Intelligent government affair service platform

Abstract

The invention relates to the technical field of task management, in particular to an intelligent government affair service platform which comprises a priority adjustment module, a process adjustment module, an intelligent text mining module and an intelligent matching optimization module. According to the invention, by analyzing historical data and user behaviors completed by government affair service tasks, accurate sorting of task priorities is realized, the task processing efficiency is effectively improved, the working process can be adjusted in real time when public health emergencies or natural disasters are dealt with, the instantaneity and adaptability of government actions are ensured, and the user experience is improved. Government affair text data are deeply mined, key information is extracted, a decision support system is optimized, the service precision is enhanced, the consistency and efficiency of the working process are remarkably improved by accurately analyzing the matching degree of text content and working requirements and optimizing information display, the processing time and potential errors are reduced, and the working efficiency is improved. Therefore, the overall response speed and quality of government service are improved.

Description

Technical Field

[0001] The present invention relates to the field of task management technology, and in particular to a smart government service platform. Background Art

[0002] The field of task management technology focuses on the development and implementation of tools and systems to help individuals and teams organize, coordinate and track various tasks and activities more effectively. It includes software and hardware solutions designed to improve productivity, optimize resource allocation and enhance decision-making processes. Core functions typically include task allocation, progress tracking, resource management, collaboration tools and data analysis systems that can support everything from simple personal to-do lists to complex enterprise-level project management.

[0003] Among them, the smart government service platform is an integrated solution that aims to optimize the provision of government services by using digitalization, automation and data-driven decision-making. It integrates multiple functions such as task management, data processing, process automation and citizen interaction to support government agencies to manage internal operations and provide public services more efficiently. The main uses include simplifying administrative processes, improving public transparency, strengthening interaction between the government and citizens, and improving the accessibility and responsiveness of government services. Through the smart government platform, the government can better respond to public needs, optimize resource allocation, and solve public problems in a timely manner.

[0004] Traditional government service platforms are slow to respond when dealing with emergencies, especially in the case of public health emergencies or natural disasters. The fixed workflow of traditional platforms is difficult to quickly adapt to changes in the situation, affecting the timeliness and effectiveness of processing. Traditional platforms also show limitations in analyzing and utilizing large-scale government text data. They are unable to quickly extract useful information from complex data, resulting in insufficient information support in the decision-making process. This lack of information extraction and analysis limits the government's ability to accurately match government needs and optimize resource allocation, affecting service quality and public satisfaction. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and propose a smart government service platform.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solution: a smart government service platform, the platform comprising:

[0007] The priority adjustment module is based on the current government service task list of the government service platform, combined with user behavior, to analyze the historical data of government service task completion, and calculate the urgency index of official document approval tasks, evaluate the priority of differentiated tasks, and prioritize them to obtain an optimized task list;

[0008] The process adjustment module analyzes the data impact of public health events and natural disaster emergencies based on the optimized task list, dynamically adjusts the workflow, and obtains an adaptive process configuration;

[0009] The intelligent text mining module extracts and analyzes key information from contract documents and resume information based on service request government text data, identifies and sorts keywords and phrases in the text, analyzes the importance and relevance of each information item, calculates information value indicators, and obtains text information analysis results;

[0010] Based on the text information analysis results and adaptive process configuration, the intelligent matching optimization module extracts and classifies the text content required for the workflow, calculates the matching degree between the text content and the government work content, adjusts the display of the text content in the workflow, and generates text matching information.

[0011] The present invention is improved in that the method for calculating the urgency index of the document approval task is:

[0012] Based on the historical data of user behavior and government service task completion, a historical behavior dataset is constructed to extract features associated with document approval tasks, including task submission time, completion time, and user feedback, to obtain a historical feature set.

[0013] Based on the historical feature set, the formula:

[0014]

[0015] Calculate the time sensitivity score S of the task t ;

[0016] Among them, S t is the time sensitivity score of the task, t start and t finish Represent the start and completion time of the task, k S is the curve steepness adjustment parameter, b S and d S are the offset parameter and the square sum base, respectively, and e is the base of the natural logarithm;

[0017] Based on the time sensitivity score and the impact of user feedback on urgency, the formula is:

[0018]

[0019] Calculate the urgency index U to obtain the urgency index of the document approval task;

[0020] Among them, U represents the urgency index of the task, F u represents the rating of task urgency in user feedback, αU and β U is the weight parameter, γ U is a normalization coefficient used to balance the impact of different task types.

[0021] The present invention is improved in that the steps of obtaining the optimization task list are:

[0022] Based on the urgency index of the document approval task, the formula is:

[0023]

[0024] Calculate the priority of task processing and obtain the task priority P;

[0025] Among them, P is the task priority, U is the urgency index, a is the weight parameter of the urgency, b is the baseline offset, c is the adjustment parameter, and e is the base of the natural logarithm;

[0026] According to the task priority P, by comparing the P values ​​of the tasks, the tasks are sorted in descending order to obtain an optimized task list.

[0027] The present invention is improved in that the method for analyzing the data impact of public health events and natural disaster emergencies is:

[0028] Collect historical data on public health events and natural disasters, including the time, duration, scope and severity of the events, to generate a historical data set;

[0029] Based on the historical data set, the event data associated with the current optimization task list is filtered, and the severity and spatiotemporal correlation of the event are combined, and the formula is used:

[0030] W=α W ×I+β W ×D+γ W ×T

[0031] The data is weighted to obtain weighted event data:

[0032] Among them, W is the weighted event data, I represents the severity of the event, D represents the duration of the event, T represents the proximity of the event to the current task time, and α W , β W and γ W is the weight coefficient;

[0033] Based on the weighted event data, by the formula:

[0034]

[0035] Calculate the impact Z to assess the impact of the event on the current workflow;

[0036] Among them, Z is the influence, K is the influence sensitivity adjustment parameter, μ is the model offset parameter, and e is the base of the natural logarithm.

[0037] The present invention is improved in that the step of obtaining the adaptive process configuration is:

[0038] Based on the impact degree Z, identifying workflow parameters that need to be adjusted, including task resource parameters and decision thresholds, and obtaining adjustment requirements;

[0039] Based on the adjustment requirements, the formula:

[0040] C new =C old ×(1+λ×Z+σ×Z 2 )

[0041] Calculate updated configuration parameters;

[0042] Among them, C new is the updated configuration parameter, C old is the current workflow configuration parameter, λ is the adjustment sensitivity coefficient, σ is the adjustment smoothing coefficient, and Z is the influence degree;

[0043] The updated configuration parameters are implemented into the actual workflow, and the workflow is dynamically adjusted to obtain an adaptive workflow configuration.

[0044] The present invention is improved in that the method for marking and sorting keywords and phrases in a text is:

[0045] Based on the service request government text data, the government text data is standardized and preprocessed, including removing stop words, punctuation marks and irrelevant characters to obtain a cleaned text dataset;

[0046] Based on the cleaned text dataset, the word frequency-inverse document frequency method is applied, through the formula:

[0047]

[0048] Calculate the importance score of each word in the text and identify the keywords in the text;

[0049] Where TF-IDF(t) is the importance score of term t, TF(t) is the frequency of term t in the document, DF(t) is the frequency of documents containing term t, and N is the total number of documents;

[0050] Based on the cleaned text data set, using the statistical association analysis method, through the formula:

[0051]

[0052] Calculate the relevance score of phrases and identify key phrases in the text;

[0053] Where Assoc(p) is the relevance score, p is the candidate phrase, and p i are other phrases that co-occur with p, Freq(p i ,p) and Freq(p i ) are the co-occurrence frequency and independent frequency, α A is the smoothing parameter, J is the total number of phrases;

[0054] The keywords and key phrases are sorted according to the importance scores and relevance scores to generate a sorted list of keywords and phrases.

[0055] The present invention is improved in that the steps of obtaining the text information analysis result are:

[0056] Based on the importance score and relevance score, the formula is:

[0057] S=λ 1 ×TF-IDF+λ 2 ×Assoc

[0058] Calculate a composite score for each keyword and phrase;

[0059] Among them, S is the comprehensive score, TF-IDF is the importance score, Assoc is the relevance score, and λ 1 and λ 2 is the weight coefficient;

[0060] Based on the comprehensive score, by the formula:

[0061]

[0062] Calculate the information value of each keyword and phrase, summarize and sort the information value of keywords and phrases, obtain a list of information value indicators, measure the value of text content, and generate text information analysis results;

[0063] Among them, V is the information value, Rel(k) is the relevance score between keyword k and the document topic, Pos(k) is the position factor of the keyword in the document, Length(k) is the length of keyword k, α is the adjustment parameter, and e is the base of the natural logarithm.

[0064] The present invention is improved in that the step of obtaining the text matching information is:

[0065] Based on the text information analysis results and the adaptive process configuration, extract the text content required by the workflow, classify the text content according to the differentiated links of the workflow, and obtain the classified text content;

[0066] Based on the classified text content, the formula is:

[0067]

[0068] Calculate the matching score M to identify the matching degree between the text content and the government work content;

[0069] Among them, M is the matching score, W i Indicates the text content T i The weight of Sim(T i ,G) represents the text content T i Similarity with the government work content G, Rel(G j ) indicates the content of government affairs work G j relevance, n represents the number of text contents, m represents the number of work contents, i represents the text content index, and j represents the work content index;

[0070] According to the matching score M, by the formula:

[0071]

[0072] Calculate the display priority G, adjust the display order of the text content in the workflow, and obtain text matching information;

[0073] Among them, Imp(T) is the importance of the text content, Pos(T) is the position of the text content in the workflow, and β G and γ G To adjust the parameters, G is the display priority, which indicates the display order of text content in the workflow.

[0074] Compared with the prior art, the advantages and positive effects of the present invention are:

[0075] In the present invention, by analyzing the historical data and user behavior of completed government service tasks, accurate sorting of task priorities is achieved, effectively improving the efficiency of task processing. When responding to public health emergencies or natural disasters, the work process can be adjusted in real time to ensure the immediacy and adaptability of government actions, deeply mine government text data and extract key information, optimize the decision support system, enhance the accuracy of services, and significantly improve the consistency and efficiency of the work process by accurately analyzing the degree of match between text content and work needs and optimizing information display, reducing processing time and potential errors, thereby improving the overall response speed and quality of government services. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 is a flow chart of the platform of the present invention;

[0077] Figure 2 A flow chart of calculating the urgency index of a document approval task according to the present invention;

[0078] Figure 3 A flowchart for obtaining an optimization task list for the present invention;

[0079] Figure 4 Flowchart for analyzing data impacts of public health events and natural disaster emergencies for the present invention;

[0080] Figure 5 A flow chart for obtaining an adaptive process configuration for the present invention;

[0081] Figure 6 A flow chart for identifying and sorting keywords and phrases in a text for the present invention;

[0082] Figure 7 A flowchart of obtaining text information analysis results for the present invention;

[0083] Figure 8 The present invention is a flowchart for obtaining text matching information. DETAILED DESCRIPTION

[0084] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0085] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating positions or positional relationships, are based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0086] Example

[0087] See also Figure 1 The present invention provides a technical solution: a smart government service platform, the platform includes:

[0088] The priority adjustment module is based on the current government service task list of the government service platform, combined with user behavior, to analyze the historical data of government service task completion, and calculate the urgency index of official document approval tasks, evaluate the priority of differentiated tasks, and prioritize them to obtain an optimized task list;

[0089] The process adjustment module analyzes the data impact of public health events and natural disaster emergencies based on the optimized task list, dynamically adjusts the workflow, and obtains an adaptive process configuration;

[0090] The intelligent text mining module extracts and analyzes key information from contract documents and resume information based on service request government text data, identifies and sorts keywords and phrases in the text, analyzes the importance and relevance of each information item, calculates information value indicators, and obtains text information analysis results;

[0091] Based on the text information analysis results and adaptive process configuration, the intelligent matching optimization module extracts and classifies the text content required for the workflow, calculates the matching degree between the text content and the government work content, adjusts the display of the text content in the workflow, and generates text matching information.

[0092] The optimized task list includes the task urgency ranking and the expected completion time of the task. The adaptive process configuration includes the adjusted workflow steps, real-time response rules and emergency handling strategies. The text information analysis results are specifically key information extraction identifiers, text value scores and correlation analysis data. The text matching information includes the matching degree between the text content and the work item, the optimization parameter set and the work efficiency improvement index.

[0093] See also Figure 2 ,The method for calculating the urgency index of the document approval task is;

[0094] Based on the historical data of user behavior and government service task completion, a historical behavior dataset is constructed to extract features associated with document approval tasks, including task submission time, completion time, and user feedback, to obtain a historical feature set.

[0095] Based on the historical feature set, through the formula:

[0096]

[0097] Calculate the time sensitivity score S of the task t ;

[0098] Among them, S t is the time sensitivity score of the task, t start and t finish Represent the start and completion time of the task, k Sis the curve steepness adjustment parameter, b S and d S are the offset parameter and the square sum base, respectively, and e is the base of the natural logarithm;

[0099] Based on the time sensitivity score and the impact of user feedback on urgency, the formula is:

[0100]

[0101] Calculate the urgency index U to obtain the urgency index of the document approval task;

[0102] Among them, U represents the urgency index of the task, F u represents the rating of task urgency in user feedback, α U and β U is the weight parameter, γ U is a normalization coefficient used to balance the impact of different task types.

[0103] formula:

[0104]

[0105] The meaning and acquisition method of parameters:

[0106] t start and t finish : Represents the start time and completion time of the document approval task. The data is directly obtained from the task tracking database of the government service system.

[0107] k S : A parameter that adjusts the steepness of the curve. Its value is based on statistical analysis of historical data and is determined by optimizing the model to best fit the historical task completion time sensitivity distribution.

[0108] b S and d S :b S is the offset parameter, which is used to adjust the displacement of the curve along the time axis, and d S are the square and cardinality, both of which are obtained through regression analysis of historical data to enhance the adaptability and discrimination of the model for different task types.

[0109] α U and β U : Weight parameter is used to measure the time sensitivity score S t and user feedback u The relative importance in the urgency index U is determined through expert evaluation or analysis of previous cases.

[0110] γ U: Normalization coefficient, used to balance the influence of various factors and ensure that the urgency index U is within a reasonable numerical range, and is adjusted based on the results of model validation.

[0111] Calculation example:

[0112] Assume that the start time of a document approval task is t start = 2 hours, completion time t finish = 5 hours, and assuming k S =0.1, b S =1 and d S =4.

[0113] Calculate the time sensitivity score S t :

[0114]

[0115] Assume that we get the task urgency score F from user feedback u =0.8, and α U =0.6,β U =0.4,γ U =2:

[0116] Calculate the urgency index U:

[0117]

[0118] The urgency of the task can be determined by the urgency index U.

[0119] See also Figure 3 , the steps to obtain the optimization task list are:

[0120] Based on the urgency index of the document approval task, the formula is:

[0121]

[0122] Calculate the priority of task processing and obtain the task priority P;

[0123] Among them, P is the task priority, U is the urgency index, a is the weight parameter of the urgency, b is the baseline offset, c is the adjustment parameter, and e is the base of the natural logarithm;

[0124] According to the task priority P, by comparing the P values ​​of the tasks, the tasks are sorted in descending order to obtain an optimized task list.

[0125] formula:

[0126]

[0127] The meaning and acquisition method of parameters:

[0128] U: Urgency index, obtained from the previous calculation step.

[0129] a: Weight parameter, used to adjust the sensitivity of the urgency index U to the priority. This parameter is usually set by analyzing historical data to find the optimal relationship between urgency and task completion efficiency.

[0130] b: Baseline offset, used to adjust the horizontal position of the priority curve to make the priority distribution more reasonable. This offset can be set based on the average urgency index of historical urgent tasks.

[0131] c: Tuning parameter used to ensure numerical stability and flexibility in priority calculations. This parameter is usually adjusted based on stability indicators in model testing.

[0132] e: The base of the natural logarithm, the constant e is approximately equal to 2.718, which is a standard value in the calculation of exponential functions and is used for smoothing and nonlinear transformation priority values.

[0133] Calculation example:

[0134] Assume that the given parameters are as follows: U=2.5, a=0.8, b=1.0, c=0.05.

[0135] Calculation process:

[0136] According to the formula Substituting the above parameters into the formula:

[0137] Calculate the exponential part:

[0138] -a(U+b)=-0.8(2.5+1.0)=-0.8×3.5=-2.8.

[0139] Calculate the exponential power of e: e -2.8 .

[0140] Using an approximation, e -2.8 ≈0.061.

[0141] E -2.8 Substituting into the formula:

[0142]

[0143] The calculation example shows the use of the urgency index U and other parameters to calculate the priority P of a task, where a high value of P indicates that the task has a high priority and should be processed first.

[0144] See also Figure 4, the method for analyzing the data impact of public health events and natural disaster emergencies is:

[0145] Collect historical data on public health events and natural disasters, including the time, duration, scope and severity of the events, to generate a historical data set;

[0146] Based on the historical data set, filter the event data associated with the current optimization task list, combine the severity and spatiotemporal correlation of the event, and use the formula:

[0147] W=α W ×I+β W ×D+γ W ×T

[0148] The data is weighted to obtain weighted event data:

[0149] Among them, W is the weighted event data, I represents the severity of the event, D represents the duration of the event, T represents the proximity of the event to the current task time, and α W , β W and γ W is the weight coefficient;

[0150] Based on the weighted event data, the formula:

[0151]

[0152] Calculate the impact Z to assess the impact of the event on the current workflow;

[0153] Among them, Z is the influence, K is the influence sensitivity adjustment parameter, μ is the model offset parameter, and e is the base of the natural logarithm.

[0154] The meaning and acquisition method of parameters:

[0155] formula:

[0156] W=α W ×I+β W ×D+γ W ×T

[0157]

[0158] W: Weighted event data. The calculation method involves weighting the severity I of public health events and natural disasters, the number of days D they last, and the temporal proximity T between the event and the current task.

[0159] α W , β W and γ W: Weight coefficient, determined based on statistical analysis of historical data or expert experience, to ensure that the weight reflects the actual impact on the workflow.

[0160] K: Parameter for adjusting the sensitivity of impact, which is used to adjust the sensitivity of impact calculation and can be determined based on model optimization or regression analysis to best fit historical event data and its impact on the workflow.

[0161] μ: Model offset parameter, which is mainly used to adjust the baseline in the calculation process to ensure that the impact calculation when no event occurs does not deviate from the actual situation. It is set based on the median or mean value in the data set.

[0162] Calculation example:

[0163] Assume a natural disaster event with the following parameters:

[0164] Severity I = 7, duration D = 5 days, proximity between the event and the current task time T = 2, weight coefficient α W =0.5,β W =0.3,γ W =0.2, influence sensitivity parameter K=1.5, model offset parameter μ=0.8.

[0165] Calculate weighted event data;

[0166] W=α W ×I+β W ×D+γ W ×T

[0167] =0.5×7+0.3×5+0.2×2

[0168] =3.5+1.5+0.4

[0169] =5.4

[0170] Calculate the impact:

[0171]

[0172] Calculate W·K+μ:

[0173] W·K+μ=5.4×1.5+0.8=8.1+0.8=8.9

[0174] Calculate the exponent part and the denominator:

[0175] e -(8.9) ≈e -8.9 ≈0.0001359

[0176]

[0177] The calculated result is 2.98. The larger the value, the greater the impact of the event.

[0178] See also Figure 5 , the steps to obtain the adaptive process configuration are:

[0179] Based on the impact Z, identify the workflow parameters that need to be adjusted, including task resource parameters and decision thresholds, and obtain adjustment requirements;

[0180] Based on the adjustment requirements, through the formula:

[0181] C new =C old ×(1+λ×Z+σ×Z 2 )

[0182] Calculate updated configuration parameters;

[0183] Among them, C new is the updated configuration parameter, C old is the current workflow configuration parameter, λ is the adjustment sensitivity coefficient, σ is the adjustment smoothing coefficient, and Z is the influence degree;

[0184] The updated configuration parameters are implemented into the actual workflow, and the workflow is dynamically adjusted to obtain an adaptive process configuration.

[0185] formula:

[0186] C new =C old ×(1+λ×Z+σ×Z 2 )

[0187] The meaning and acquisition method of parameters:

[0188] C old :Configuration parameters of the current workflow, indicating the process configuration values ​​that have been set and are in use before adjustment. The parameters are obtained from the organization's process management system or process documentation.

[0189] λ: Adjustment sensitivity coefficient, which is a preset parameter used to adjust the configuration of the workflow according to the impact Z. It determines how the impact affects the configuration change linearly and is set through historical data analysis or simulation testing.

[0190] σ: Adjustment smoothing coefficient, used to refine configuration adjustments, taking into account the quadratic term of the impact to provide smoother configuration adjustments, especially when the impact is high. The coefficient is determined through optimization models or feedback adjustment experiments to ensure that the adjustment does not cause system instability due to sudden changes.

[0191] Z: Impact, calculated from the previous analysis step, reflects the degree of influence of a specific event on the workflow.

[0192] Calculation example:

[0193] Assumption: C old =100, λ=0.1, σ=0.05, Z=3.

[0194] Calculation process:

[0195] Calculate the new configuration parameter C new :

[0196] C new =C old ×(1+λ×Z+σ×Z 2 )

[0197] C new =100×(1+0.1×3+0.05×3 2 )

[0198] C new =100×(1+0.3+0.05×9)

[0199] C new =100×(1+0.3+0.45)

[0200] C new =100×1.75

[0201] C new =175

[0202] The calculation example shows the known influence Z, sensitivity coefficient λ and smoothing coefficient σ, as well as the current configuration value C old , calculate the new workflow configuration C new . Ensure dynamic adaptation of workflows to respond to the impact of external events and improve the responsiveness and efficiency of the overall process.

[0203] See also Figure 6 , the method for identifying and sorting keywords and phrases in the text is:

[0204] Based on the service request government text data, the government text data is standardized and preprocessed, including removing stop words, punctuation marks and irrelevant characters to obtain a cleaned text dataset;

[0205] Based on the cleaned text dataset, the word frequency-inverse document frequency method is applied, through the formula:

[0206]

[0207] Calculate the importance score of each word in the text and identify the keywords in the text;

[0208] Where TF-IDF(t) is the importance score of term t, TF(t) is the frequency of term t in the document, DF(t) is the frequency of documents containing term t, and N is the total number of documents;

[0209] Based on the cleaned text data set, using the statistical correlation analysis method, through the formula:

[0210]

[0211] Calculate the relevance score of phrases and identify key phrases in the text;

[0212] Where Assoc(p) is the relevance score, p is the candidate phrase, and p i are other phrases that co-occur with p, Freq(p i ,p) and Freq(p i ) are the co-occurrence frequency and independent frequency, α A is the smoothing parameter;

[0213] Sort the keywords and key phrases according to importance scores and relevance scores to generate a ranked list of keywords and phrases.

[0214] formula:

[0215]

[0216] The meaning and acquisition method of parameters:

[0217] TF(t): The frequency of term t in the document, that is, the number of times the term appears in a particular document. The parameters are directly obtained from the term frequency statistics of the document.

[0218] DF(t): The frequency of documents containing term t, which indicates the total number of documents containing the term. The parameter is obtained by counting the occurrence of each term in the entire document collection.

[0219] N: Total number of documents, that is, the total number of documents in the entire document collection. This parameter is known in advance or obtained by simple counting.

[0220] p: candidate phrase, indicating the phrase to be analyzed.

[0221] p i : Other phrases that co-occur with p, indicating other phrases that appear in the same document.

[0222] Freq(p i ,p): co-occurrence frequency, indicating the candidate phrase p and the co-occurrence phrase p iThe number of occurrences in the same document is obtained by counting the co-occurrence matrix.

[0223] Freq(p i ): independent frequency, indicating the co-occurrence phrase p i The number of times a word appears in all documents is obtained directly from the word frequency statistics of the document.

[0224] α A : Smoothing parameter, used to ensure that the relevance calculation of low-frequency phrases is more stable. The parameter is usually determined empirically or through verification.

[0225] Calculation example:

[0226] Assumptions: TF(t) = 5, DF(t) = 10, N = 100 documents, candidate phrase p = p co-occurs with p i , candidate phrase p and co-occurrence phrase p i The co-occurrence frequency Freq(p i ,p)=3, co-occurring phrase p i The independent frequency Freq(p i )=20, smoothing parameter α A =1.

[0227] Calculate TF-IDF:

[0228]

[0229] Substitute the parameters:

[0230]

[0231] calculate:

[0232]

[0233] Calculate the correlation:

[0234]

[0235] Assume J = 1 (i.e. there is only one co-occurring phrase) and substitute the parameters:

[0236]

[0237] Assume that the independent frequency of candidate phrase p is Freq(p)=10:

[0238]

[0239] The calculation example shows how to calculate the TF-IDF value of a keyword and the relevance of a candidate phrase based on given parameters and formulas, ensuring the accuracy and rationality of the calculation and providing a basis for subsequent keyword and phrase ranking.

[0240] See also Figure 7 , the steps to obtain the text information analysis results are:

[0241] Based on the importance score and relevance score, the formula is:

[0242] S=λ 1 ×TF-IDF+λ 2 ×Assoc

[0243] Calculate a composite score for each keyword and phrase;

[0244] Among them, S is the comprehensive score, TF-IDF is the importance score, Assoc is the relevance score, and λ 1 and λ 2 are weight coefficients, which are used to adjust the influence of TF-IDF score and relevance score respectively;

[0245] Based on the comprehensive score, by the formula:

[0246]

[0247] Calculate the information value of each keyword and phrase, summarize and sort the information value of keywords and phrases, obtain a list of information value indicators, measure the value of text content, and generate text information analysis results;

[0248] Among them, V is the information value, Rel(k) is the relevance score between keyword k and the document topic, Pos(k) is the position factor of the keyword in the document, Length(k) is the length of keyword k, α is the adjustment parameter, and e is the base of the natural logarithm.

[0249] formula:

[0250] S=λ 1 ×TF-IDF+λ 2 ×Assoc

[0251]

[0252] The meaning and acquisition method of parameters:

[0253] λ 1 : Adjust the weight coefficient of the TF-IDF score, which is determined by model optimization or expert experience to balance the impact of the TF-IDF score in the comprehensive score;

[0254] λ 2: Adjust the weight coefficient of the relevance score, which is determined through model optimization or expert experience to balance the impact of the relevance score in the comprehensive score;

[0255] TF-IDF: The term frequency-inverse document frequency score calculated from the keyword extraction step, indicating the importance of the keyword in the document;

[0256] Assoc: The relevance score calculated from the key phrase identification step, indicating the relevance of the keyword to other phrases;

[0257] S: Comprehensive score, which is the total score obtained by combining TF-IDF and relevance score, and is used for subsequent information value calculation.

[0258] Rel(k): The relevance score between keyword k and the document topic, obtained through topic modeling or similarity calculation, used to measure the relevance between the keyword and the document topic.

[0259] Pos(k): The position factor of keyword k in the document, which indicates the position of the keyword in the document. Words with higher positions usually have higher weights.

[0260] Length(k): The length of keyword k, used to consider the impact of long and short words on information value.

[0261] α: Adjustment parameter used to adjust the impact of keyword length on information value and ensure the balance of keywords of different lengths in the calculation.

[0262] Calculation Example

[0263] Assume TF-IDF is 3, Assoc is 2, and λ 1 =0.5,λ 2 =0.7, Rel(k)=4, Pos(k)=2, Length(k)=5, α=0.1.

[0264] Calculate the comprehensive score S

[0265] S=λ 1 ×TF-IDF+λ 2 ×Assoc

[0266] S = 0.5 × 3 + 0.7 × 2

[0267] S=1.5+1.4

[0268] S=2.9

[0269] Calculate the information value index V

[0270]

[0271]

[0272] The calculation example shows how to calculate the comprehensive score and information value index of keywords based on given parameters and formulas, ensuring the accuracy and rationality of the calculation, thus providing a basis for subsequent information processing and analysis.

[0273] See also Figure 8 , the steps to obtain text matching information are:

[0274] Based on the text information analysis results and the adaptive process configuration, the text content required by the workflow is extracted, and the text content is classified according to the differentiated links of the workflow to obtain the classified text content;

[0275] Based on the classified text content, the formula is:

[0276]

[0277] Calculate the matching score M to identify the matching degree between the text content and the government work content;

[0278] Among them, M is the matching score, W i Indicates the text content T i The weight of Sim(T i ,G) represents the text content T i Similarity with the government work content G, Rel(G j ) indicates the content of government affairs work G j relevance, n represents the number of text contents, m represents the number of work contents, i represents the text content index, and j represents the work content index;

[0279] According to the matching score M, through the formula:

[0280]

[0281] Calculate the display priority G, adjust the display order of the text content in the workflow, and obtain text matching information;

[0282] Among them, Imp(T) is the importance of the text content, Pos(T) is the position of the text content in the workflow, and β G and γ G To adjust the parameters, G is the display priority, which indicates the display order of text content in the workflow.

[0283] formula:

[0284]

[0285] The meaning and acquisition method of parameters:

[0286] W i :Text content T i The weight is set through historical data or expert experience, indicating the importance of text content in calculating the matching degree.

[0287] Sim(T i ,G): text content T i The similarity with the government work content G is usually obtained through a similarity calculation method.

[0288] Rel(G j ):Government affairs work content j The relevance is obtained through topic modeling or expert scoring, which indicates the importance of the work content in the overall government work.

[0289] M: Matching score, indicating the degree of match between the text content and the government work content.

[0290] β G and γ G : Adjustment parameters, used to adjust the index impact, preset according to experts.

[0291] Calculation example:

[0292] Setting: Text content T 1 and T 2 The weight W 1 =2 and W 2 =3, text content T 1 and T 2 The similarity Sim(T 1 ,G)=0.8 and Sim(T 2 ,G)=0.7, the relevance of the government work content G Rel(G 1 )=1 and Rel(G 2 )=2, the importance of the text content Imp(T 1 )=5 and Imp(T 2 )=4, the position of the text content in the workflow Pos(T 1 )=1 and Pos(T 2 )=2, adjust the parameter β G = 0.5 and γ G =1.

[0293] Calculate the matching score M

[0294]

[0295] Calculate the display priority P

[0296]

[0297] The calculation example shows how to calculate the matching score and display priority based on the given parameters and formulas, ensuring the accuracy and rationality of the calculation, thus providing a basis for the subsequent text content display.

[0298] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. Smart government service platform, characterized by: The platform includes: The priority adjustment module is based on the current government service task list of the government service platform, combined with user behavior, to analyze the historical data of government service task completion, and calculate the urgency index of official document approval tasks, evaluate the priority of differentiated tasks, and prioritize them to obtain an optimized task list; The process adjustment module analyzes the data impact of public health events and natural disaster emergencies based on the optimized task list, dynamically adjusts the workflow, and obtains an adaptive process configuration; The intelligent text mining module extracts and analyzes key information from contract documents and resume information based on service request government text data, identifies and sorts keywords and phrases in the text, analyzes the importance and relevance of each information item, calculates information value indicators, and obtains text information analysis results; Based on the text information analysis results and adaptive process configuration, the intelligent matching optimization module extracts and classifies the text content required for the workflow, calculates the matching degree between the text content and the government work content, adjusts the display of the text content in the workflow, and generates text matching information.

2. The smart government service platform according to claim 1 is characterized in that: The method for calculating the urgency index of the document approval task is as follows: Based on the historical data of user behavior and government service task completion, a historical behavior dataset is constructed to extract features associated with document approval tasks, including task submission time, completion time, and user feedback, to obtain a historical feature set. Based on the historical feature set, the formula: Calculate the time sensitivity score S of the task t ; Among them, S t is the time sensitivity score of the task, t start and t finish Represent the start and completion time of the task, k S is the curve steepness adjustment parameter, b S and d S are the offset parameter and the square sum base, respectively, and e is the base of the natural logarithm; Based on the time sensitivity score and the impact of user feedback on urgency, the formula is: Calculate the urgency index U to obtain the urgency index of the document approval task; Among them, U represents the urgency index of the task, F u represents the rating of task urgency in user feedback, α U and β U is the weight parameter, γ U is a normalization coefficient used to balance the impact of different task types.

3. The smart government service platform according to claim 2 is characterized in that: The steps for obtaining the optimization task list are as follows: Based on the urgency index of the document approval task, the formula is: Calculate the priority of task processing and obtain the task priority P; Among them, P is the task priority, U is the urgency index, a is the weight parameter of the urgency, b is the baseline offset, c is the adjustment parameter, and e is the base of the natural logarithm; According to the task priority P, by comparing the P values ​​of the tasks, the tasks are sorted in descending order to obtain an optimized task list.

4. The smart government service platform according to claim 1, characterized in that: The method for analyzing the data impact of public health events and natural disaster emergencies is: Collect historical data on public health events and natural disasters, including the time, duration, scope and severity of the events, to generate a historical data set; Based on the historical data set, the event data associated with the current optimization task list is filtered, and the severity and spatiotemporal correlation of the event are combined, and the formula is used: W=a W ×I+β W ×D+γ W ×T The data is weighted to obtain weighted event data: Among them, W is the weighted event data, I represents the severity of the event, D represents the duration of the event, T represents the proximity of the event to the current task time, and α W , β W and γ W is the weight coefficient; Based on the weighted event data, by the formula: Calculate the impact Z to assess the impact of the event on the current workflow; Among them, Z is the influence, K is the influence sensitivity adjustment parameter, μ is the model offset parameter, and e is the base of the natural logarithm.

5. The smart government service platform according to claim 4 is characterized in that: The steps for obtaining the adaptive process configuration are: Based on the impact degree Z, identifying workflow parameters that need to be adjusted, including task resource parameters and decision thresholds, and obtaining adjustment requirements; Based on the adjustment requirements, the formula: C new =C old ×(1+λ×Z+σ×Z 2 ) Calculate updated configuration parameters; Among them, C new is the updated configuration parameter, C old is the current workflow configuration parameter, λ is the adjustment sensitivity coefficient, σ is the adjustment smoothing coefficient, and Z is the influence degree; The updated configuration parameters are implemented into the actual workflow, and the workflow is dynamically adjusted to obtain an adaptive workflow configuration.

6. The smart government service platform according to claim 1, characterized in that: The method for marking and sorting keywords and phrases in the text is: Based on the service request government text data, the government text data is standardized and preprocessed, including removing stop words, punctuation marks and irrelevant characters to obtain a cleaned text dataset; Based on the cleaned text dataset, the word frequency-inverse document frequency method is applied, through the formula: Calculate the importance score of each word in the text and identify the keywords in the text; Where TF-IDF(t) is the importance score of term t, TF(t) is the frequency of term t in the document, DF(t) is the frequency of documents containing term t, and N is the total number of documents; Based on the cleaned text data set, using the statistical association analysis method, through the formula: Calculate the relevance score of phrases and identify key phrases in the text; Among them, Assoc(p) is the relevance score, p is the candidate phrase, pi is the other phrases that co-occur with p, Freq(pi,p) and Freq(pi) are the co-occurrence frequency and independent frequency respectively, α A is the smoothing parameter, J is the total number of phrases; The keywords and key phrases are sorted according to the importance scores and relevance scores to generate a sorted list of keywords and phrases.

7. The smart government service platform according to claim 6, characterized in that: The steps for obtaining the text information analysis result are: Based on the importance score and relevance score, the formula is: S = λ1×TF-IDF+λ2×Assoc Calculate a composite score for each keyword and phrase; Among them, S is the comprehensive score, TF-IDF is the importance score, Assoc is the relevance score, λ1 and λ2 are weight coefficients; Based on the comprehensive score, by the formula: Calculate the information value of each keyword and phrase, summarize and sort the information value of keywords and phrases, obtain a list of information value indicators, measure the value of text content, and generate text information analysis results; Where V is the information value, Rel(k) is the relevance score between keyword k and the document topic, Pos(k) is the position factor of the keyword in the document, Length(k) is the length of keyword k, α is the adjustment parameter, and e is the base of the natural logarithm.

8. The smart government service platform according to claim 1, characterized in that: The steps for obtaining the text matching information are: Based on the text information analysis results and the adaptive process configuration, extract the text content required by the workflow, classify the text content according to the differentiated links of the workflow, and obtain the classified text content; Based on the classified text content, the formula is: Calculate the matching score M to identify the matching degree between the text content and the government work content; Among them, M is the matching score, W i Indicates the text content T i The weight of Sim(T i ,G) represents the text content T i Similarity with the government work content G, Rel(G j ) indicates the content of government affairs work G j relevance, n represents the number of text contents, m represents the number of work contents, i represents the text content index, and j represents the work content index; According to the matching score M, by the formula: Calculate the display priority G, adjust the display order of the text content in the workflow, and obtain text matching information; Among them, Imp(T) is the importance of the text content, Pos(T) is the position of the text content in the workflow, and β G and γ G To adjust the parameters, G is the display priority, which indicates the display order of text content in the workflow.

Images