Intelligent government affair service platform based on AI and big data and management method thereof

Abstract

The invention relates to the technical field of government affair services, and discloses an intelligent government affair service platform based on AI and big data and a management method thereof, and the platform comprises a data collection unit, a feature extraction unit, a service analysis unit, and a result display unit. Structured and unstructured data can be collected, various government affair information sources are covered, and a foundation is laid for accurate analysis; service handling key features can be accurately obtained for structured data, and keywords and emotional tendencies can be effectively extracted for unstructured data through an algorithm and an emotional dictionary; the transaction efficiency analysis is integrated with multiple indexes to judge the business efficiency, and the average duration, the abnormal duration factor and the like are considered; the satisfaction analysis is quantitatively evaluated through text emotional tendency; analysis results are clearly presented to related departments, if the efficiency does not reach the standard, personnel adjustment or procedure simplification can be suggested, and if the satisfaction degree is low, negative keywords can be extracted to help the departments improve services, and the government affair quality and satisfaction degree are improved.

Description

Technical Field

[0001] The present invention relates to the field of government service technology, and in particular to an AI and big data-based smart government service platform and a management method thereof. Background Art

[0002] With the rapid development of information technology, the digital transformation of government services has become an inevitable trend. The traditional government service management model has gradually exposed many problems in the face of growing data volume and diversified service needs.

[0003] On the one hand, government data comes from a wide range of sources, including structured data and unstructured data. These data contain rich information, but in the traditional model, due to the lack of effective integration and analysis methods, the data is often isolated and cannot fully tap its value, making it difficult for government departments to fully and accurately understand the actual situation of government services. On the other hand, for the evaluation of the quality of government services, different types of businesses have different complexities and process requirements. Simple time indicators cannot accurately locate inefficient links. For example, some businesses may be abnormally time-consuming for individual processing personnel, but traditional evaluation methods cannot effectively identify them. At the same time, there is a lack of effective quantitative methods to measure citizens' satisfaction with government services. Citizens' feedback information is mostly presented in text form, which contains a lot of complex emotional tendencies. Traditional analysis methods are difficult to accurately extract citizens' true attitudes towards government services from these unstructured texts, and cannot determine whether citizens are dissatisfied with specific business links or have objections to the overall service quality, thus failing to provide targeted directions for the improvement of government services.

[0004] To sum up, the existing government service management and evaluation model can no longer meet the development needs of government services in the digital age. There is an urgent need for a smart government service platform based on advanced technology to achieve accurate analysis of government data. Summary of the invention

[0005] The purpose of the present invention is to provide a smart government service platform based on AI and big data and a management method thereof, which solves the technical problems raised in the background technology.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] A smart government service platform based on AI and big data, including:

[0008] A data collection unit, used for collecting target data from multiple government data sources, where the data types of the target data include structured data and unstructured data;

[0009] The feature extraction unit is used to extract features from the target data. For structured data, the corresponding features of the business handling type, handling time, and handling personnel information are extracted; for unstructured data, the corresponding features of keywords and sentiment tendencies in the text are extracted;

[0010] The service analysis unit is used to evaluate and analyze the government services based on the results extracted by the feature extraction unit, and to determine whether the efficiency of the corresponding type of business is up to standard based on the evaluation and analysis results, and to determine the satisfaction of citizens with the government services;

[0011] The result display unit is used to display the analysis results of the service analysis unit to relevant departments.

[0012] As a further solution of the present invention: wherein, structured data includes but is not limited to resident information, business processing information, and enterprise registration information in tabular form; unstructured data includes but is not limited to government application documents, text content of citizen consultation feedback, and image videos of government processing sites.

[0013] As a further solution of the present invention: the feature extraction unit is as follows:

[0014] Step T1. Business handling type

[0015] Directly obtain it based on the business classification field in the business processing form;

[0016] Step T2. Processing time

[0017] Select a business to be processed, extract the start time and end time of the business processing, and mark them as TK and TJ respectively;

[0018] Then, the processing time BS of each business is calculated by: BS = TJ-TK;

[0019] And so on, calculate the processing time of each business;

[0020] Step T3, information of the person handling the matter

[0021] Directly obtain from the Handler field in the business handling record;

[0022] Step T4: Keyword extraction

[0023] Use the term frequency-inverse document frequency algorithm to calculate the TF-IDF value of each word in the text data, and then determine the keywords based on the TF-IDF value;

[0024] Step T5: Sentiment extraction

[0025] Extract the preset sentiment dictionary, traverse all the words in the corresponding text data, and then determine and calculate the sentiment tendency of the corresponding text data according to the predetermined sentiment polarity value in the sentiment dictionary.

[0026] As a further solution of the present invention: the specific method in Step T4 is as follows:

[0027] Step T4.1, perform word segmentation on the text data; wherein, word segmentation is to divide the continuous text sequence into independent words according to certain rules;

[0028] Step T4.2, count the number of times each word appears in the document;

[0029] Step T4.3, pass Calculate the frequency TF of the i-th word in the j-th document i，j ;

[0030] Among them, C i，j is the number of times the i-th word appears in the j-th document, i represents the word, i=1, 2, ...n, n represents the number of all words in the j-th document, j represents the document, j=1, 2, ...m, m represents the number of documents;

[0031] Step T4.4, select each word one by one, and count the number of documents containing each word in the document, and record it as D i ;

[0032] StepT4.5, then pass Calculate the inverse document frequency (IDF) of each word i ;

[0033] Step T4.6, pass Calculate the word frequency-inverse document frequency TF-IDF of each word i，j ;

[0034] Step T4.7. TF-IDF i，j Compare with the corresponding preset keyword determination threshold GY;

[0035] If TF-IDF i，j >GY, then the corresponding word is determined to be a keyword; if TF-IDF i，j ≤GY, then the corresponding word is judged not to be a keyword.

[0036] As a further solution of the present invention: wherein the sentiment dictionary contains various sentiment words and corresponding sentiment polarity values; the sentiment polarity value is a quantitative representation of the sentiment tendency contained in the sentiment word, which reflects the positive, negative or neutral degree of the sentiment through a specific numerical value; wherein the positive sentiment is assigned a positive number, which is +1; the negative sentiment is assigned a negative number, which is -1; and the neutral sentiment is assigned a value of 0;

[0037] As a further solution of the present invention: the specific method in Step T5 is as follows;

[0038] Step T5.1, extract the text data after word segmentation, and then compare all the words in it with various sentiment words in the sentiment dictionary one by one to determine whether the corresponding word is in the sentiment dictionary;

[0039] Select a word in the text data;

[0040] If the word is not in the sentiment dictionary, the next word is extracted and compared with various sentiment words in the sentiment dictionary;

[0041] If the word is in the sentiment dictionary, the sentiment word corresponding to the word is obtained in the sentiment dictionary, and then the sentiment polarity value of the sentiment word is extracted, and the sentiment polarity value is assigned to the word in the text data;

[0042] Step T5.2, in the text data, extract all words assigned with sentiment polarity values, and add up the sentiment polarity values ​​assigned to each word to obtain the sum of the sentiment polarity values ​​of the text data;

[0043] Step T5.3, then determine the sentiment tendency of the corresponding text data based on the sum of the sentiment polarity values;

[0044] If the sum of the sentiment polarity values ​​of the text data is greater than 0, the sentiment tendency of the corresponding text data is determined to be positive;

[0045] If the sum of the sentiment polarity values ​​of the text data is less than 0, the sentiment tendency of the corresponding text data is determined to be negative;

[0046] If the sum of the sentiment polarity values ​​of the text data is equal to 0, the sentiment tendency of the corresponding text data is determined to be neutral.

[0047] As a further solution of the present invention: the service analysis unit is as follows:

[0048] Step F1. Processing efficiency analysis

[0049] Determine whether the business processing efficiency meets the standards based on the average processing time of different business types;

[0050] Step F2. Satisfaction Analysis

[0051] Analyze citizens' satisfaction with government services based on the sentiment trends in unstructured data.

[0052] As a further solution of the present invention: the method in Step F1 is as follows:

[0053] StepF1.1, select a business handling type;

[0054] In this business processing type, extract the processing time of all processing businesses and mark it as BS v , v = 1, 2, ... g, g is the number of all the services handled in this service handling type;

[0055] StepF1.2, through Calculate the average processing time BSP for all services in this service type and mark it as the first efficiency evaluation indicator;

[0056] Compare the first efficiency evaluation index BSP with the preset first efficiency index threshold XPY1:

[0057] If BSP>XPY1, it means that the processing efficiency of this business type does not meet the standard, otherwise it meets the standard;

[0058] StepF1.3, through Calculate the standard deviation BSB of the corresponding processing time of all the processing services in this business processing type;

[0059] StepF1.4. The processing time BS corresponding to each processing business in this business processing type v Combine BSP and BSB to analyze the business with abnormal processing time;

[0060] Specifically:

[0061] If BS v -BSP>γ×BSB, the processing time of the corresponding business will be marked as abnormal processing time;

[0062] Wherein, γ is a preset multiple value, and the value of γ is 3;

[0063] Step F1.5, extract all the business transactions marked as abnormal processing time in the business processing type, count their number, and mark their number as f;

[0064] Then, through YB=f / g, calculate the abnormal processing time factor YB in this business processing type and mark it as the second efficiency evaluation indicator;

[0065] Compare the second efficiency evaluation index YB with the preset second efficiency index threshold XPY2:

[0066] If YB>XPY2, it means that the processing efficiency of this business type does not meet the standard, otherwise it meets the standard;

[0067] Step F1.6, extract the first efficiency evaluation index BSP and the second efficiency evaluation index YB;

[0068] Then through Calculate the comprehensive efficiency evaluation index XPZ;

[0069] In the formula, β1 and β2 are preset weight values, μ is the correction factor, and its value is 0.135;

[0070] Compare the comprehensive efficiency evaluation index XPZ with the comprehensive efficiency index threshold XPY0:

[0071] If XPZ>XPY0, it means that the processing efficiency of this business type does not meet the standard, otherwise it meets the standard.

[0072] As a further solution of the present invention: wherein, when the efficiency of the corresponding business type is evaluated based on the first efficiency evaluation index, the second efficiency evaluation index and the comprehensive efficiency evaluation index, the evaluation result gives priority to the comparison result of the comprehensive efficiency evaluation index, followed by the comparison result of the second efficiency evaluation index, and finally the comparison result of the first efficiency evaluation index.

[0073] As a further solution of the present invention: the method in Step F2 is as follows:

[0074] StepF2.1, in all text data within the specified period, mark the total number of texts in all text data as R0;

[0075] Step F2.2, extract the number of texts judged as positive sentiment texts and mark them as R1, and extract the number of texts judged as negative sentiment texts and mark them as R2;

[0076] StepF2.3, then pass and Calculate citizens’ satisfaction MR1 and dissatisfaction MR2 respectively.

[0077] As a further embodiment of the present invention:

[0078] If the business processing efficiency does not meet the standards, suggestions for optimizing business processes will be issued to relevant departments, including adjusting staffing and simplifying processing procedures.

[0079] If citizen satisfaction is low, the keywords of negative feedback will be extracted and displayed to the relevant departments. The relevant departments will make corresponding improvement measures based on the keywords of negative feedback, including strengthening training for specific businesses and improving service attitudes.

[0080] A management method for a smart government service platform based on AI and big data, the method is implemented by the smart government service platform based on AI and big data, and the method comprises the following steps:

[0081] Step 1: Information Collection

[0082] Collect target data from multiple government data sources, where the target data types include structured data and unstructured data;

[0083] Step 2: Feature Extraction

[0084] Extract features from the target data. For structured data, extract the corresponding features of business handling type, handling time, and handling personnel information; for unstructured data, extract the corresponding features of keywords and sentiment tendency in the text;

[0085] Step 3: Service Analysis

[0086] Based on the results of feature extraction, the government services are evaluated and analyzed. Based on the evaluation and analysis results, it is determined whether the efficiency of the corresponding type of business is up to standard, and the citizens' satisfaction with the government services is determined;

[0087] Step 4: Result output

[0088] Used to present service analysis results to relevant departments.

[0089] Beneficial effects of the present invention:

[0090] Comprehensive data collection and utilization: The present invention can collect target data covering structured and unstructured data from multiple government data sources, including but not limited to resident information, business processing information, government application documents, citizen consultation feedback content, government processing site images and videos, etc. This allows government service evaluation to be based on comprehensive and rich data, avoiding inaccurate evaluation problems caused by missing or one-sided data, and providing a solid data foundation for accurate analysis and optimization of government services.

[0091] Accurate feature extraction: For structured data: key features such as business processing type, processing time, and processing personnel information can be accurately extracted. By directly obtaining the business classification field in the business processing form to determine the business processing type, using the start and end time of the business processing to calculate the processing time, and obtaining the processing personnel information from the processing person field in the business processing record, these methods are simple and efficient, and can provide a direct basis for subsequent processing efficiency analysis. For unstructured data: use advanced algorithms to extract keywords and sentiment tendency features. Using the word frequency-inverse document frequency algorithm to determine keywords can highlight the key content of the text; by comparing with the sentiment dictionary containing sentiment polarity values ​​to extract sentiment tendencies, it can accurately quantify the emotional attitudes of citizens towards government services and provide strong support for satisfaction analysis.

[0092] In terms of processing efficiency analysis: taking into account the average processing time and abnormal processing time factors, by comparing with the corresponding threshold, it can not only determine whether the business processing efficiency meets the standard, but also accurately identify the business processing links that may have problems. Comprehensive evaluation is carried out using efficiency comprehensive evaluation indicators, and the priority of the results of different evaluation indicators is clarified, making the evaluation of processing efficiency more scientific and reasonable, which helps to discover potential process problems and optimization directions.

[0093] Satisfaction analysis: By counting the number of positive and negative emotional texts in the text data within a specified period, the citizens' satisfaction and dissatisfaction are calculated, which intuitively reflects the citizens' overall evaluation of government services and provides a quantitative reference for improving service quality.

[0094] Effective result display and feedback mechanism: The result display unit can clearly display the analysis results of the service analysis unit to the relevant departments, realizing efficient information transmission. If the business handling efficiency is not up to standard, it can issue suggestions to the relevant departments to optimize the business process, such as adjusting staffing, simplifying handling procedures, etc. If the citizen satisfaction is low, the negative feedback keywords are extracted and displayed to the relevant departments, prompting them to take improvement measures for the problem, such as strengthening specific business training and improving service attitude, which will help government departments to adjust and optimize services in a timely manner and improve the quality of government services and citizen satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] The present invention will be further described below in conjunction with the accompanying drawings.

[0096] Figure 1 It is a system block diagram of a smart government service platform based on AI and big data in the present invention.

[0097] Figure 2 It is a flow chart of a management method of a smart government service platform based on AI and big data in the present invention. DETAILED DESCRIPTION

[0098] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0099] Embodiment 1

[0100] See also Figure 1 As shown, the present invention is a smart government service platform based on AI and big data and a management method thereof, comprising:

[0101] A data collection unit, used for collecting target data from multiple government data sources, where the target data includes structured data and unstructured data;

[0102] Among them, structured data includes but is not limited to resident information, business processing information, and enterprise registration information in tabular form;

[0103] A feature extraction unit is used to extract features from structured data to extract corresponding features of business handling type, handling time, and handling personnel information;

[0104] The specific method is as follows:

[0105] Step T1. Business handling type

[0106] Directly obtain it based on the business classification field in the business processing form;

[0107] Step T2. Processing time

[0108] Select a business to be processed, extract the start time and end time of the business processing, and mark them as TK and TJ respectively;

[0109] Then, the processing time BS of each business is calculated by: BS = TJ-TK;

[0110] And so on, calculate the processing time of each business;

[0111] Step T3, information of the person handling the matter

[0112] Directly obtain from the Handler field in the business handling record;

[0113] The service analysis unit is used to evaluate and analyze government services based on the results extracted by the feature extraction unit; the evaluation and analysis includes: processing efficiency analysis

[0114] Determine whether the business processing efficiency meets the standards based on the average processing time of different business types;

[0115] The specific method is as follows:

[0116] StepF1.1, select a business handling type;

[0117] In this business processing type, extract the processing time of all processing businesses and mark it as BS v , v = 1, 2, ... g, g is the number of all the services handled in this service handling type;

[0118] StepF1.2, through Calculate the average processing time BSP for all services in this service type and mark it as the first efficiency evaluation indicator;

[0119] Compare the first efficiency evaluation index BSP with the preset first efficiency index threshold XPY1:

[0120] If BSP>XPY1, it means that the processing efficiency of this business type does not meet the standard, otherwise it meets the standard;

[0121] StepF1.3, through Calculate the standard deviation BSB of the corresponding processing time of all the processing services in this business processing type;

[0122] StepF1.4. The processing time BS corresponding to each processing business in this business processing type v Combine BSP and BSB to analyze the business with abnormal processing time;

[0123] Specifically:

[0124] If BS v -BSP>γ×BSB, the processing time of the corresponding business will be marked as abnormal processing time;

[0125] Wherein, γ is a preset multiple value, and the value of γ is 3;

[0126] Step F1.5, extract all the business transactions marked as abnormal processing time in the business processing type, count their number, and mark their number as f;

[0127] Then, through YB=f / g, calculate the abnormal processing time factor YB in this business processing type and mark it as the second efficiency evaluation indicator;

[0128] Compare the second efficiency evaluation index YB with the preset second efficiency index threshold XPY2:

[0129] If YB>XPY2, it means that the processing efficiency of this business type does not meet the standard, otherwise it meets the standard;

[0130] Step F1.6, extract the first efficiency evaluation index BSP and the second efficiency evaluation index YB;

[0131] Then through Calculate the comprehensive efficiency evaluation index XPZ;

[0132] In the formula, β1 and β2 are preset weight values, μ is the correction factor, and its value is 0.135;

[0133] Compare the comprehensive efficiency evaluation index XPZ with the comprehensive efficiency index threshold XPY0:

[0134] If XPZ>XPY0, it means that the processing efficiency of this business type does not meet the standard, otherwise it meets the standard;

[0135] Among them, when the efficiency evaluation of the corresponding business type is carried out according to the first efficiency evaluation index, the second efficiency evaluation index and the comprehensive efficiency evaluation index, the evaluation result shall give priority to the comparison result of the comprehensive efficiency evaluation index, followed by the comparison result of the second efficiency evaluation index, and finally the comparison result of the first efficiency evaluation index;

[0136] The result display unit is used to display the analysis results of the service analysis unit to relevant departments.

[0137] In this embodiment, by comprehensively collecting structured data from government data sources, covering resident information, business processing information, enterprise registration information and other aspects, a rich data foundation is provided for the accurate analysis of government services. By extracting features from structured data, key features such as business processing type, processing time and processing personnel information can be accurately obtained, which helps to gain an in-depth understanding of the specific situation of government services. The service analysis unit conducts multi-dimensional evaluation and analysis of government services, especially processing efficiency analysis. By calculating indicators such as average processing time, standard deviation and abnormal processing time factor, it can accurately judge whether the business processing efficiency meets the standards, providing a scientific basis for optimizing government services. The result display unit displays the analysis results to relevant departments, which is conducive to the relevant departments to understand the status of government services in a timely manner, so as to take targeted improvement measures and improve the quality and efficiency of government services.

[0138] Embodiment 2

[0139] As the second embodiment of the present invention, when the present application is implemented, compared with the first embodiment, the technical solution of the present embodiment is different from that of the first embodiment only in that:

[0140] Target data also includes unstructured data;

[0141] Among them, unstructured data includes but is not limited to government application documents, text content of citizen consultation feedback, and images and videos of government affairs processing sites;

[0142] The feature extraction unit is also used to extract features from unstructured data, extracting keywords and corresponding features of sentiment tendencies in the text;

[0143] Step 1: Use the term frequency-inverse document frequency algorithm to calculate the TF-IDF value of each word in the text data, and then determine the keywords based on the TF-IDF value;

[0144] The specific method is as follows:

[0145] Step T4.1, perform word segmentation on the text data; wherein, word segmentation is to divide the continuous text sequence into independent words according to certain rules;

[0146] Step T4.2, count the number of times each word appears in the document;

[0147] Step T4.3, pass Calculate the frequency TF of the i-th word in the j-th document i，j ;

[0148] Among them, C i，j is the number of times the i-th word appears in the j-th document, i represents the word, i=1, 2, ...n, n represents the number of all words in the j-th document, j represents the document, j=1, 2, ...m, m represents the number of documents;

[0149] Step T4.4, select each word one by one, and count the number of documents containing each word in the document, and record it as D i ;

[0150] StepT4.5, then pass Calculate the inverse document frequency (IDF) of each word i ;

[0151] Step T4.6, through TF-IDF i，j =TF i，j ×IDF i , calculate the word frequency-inverse document frequency TF-IDF of each word i，j ;

[0152] Step T4.7. TF-IDF i，j Compare with the corresponding preset keyword determination threshold GY;

[0153] If TF-IDF i，j >GY, then the corresponding word is determined to be a keyword;

[0154] If TF-IDF i，j ≤GY, then the corresponding word is judged not to be a keyword;

[0155] Step 2: Extract sentiment tendency

[0156] Extract a preset sentiment dictionary, traverse all words in the corresponding text data, and then determine and calculate the sentiment tendency of the corresponding text data according to the predetermined sentiment polarity value in the sentiment dictionary;

[0157] Among them, the sentiment dictionary contains various sentiment words and corresponding sentiment polarity values;

[0158] The sentiment polarity value is a quantitative representation of the sentiment tendency contained in the sentiment word. It reflects the positive, negative or neutral degree of the sentiment through a specific numerical value.

[0159] Among them, positive emotions are assigned a positive number, with a value of +1; negative emotions are assigned a negative number, with a value of -1; neutral emotions are assigned a value of 0;

[0160] The specific method is as follows;

[0161] Step T5.1, extract the text data after word segmentation, and then compare all the words in it with various sentiment words in the sentiment dictionary one by one to determine whether the corresponding word is in the sentiment dictionary;

[0162] Select a word in the text data;

[0163] If the word is not in the sentiment dictionary, the next word is extracted and compared with various sentiment words in the sentiment dictionary;

[0164] If the word is in the sentiment dictionary, the sentiment word corresponding to the word is obtained in the sentiment dictionary, and then the sentiment polarity value of the sentiment word is extracted, and the sentiment polarity value is assigned to the word in the text data;

[0165] Step T5.2, in the text data, extract all words assigned with sentiment polarity values, and add up the sentiment polarity values ​​assigned to each word to obtain the sum of the sentiment polarity values ​​of the text data;

[0166] Step T5.3, then determine the sentiment tendency of the corresponding text data based on the sum of the sentiment polarity values;

[0167] If the sum of the sentiment polarity values ​​of the text data is greater than 0, the sentiment tendency of the corresponding text data is determined to be positive;

[0168] If the sum of the sentiment polarity values ​​of the text data is less than 0, the sentiment tendency of the corresponding text data is determined to be negative;

[0169] If the sum of the sentiment polarity values ​​of the text data is equal to 0, the sentiment tendency of the corresponding text data is determined to be neutral;

[0170] In the service analysis unit, the evaluation analysis also includes: satisfaction analysis

[0171] Analyze citizens’ satisfaction with government services based on sentiment trends in unstructured data;

[0172] StepF2.1, in all text data within the specified period, mark the total number of texts in all text data as R0;

[0173] Step F2.2, extract the number of texts judged as positive sentiment texts and mark them as R1, and extract the number of texts judged as negative sentiment texts and mark them as R2;

[0174] StepF2.3, then pass and Calculate citizens’ satisfaction MR1 and dissatisfaction MR2 respectively.

[0175] Based on the first embodiment, this embodiment adds the collection and processing of unstructured data, including government application documents, text content of citizen consultation feedback, and images and videos of government affairs handling sites, etc., further enriching the data sources and making the analysis results more comprehensive and accurate. The word frequency-inverse document frequency algorithm is used to extract keywords from the text, which can quickly and accurately grasp the core content of the text, providing an important basis for subsequent sentiment tendency analysis and satisfaction analysis. By extracting sentiment tendency, determining the sentiment tendency of text data, and then conducting satisfaction analysis, it is possible to intuitively understand the satisfaction of citizens with government services, and provide a clear direction for relevant departments to improve services.

[0176] Embodiment 3

[0177] As the third embodiment of the present invention, when the present application is specifically implemented, compared with the first and second embodiments, the technical solution of this embodiment is to combine the solutions of the first and second embodiments, and the difference between the technical solution of this embodiment and the first and second embodiments is only that in this embodiment; according to the analysis result of the service analysis unit, a corresponding decision is executed;

[0178] If the business processing efficiency does not meet the standard, suggestions for optimizing business processes will be issued to relevant departments, including adjusting staff allocation, simplifying handling procedures, etc. Among them, the adjustment of staff allocation will be implemented in conjunction with the information of the handling personnel;

[0179] If the citizen satisfaction is low, the keywords of the negative feedback will be extracted and displayed to the relevant departments. The relevant departments will make corresponding improvement measures based on the keywords of the negative feedback, including strengthening training for specific businesses, improving service attitudes, etc.

[0180] This embodiment combines the analysis results of Embodiment 1 and Embodiment 2 to execute corresponding decisions, thereby improving the pertinence and effectiveness of government services. When the business handling efficiency is not up to standard, suggestions for optimizing business processes are put forward, such as adjusting staffing, simplifying handling procedures, etc., which help to improve business handling efficiency and enhance citizen satisfaction. When citizen satisfaction is low, keywords of negative feedback are extracted and displayed to relevant departments. Relevant departments can make corresponding improvement measures for these keywords, such as strengthening training for specific businesses, improving service attitudes, etc., thereby effectively improving the quality of government services.

[0181] Embodiment 4

[0182] As the fourth embodiment of the present invention, when the present application is specifically implemented, compared with the first, second and third embodiments, the technical solution of this embodiment is to combine and implement the solutions of the above-mentioned first, second and third embodiments.

[0183] This embodiment combines the solutions of Embodiment 1, Embodiment 2 and Embodiment 3 to achieve comprehensive and in-depth analysis and accurate decision-making of government services. It can make full use of structured data and unstructured data to evaluate government services from multiple perspectives such as handling efficiency and citizen satisfaction, provide more comprehensive and accurate decision-making basis for relevant departments, and further improve the quality and efficiency of government services.

[0184] See also Figure 2 As shown, the present invention also provides a management method for a smart government service platform based on AI and big data, which is implemented by a smart government service platform based on AI and big data, and the method includes the following steps:

[0185] Step 1: Information Collection

[0186] Collect target data from multiple government data sources, where the target data types include structured data and unstructured data;

[0187] Step 2: Feature Extraction

[0188] Extract features from the target data. For structured data, extract the corresponding features of business handling type, handling time, and handling personnel information; for unstructured data, extract the corresponding features of keywords and sentiment tendency in the text;

[0189] Step 3: Service Analysis

[0190] Based on the results of feature extraction, the government services are evaluated and analyzed. Based on the evaluation and analysis results, it is determined whether the efficiency of the corresponding type of business is up to standard, and the citizens' satisfaction with the government services is determined;

[0191] Step 4: Result output

[0192] Used to present service analysis results to relevant departments.

[0193] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0194] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A smart government service platform based on AI and big data, characterized by: include: A data collection unit, used for collecting target data from multiple government data sources, where the data types of the target data include structured data and unstructured data; The feature extraction unit is used to extract features from the target data. For structured data, the corresponding features of the business handling type, handling time, and handling personnel information are extracted; for unstructured data, the corresponding features of keywords and sentiment tendencies in the text are extracted; The service analysis unit is used to evaluate and analyze the government services based on the results extracted by the feature extraction unit. The evaluation and analysis includes processing efficiency analysis and satisfaction analysis. Then, based on the evaluation and analysis results, it is judged whether the processing efficiency of the corresponding type of business meets the standards, and the satisfaction of citizens with the government services is determined; The result display unit is used to display the analysis results of the service analysis unit to relevant departments.

2. According to claim 1, a smart government service platform based on AI and big data is characterized in that: The feature extraction unit is as follows: Step T1. Directly obtain the business processing type according to the business classification field in the business processing form; Step T2, select a business to be handled, extract the start time and end time of the business handling, and mark them as TK and TJ respectively; Then, the processing time BS of each business is calculated by: BS = TJ-TK; And so on, calculate the processing time of each business; Step T3, directly obtain the information of the person handling the business from the person handling field in the business handling record; Step T4, use the word frequency-inverse document frequency algorithm to calculate the TF-IDF value of each word in the text data, and then determine the keywords based on the TF-IDF value; Step T5, extract the preset sentiment dictionary, traverse all the words in the corresponding text data, and then determine and calculate the sentiment tendency of the corresponding text data according to the predetermined sentiment polarity value in the sentiment dictionary.

3. According to claim 2, a smart government service platform based on AI and big data is characterized in that: The specific method in StepT4 is as follows: Step T4.1, perform word segmentation on the text data; wherein, word segmentation is to divide the continuous text sequence into independent words according to certain rules; Step T4.2, count the number of times each word appears in the document; Step T4.3, pass Calculate the frequency TF of the i-th word in the j-th document i，j ; Among them, C i，j is the number of times the i-th word appears in the j-th document, i represents the word, i=1, 2, ...n, n represents the number of all words in the j-th document, j represents the document, j=1, 2, ...m, m represents the number of documents; Step T4.4, select each word one by one, and count the number of documents containing each word in the document, and record it as D i ; StepT4.5, then pass Calculate the inverse document frequency (IDF) of each word i ; Step T4.6, through TF-IDF i，j =TF i，j ×IDF i , calculate the word frequency-inverse document frequency TF-IDF of each word i，j ; Step T4.

7. TF-IDF i，j Compare with the corresponding preset keyword determination threshold GY; If TF-IDF i，j >GY, then the corresponding word is determined to be a keyword; if TF-IDF i，j ≤GY, then the corresponding word is judged not to be a keyword.

4. According to claim 2, a smart government service platform based on AI and big data is characterized in that: in, The sentiment dictionary contains various sentiment words and their corresponding sentiment polarity values. The sentiment polarity value is a quantitative representation of the sentiment tendency contained in the sentiment words. It reflects the degree of positivity, negativity or neutrality of the sentiment through specific numerical values. Among them, positive sentiment is assigned a positive number with a value of +1; negative sentiment is assigned a negative number with a value of -1; and neutral sentiment is assigned a value of 0.

5. According to claim 4, a smart government service platform based on AI and big data is characterized in that: The specific method in StepT5 is as follows; Step T5.1, extract the text data after word segmentation, and then compare all the words in it with various sentiment words in the sentiment dictionary one by one to determine whether the corresponding word is in the sentiment dictionary; Select a word in the text data; If the word is not in the sentiment dictionary, the next word is extracted and compared with various sentiment words in the sentiment dictionary; If the word is in the sentiment dictionary, the sentiment word corresponding to the word is obtained in the sentiment dictionary, and then the sentiment polarity value of the sentiment word is extracted, and the sentiment polarity value is assigned to the word in the text data; Step T5.2, in the text data, extract all words assigned with sentiment polarity values, and add up the assigned sentiment polarity values ​​of each word to obtain the sum of the sentiment polarity values ​​of the text data; Step T5.3, then determine the sentiment tendency of the corresponding text data based on the sum of the sentiment polarity values; If the sum of the sentiment polarity values ​​of the text data is greater than 0, the sentiment tendency of the corresponding text data is determined to be positive; If the sum of the sentiment polarity values ​​of the text data is less than 0, the sentiment tendency of the corresponding text data is determined to be negative; If the sum of the sentiment polarity values ​​of the text data is equal to 0, the sentiment tendency of the corresponding text data is determined to be neutral.

6. The smart government service platform based on AI and big data according to claim 2 is characterized in that: The processing efficiency analysis method in the service analysis unit is as follows: StepF1.1, select a business handling type; In this business processing type, extract the processing time of all processing businesses and mark it as BS v , v = 1, 2, ... g, g is the number of all the services handled in this service handling type; StepF1.2, calculate the average processing time of all the services in this service type, and mark it as the efficiency first evaluation indicator BSP; StepF1.3, calculate the standard deviation of the corresponding processing time of all the processing services in this business processing type, and mark it as BSB; StepF1.

4. The processing time BS corresponding to each processing business in this business processing type v Combine BSP and BSB to analyze the business with abnormal processing time; Specifically: If BS v -BSP>γ×BSB, the processing time of the corresponding business will be marked as abnormal processing time; Among them, γ is the preset multiple value; Step F1.5, extract all the business transactions marked as abnormal processing time in the business processing type, count their number, and mark their number as f; Then, through YB=f / g, calculate the abnormal processing time factor YB in this business processing type and mark it as the second efficiency evaluation indicator; Step F1.6, extract the first efficiency evaluation index BSP and the second efficiency evaluation index YB; Then through Calculate the comprehensive efficiency evaluation index XPZ; In the formula, β1 and β2 are preset weight values, and μ is the correction factor, which is 0.

135.

7. The smart government service platform based on AI and big data according to claim 6 is characterized in that: Compare the second efficiency evaluation index YB with the preset second efficiency index threshold XPY2: If YB>XPY2, it means that the processing efficiency of this business type does not meet the standard, otherwise it meets the standard; Compare the first efficiency evaluation index BSP with the preset first efficiency index threshold XPY1: If BSP>XPY1, it means that the processing efficiency of this business type does not meet the standard, otherwise it meets the standard; Compare the comprehensive efficiency evaluation index XPZ with the comprehensive efficiency index threshold XPY0: If XPZ>XPY0, it means that the processing efficiency of this business type does not meet the standard, otherwise it meets the standard.

8. The smart government service platform based on AI and big data according to claim 6 is characterized in that: The satisfaction analysis method in the service analysis unit is as follows: StepF2.1, in all text data within the specified period, mark the total number of texts in all text data as R0; Step F2.2, extract the number of texts judged as positive sentiment texts and mark them as R1, and extract the number of texts judged as negative sentiment texts and mark them as R2; StepF2.3, then pass and Calculate citizens’ satisfaction MR1 and dissatisfaction MR2 respectively.

9. The smart government service platform based on AI and big data according to claim 8 is characterized in that: The display results of the result display unit include the first efficiency evaluation index, the second efficiency evaluation index, the comprehensive efficiency evaluation index, and the satisfaction and dissatisfaction of citizens; When conducting efficiency evaluation of corresponding business types based on the first efficiency evaluation indicator, the second efficiency evaluation indicator and the comprehensive efficiency evaluation indicator, the evaluation results shall give priority to the comparison results of the comprehensive efficiency evaluation indicator, followed by the comparison results of the second efficiency evaluation indicator, and finally the comparison results of the first efficiency evaluation indicator.

10. A management method for a smart government service platform based on AI and big data, characterized in that: The method is implemented through an intelligent government service platform based on AI and big data as described in any one of claims 1-9.

Images