Government affair data-driven policy optimization suggestion method and system
By leveraging big data analytics and artificial intelligence technologies, and utilizing multimodal data mining to extract precise policy optimization suggestions from government data, the problem of insufficient utilization of government data has been solved, thereby improving the scientific nature and efficiency of government decision-making.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-05
- Publication Date
- 2026-03-13
AI Technical Summary
Existing technologies are unable to effectively utilize massive amounts of government data, resulting in a lack of scientific rigor and precision in government decision-making. Traditional policy-making methods are also unable to capture dynamic social changes and public needs.
By introducing big data analytics and artificial intelligence technologies, multimodal data is collected, analyzed, and mined to generate policy optimization suggestions using large language models. Machine learning and deep learning algorithms are combined to perform data classification, clustering, and correlation analysis to identify policy issues and generate targeted recommendations.
To improve the scientific rigor and precision of government decision-making, enhance the targeted nature of policies, improve government governance efficiency, and provide data-driven decision support.
Smart Images
Figure CN121660137A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of computer data processing technology, and in particular to a method and system for policy optimization recommendations driven by government data. Background Technology
[0002] With the rapid advancement of information technology and the full arrival of the big data era, government departments have achieved a qualitative leap in their data collection, storage, and processing capabilities. Against this backdrop, government data has experienced explosive growth, covering numerous aspects such as socio-economic development, public services, and market regulation. However, the abundance of data does not equate to the effective utilization of information. Currently, how to extract valuable information from this massive amount of data to further improve the scientific nature of government decision-making and the accuracy of policy implementation has become an urgent problem to be solved.
[0003] Traditional policy-making methods often rely on the experience and intuition of decision-makers, or on inferences based on limited data samples. This approach proves inadequate in complex and ever-changing social environments, often failing to capture timely social dynamics or fully reflect the true needs and desires of the people. Therefore, relying solely on traditional methods is no longer sufficient to meet the demands of modern government decision-making.
[0004] At the same time, large-scale modeling technology has made rapid progress. These large models, possessing hundreds of millions of parameters, are able to capture more complex data features and patterns through deep learning of massive datasets. They have demonstrated astonishing capabilities in fields such as natural language processing, image recognition, and predictive analytics, and are gradually moving from academia to industry, and from the laboratory to the market. In particular, large-scale models have shown unprecedented advantages in processing unstructured data and uncovering potential patterns hidden within large datasets.
[0005] Given the superior performance of large-scale models in data processing and analysis, applying them to in-depth mining of government data and generation of policy optimization recommendations could potentially provide unprecedented data support and scientific basis for government decision-making. By training and optimizing large-scale models, they can accurately identify socio-economic trends, public opinion, and policy implementation effects, thereby providing policymakers with more precise and scientific decision-making references.
[0006] Therefore, leveraging the development of large-scale modeling technology, developing a data-driven intelligent policy optimization system for government affairs is not only technically feasible but also of paramount importance for improving the transparency, scientific rigor, and effectiveness of government decision-making. Such a system can fully utilize government data resources to achieve data-driven decision support, helping government departments better serve the public and promoting the modernization of social governance. Summary of the Invention
[0007] One of the objectives of this invention is to provide a policy optimization suggestion method driven by government data. By introducing advanced big data analysis and artificial intelligence technologies, it realizes the automatic collection, sorting and analysis of government data, providing strong scientific support for government decision-making.
[0008] This invention provides a government data-driven policy optimization suggestion method, comprising:
[0009] Collect multimodal data from multiple government systems;
[0010] Analyze and mine multimodal data to obtain analysis and mining results;
[0011] Based on the large language model and the analysis and mining results, policy optimization recommendations are determined.
[0012] Provide policy optimization suggestions.
[0013] Preferably, the multimodal data includes at least: socioeconomic data, public opinion survey data, and environmental data.
[0014] Preferably, the step of analyzing and mining multimodal data to obtain analysis and mining results includes:
[0015] Based on machine learning and deep learning algorithms, multimodal data is classified, clustered, and analyzed for correlation to obtain analysis and mining results. The analysis and mining results reveal the inherent connections and potential patterns among the data in the multimodal data.
[0016] Preferably, the step of determining policy optimization suggestions based on the large language model and the analysis and mining results includes:
[0017] The large language model identifies problems and shortcomings in the policy implementation process based on the analysis and mining results, assesses the actual impact of policies on various aspects of society, economy, and environment, and generates corresponding policy optimization suggestions.
[0018] Preferred methods for policy optimization recommendations driven by government data also include:
[0019] Track information on the effectiveness of policy optimization recommendations.
[0020] This invention provides a government data-driven policy optimization suggestion system, comprising:
[0021] The collection module is used to collect multimodal data from multiple government systems;
[0022] The analysis and mining module is used to analyze and mine multimodal data to obtain analysis and mining results;
[0023] The determination module is used to determine policy optimization recommendations based on the analysis and mining results of a large language model;
[0024] The output module is used to output policy optimization suggestions.
[0025] An electronic device provided in this embodiment of the invention includes:
[0026] At least one processor; and
[0027] A memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to execute the government data-driven policy optimization recommendation method described above.
[0028] This invention provides a computer-readable storage medium storing computer instructions, which are used to cause a processor to execute the above-described government data-driven policy optimization suggestion method.
[0029] The present invention provides a computer program product, which includes computer instructions. When executed by an electronic device, the electronic device executes the computer program code of the government data-driven policy optimization suggestion method described above.
[0030] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings.
[0031] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description
[0032] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:
[0033] Figure 1 This is a schematic diagram of a policy optimization suggestion method driven by government data in an embodiment of the present invention;
[0034] Figure 2 This is a schematic diagram of a government data-driven policy optimization suggestion system according to an embodiment of the present invention. Detailed Implementation
[0035] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.
[0036] This invention provides a government data-driven method for policy optimization recommendations, such as... Figure 1 As shown, it includes:
[0037] S1. Collect multimodal data from multiple government systems;
[0038] S2. Analyze and mine the multimodal data to obtain the analysis and mining results;
[0039] S3. Based on the large language model, determine policy optimization recommendations according to the analysis and mining results;
[0040] S4. Provide policy optimization suggestions.
[0041] The module comprehensively and automatically collects relevant data from various government systems. The collected data types include, but are not limited to, socioeconomic data, public opinion survey data, and environmental data, providing a rich source of information for subsequent data processing and policy analysis. The module's built-in data cleaning and integration mechanisms automatically remove outliers and duplicate data, ensuring data accuracy and integrity. Advanced machine learning and deep learning algorithms are used to perform in-depth analysis and mining of the collected data. Specifically, the module can perform detailed classification, precise clustering, and in-depth correlation analysis of the data, thereby revealing the inherent connections and potential patterns between data. Combining the understanding capabilities of a large language model, the module conducts in-depth analysis and comprehensive evaluation of current policies. This module can not only identify problems and shortcomings in the policy implementation process but also accurately assess the actual impact of policies on various aspects such as the socio-economic environment. Through policy analysis, we can gain a clearer understanding of the policy's implementation effectiveness and areas for improvement. Based on the results of the policy analysis module and the reasoning capabilities of the large language model, targeted and actionable policy optimization suggestions are generated. These suggestions integrate expert knowledge and rich historical data, aiming to provide the government with practical decision-making support, helping the government adjust and improve current policies, and enhance policy implementation effectiveness and social satisfaction. It fully leverages the powerful capabilities of large-scale natural language processing models, playing a crucial role in multiple stages such as policy text analysis and optimization suggestion generation. Specifically, the large language model can deeply understand the semantics and contextual information of policy texts, providing more accurate policy interpretation and evaluation. Simultaneously, in generating optimization suggestions, the large language model can combine historical data and expert knowledge, engaging in logical reasoning and creative thinking to generate more comprehensive and specific policy recommendations.
[0042] Through the introduction of advanced big data analysis and artificial intelligence technologies, this invention has achieved the automatic collection, collation, and analysis of government affairs data, providing strong scientific support for government decision-making. Its beneficial effects are mainly reflected in the following aspects:
[0043] Improve the scientific nature of decision-making: The system can deeply explore the potential laws and trends in government affairs data, providing a more scientific and accurate basis for government decision-making, and avoiding the subjectivity and blindness that may exist in traditional decision-making methods.
[0044] Enhance the pertinence of policies: Through the precise analysis of government affairs data, the system can help the government formulate policies that are more in line with social reality and the needs of the people, improving the pertinence and effectiveness of policies.
[0045] Improve the efficiency of government governance: The intelligent and automated features of the system can significantly reduce manual intervention, improve the efficiency of government decision-making, and at the same time reduce the risk of decision-making errors.
[0046] Wide application prospects and promotion value: This system is not only applicable to decision-making support in government departments at all levels, but can also be widely applied in fields such as enterprise management and market research, with extremely high promotion value and application prospects.
[0047] In one embodiment, the multi-modal data at least includes: social and economic data, public opinion poll data, environmental data.
[0048] In one embodiment, the analysis and mining of the multi-modal data to obtain the analysis and mining results includes:
[0049] Based on machine learning algorithms and deep learning algorithms, classify, cluster, and associate analyze the multi-modal data to obtain the analysis and mining results; the analysis and mining results reveal the internal connections and potential laws between the data in the multi-modal data. The specific process of analysis and mining is as follows:
[0050] 1. Construct machine learning and deep learning models
[0051] Step 1. Preprocess the text data (this text data refers to the above multi-modal data):
[0052] Clean and standardize the text data for subsequent feature extraction, including: splitting the text into individual words or tokens; cleaning the text data to remove HTML tags, punctuation marks, stop words (such as "de", "he", "shi", etc.); converting all text to a unified format in lowercase.
[0053] Step 2. Construct a vocabulary
[0054] Create a list containing all unique words for subsequent TF-IDF matrix construction, including: extracting all unique words from the preprocessed text data; filtering out stop words and other unwanted words; and compiling the remaining words into a vocabulary.
[0055] Step 3: Calculate the TF-IDF value (Term Frequency-Inverse Document Frequency).
[0056] TF-IDF values are calculated for each word in each document to reflect the word's importance within the document and its uniqueness across the entire dataset. This includes: for each document, counting the frequency of each word within that document, and possibly normalizing the frequency to compare documents of different lengths, to obtain the term frequency (TF); for each word in the vocabulary, calculating the inverse logarithm of the number of documents in which it appears across the entire dataset, to obtain the inverse document frequency (IDF). A higher IDF value indicates a more unique word; and finally, multiplying the TF and IDF values for each word to obtain the TF-IDF value. This value reflects the word's importance within a specific document and its uniqueness across the entire dataset.
[0057] Step 4: Construct the TF-IDF matrix
[0058] The text data is converted into a numerical feature matrix that can be processed by machine learning algorithms. This involves using the vocabulary as the row index of the matrix and the documents as the column index. For each document, the TF-IDF value of each word in the vocabulary is filled into the corresponding matrix element (if a word does not appear in the document, its value is set to 0). The result is a sparse matrix in which most elements are 0 and the non-zero elements represent the TF-IDF values of the words in the document.
[0059] Step 5: Apply to the classification model
[0060] The extracted features are input into a support vector machine (SVM) for text classification. This includes using the extracted feature set as input to train the SVM classification model; adjusting the SVM parameters (such as kernel function, regularization parameters, etc.) to optimize the model's performance; and using the trained model to classify and predict new text data.
[0061] Step Six: Analyze and interpret the classification results:
[0062] Analyzing and interpreting classification results helps in understanding model performance, identifying potential areas for improvement, and evaluating the model's effectiveness in real-world applications. This includes: analyzing the confusion matrix (identifying misclassified samples, assessing the impact of class imbalance, calculating precision, recall, and F1 score); understanding which features contribute most to classification results through feature weights and feature selection, particularly for models using feature extraction (e.g., TF-IDF) and feature-based classifiers (e.g., SVM); and conducting in-depth analysis of misclassified samples by examining the original text and analyzing feature values to understand the reasons for the errors.
[0063] Step 7: Model Optimization
[0064] Based on the above analysis, the model is optimized to improve classification performance, including adjusting model parameters (such as C and gamma parameters in SVM) and improving feature extraction (trying different word segmentation tools, stop word lists, or TF-IDF parameters) to optimize the classification model performance; the trained model is then used to classify and predict new text data.
[0065] Step 8: Association Analysis
[0066] Using an attention mechanism model to perform correlation analysis on policy data and social data includes: constructing an attention mechanism model, receiving vector representations of policy data and social data, calculating the correlation between input data, and generating attention weights to enable the model to focus on key information, encoding input data, and extracting features; attention weight analysis, analyzing the attention weights generated by the attention layer to understand the key information the model focuses on during the correlation analysis process; and using the trained model to perform correlation analysis on policy data and social data to uncover the potential relationships between them.
[0067] Step 9: Constructing the RAG Knowledge Base System
[0068] Relevant policy and social data texts are imported into a single RAG knowledge base. An embedding model is used to convert these texts into vectors, which are then stored in a vector database. When a user enters a query, the system converts it into vector form and retrieves relevant policy texts from the vector database. This knowledge base serves as the foundation for policy big language model analysis, providing it with a wealth of knowledge and information.
[0069] In one embodiment, determining policy optimization recommendations based on a large language model and the analysis and mining results includes:
[0070] The large language model identifies problems and shortcomings in the policy implementation process based on the analysis and mining results, assesses the actual impact of policies on various aspects of society, economy, and environment, and generates corresponding policy optimization suggestions.
[0071] In one embodiment, the government data-driven policy optimization recommendation method further includes:
[0072] Track information on the effectiveness of policy optimization recommendations.
[0073] This invention provides a government data-driven policy optimization suggestion system, such as... Figure 2 As shown, it includes:
[0074] Collection module 1 is used to collect multimodal data from multiple government systems;
[0075] Analysis and mining module 2 is used to analyze and mine multimodal data to obtain analysis and mining results;
[0076] Module 3 is used to determine policy optimization recommendations based on the analysis and mining results of the large language model.
[0077] Output module 4 is used to output policy optimization suggestions.
[0078] This invention provides an electronic device, the electronic device comprising:
[0079] At least one processor; and
[0080] A memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to execute the government data-driven policy optimization recommendation method described above.
[0081] This invention provides a computer-readable storage medium storing computer instructions that, when executed by a processor, implement the government data-driven policy optimization suggestion method described above.
[0082] This invention provides a computer program product, which includes computer instructions. When executed by an electronic device, the electronic device executes the computer program code of the government data-driven policy optimization suggestion method described above.
[0083] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.
Claims
1. A policy optimization suggestion method driven by government data, characterized in that, include: Collect multimodal data from multiple government systems; Analyze and mine multimodal data to obtain analysis and mining results; Based on the large language model and the analysis and mining results, policy optimization recommendations are determined. Provide policy optimization suggestions.
2. The policy optimization suggestion method driven by government data as described in claim 1, characterized in that, The multimodal data includes at least: socioeconomic data, public opinion survey data, and environmental data.
3. The policy optimization suggestion method driven by government data as described in claim 1, characterized in that, The analysis and mining of multimodal data to obtain analysis and mining results includes: Based on machine learning and deep learning algorithms, multimodal data is classified, clustered, and analyzed for correlation to obtain analysis and mining results. The analysis and mining results reveal the inherent connections and potential patterns among the data in the multimodal data.
4. The policy optimization suggestion method driven by government data as described in claim 1, characterized in that, Based on the large language model and the analysis and mining results, policy optimization suggestions are determined, including: The large language model identifies problems and shortcomings in the policy implementation process based on the analysis and mining results, assesses the actual impact of policies on various aspects of society, economy, and environment, and generates corresponding policy optimization suggestions.
5. The policy optimization suggestion method driven by government data as described in claim 1, characterized in that, Also includes: Track information on the effectiveness of policy optimization recommendations.
6. A government data-driven policy optimization suggestion system, characterized in that, include: The collection module is used to collect multimodal data from multiple government systems; The analysis and mining module is used to analyze and mine multimodal data to obtain analysis and mining results; The determination module is used to determine policy optimization recommendations based on the analysis and mining results of a large language model; The output module is used to output policy optimization suggestions.
7. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the government data-driven policy optimization recommendation method according to any one of claims 1-5.
8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the government data-driven policy optimization recommendation method as described in any one of claims 1-5.
9. A computer program product, characterized in that, The computer program product includes computer instructions that, when executed by an electronic device, enable the electronic device to execute computer program code of the government data-driven policy optimization recommendation method as described in any one of claims 1-5.