting future improvements based on future matching degree analysis also include: Step 714: Based on future policy information and future matching degree, find the corresponding number of similar enterprises using the same policy and the number of similar enterprises matching the same policy. Step 715: Calculate the policy usage ratio by combining the number of similar enterprises using the policy with the number of similar enterprises matching it; Step 716: When the policy usage ratio reaches the preset beneficial threshold, analyze future improvement suggestions based on future matching degree; Step 717: If the policy use ratio does not reach the beneficial threshold, repeat steps 66 to 71 to obtain recommendations for future improvement.

9. The multimodal semantic and dynamic adaptation method for matching industrial policies according to claim 6, characterized in that, Methods for finding corresponding future policy information based on future enterprise information include: Step 680: Search for similar policy information based on the company's future information and pre-defined similarity rules; Step 681: Calculate the similarity matching degree based on the enterprise's future data and similar policy information; Step 682: When the similarity matching degree reaches the preset matching threshold, define the similar policy information as future policy information and output it; Step 683: When the similarity matching degree does not reach the matching threshold, combine the enterprise's future data with similar policy information to filter beneficial policy information and define it as future policy information output.

10. A multimodal semantic and dynamically adaptable industrial policy matching system, characterized in that, include: The acquisition module is used to acquire enterprise entity data, demand policy information, and the number of historical enterprise types and the actual number of historical types. A memory for storing the program of the control method for the multimodal semantic and dynamic adaptation industrial policy matching method as described in any one of claims 1 to 9; The processor loads and executes programs from memory.

Images