Management system, decision support system and equipment based on digital government big data mining analysis

Abstract

The invention discloses a management system, a decision support system and equipment based on digital government big data mining analysis, and belongs to the technical field of data analysis and management. In order to solve the problem that serious problems are weakened due to the fact that common problem texts are submerged in keyword extraction of an existing management system, complaint data are subjected to word segmentation and cleaning firstly, then clustering is carried out based on the cleaned data, and a clustering result is obtained; determining a keyword selection index and extracting keywords based on the number of words x in the text corresponding to the problem reflecting data in which the words x are located, the number of the corresponding texts and the total number of the texts corresponding to the problem reflecting data, taking the keywords as nodes, and obtaining a directed association rule graph of the keywords by utilizing a directed hypergraph spanning tree algorithm, and keyword query and display are provided. On the basis, matching of jurisdiction areas, institutions, organizations and positions and responsibilities is carried out in a question and personnel database based on keywords, and information about future occurrence of similar questions is provided based on Bayesian network prediction.

Description

Technical Field

[0001] The present invention belongs to the technical field of data analysis and management, and particularly relates to an information management system, a decision support system and a device. Background Art

[0002] Now, "Internet + government services" has become the most powerful tool for the modernization of urban governance. Data mining based on the data of "Internet + government services" can timely discover problems in all aspects of urban management, so as to quickly identify problems with a wide range of coverage and a large impact scope among all problems, provide a scientific and reasonable basis for decision-making, so that managers can solve problems in order from important to less important in sequence and in a timely manner based on limited resources and time, or arrange to solve parts of problems and allocate resources for solving problems.

[0003] With the development of Internet technology, at present, management platforms, even grid management platforms, have been basically established in each city, forming a five-level closed-loop linkage of grid, community, sub-district, county and city levels, and realizing the closed-loop management of the 7 stages of "information collection, case establishment, task dispatch, task processing, processing feedback, verification and conclusion, and performance appraisal" for the problem of "whistle blowing and reporting". Although this method provides a data basis for urban managers to solve problems and also improves the efficiency of solving problems encountered in urban management, it is undeniable that there are still many problems to be solved in urban management in aspects such as "transportation", "environmental sanitation", "noise nuisance", etc. Moreover, with the rapid development of the city, the problems to be solved in urban management are on the rise, and the scope involved in the problems encountered is also getting wider and wider, even problems that have never been encountered in previous urban management. This has increased the difficulty of urban management, and problems involving multiple districts and multiple departments will also be encountered, and it is often difficult for each district and each department to quickly form a coordinated and unified cooperation work process. This not only leads to a very long time to solve problems, but also makes it very difficult to solve problems.

[0004] With the rapid development of cities, the explosion of grid management data volume and the increase in data complexity, the data volume shows rapid growth. The data volume of urban management platforms is also astonishing. If only relying on manpower to extract problems one by one in the system, report problems, and then enable managers to arrange relevant departments and mobilize relevant resources to solve problems, the amount of data reported will be astonishing, making the workload of the management part huge. More importantly, in the management platform system, there are often some problems that are considered urgent problems to be resolved by some individual problem reporters, but are not considered urgent problems or even problems to be solved by others. For example, some individuals around the morning market may think that the noise affects their rest and life, so they file complaints in the management platform. However, for most / the vast majority of people, this may not be a problem. That is to say, some situations may be more affected by personal factors. If these are uniformly reported and solved without discrimination in the management platform, not only may administrative resources be wasted, but it may even be impossible to solve the situations that are considered "problems" due to personal factors. More importantly, this method cannot target the impact degree and scope of a certain / some problems, resulting in managers arranging relevant parts and resources to solve problems without prioritization, causing problems with a large impact degree and wide scope not to be processed in a timely manner. This not only unreasonably extends the time to solve problems but also affects the management level and credibility of managers. Of course, this can be handled by information collection and reporting personnel marking the processing levels, but this method can only mark the importance of the "problems" subjectively considered by relevant personnel. Not only is there an impact of subjective factors and cognitive levels, resulting in the processing levels not being objective enough, but this method also cannot reflect the scope of influence of the "problems" (how many people are affected), causing problems that urgently need to be solved due to a wide influence area to be marked with a reduced importance level, unreasonably extending the time to solve problems and also affecting the management level and credibility of managers. In addition, this method of reporting and solving item by item is also not conducive to the unified solution of similar problems. For example, in a certain district, there is a problem of sewage pollution in Area A, and there is the same or similar problem in Area B. If relying on manpower to report one by one / selectively report according to the marked importance level, no matter which method, since it is impossible for manpower to collect and report all information in one day, it may lead to Problem A being reported on a certain day, approved by the manager, and relevant departments and resources being arranged to solve it. On other days, when Problem B is discovered, for the same or similar problems, it needs to be reported again. Then the manager still needs to approve and arrange relevant departments and resources to solve the same or similar matters again. This not only wastes administrative resources to a certain extent but also seriously affects the scheduling efficiency of resources and extends the time to solve problems.

[0005] Of course, with the development of data mining technology, clustering and importance rating of problem - reflecting data can be achieved through computer technology, which can alleviate the above - mentioned problems to a certain extent. However, the current information importance rating methods are not well - suited for the "problem feedback / complaint" function / system / module of the urban management platform. Among them, the TF - IDF (term frequency–inverse document frequency) method, which is commonly used in information processing systems, is often used for mining information importance. However, it should be noted that this information mining method is more suitable for mining important information in documents. Because the IDF statistical method regards documents with lower occurrence frequencies as more important documents, this method will cause information with more complaints to be submerged to a certain extent. However, it is precisely the problems with more complaints that have a greater impact on the lives of the masses and a more serious impact. That is to say, this method will lead to serious problems being weakened, which may seriously affect the information processing of the "problem feedback / complaint" function / system / module of the management platform and even deviate from the actual situation to a certain extent. Summary of the Invention

[0006] In order to solve the problem that the keyword extraction of the existing management system submerges the common - problem text and weakens the processing of serious problems, the present invention proposes a management system based on digital government big data mining and analysis and a corresponding decision - making support system.

[0007] A management system based on digital government big data mining and analysis includes:

[0008] Data acquisition and word - segmentation module: According to the information of the time period to be analyzed, obtain problem - reflecting data from the management platform database. Each piece of problem - reflecting data corresponds to a problem data ID; then perform word - segmentation on the obtained problem - reflecting data and conduct part - of - speech tagging.

[0009] Data cleaning module: Used to clean the word - segmentation results.

[0010] Location area determination module: Based on the cleaned data, taking each piece of problem - reflecting data as a unit, determine the location area where each piece of problem - reflecting data occurs, and "bind" the location area to the problem - reflecting data, that is, establish a mapping relationship between the problem - reflecting data and the location area corresponding to the problem, and store it.

[0011] Basic information database: Store the administrative geographical factors and corresponding regional boundaries within the administrative region, natural geographical factors and corresponding regional boundaries, streets and corresponding segmented regional boundaries, and infrastructure and corresponding geographical locations

[0012] Keyword extraction module: Based on the cleaned data, cluster the feature words according to nouns; then, based on the clustered feature words, extract keywords; in the process of extracting keywords, use keyword selection indicators to select keywords, and the keyword selection indicators are as follows:

[0013]

[0014] Among them, n x represents the number of the word x in the text corresponding to the problem-reflecting data where the word x is located, N represents the total number of texts corresponding to the problem-reflecting data, and T x represents the number of texts corresponding to the problem-reflecting data where the word x is located; λ is a proportionality coefficient.

[0015] Directed association rule graph generation module: Use the keywords as nodes and utilize the directed hypergraph spanning tree algorithm to obtain the directed association rule graph of the keywords;

[0016] Keyword query and display module: Used to receive the query word information input by the manager user, query the keyword information that best matches the query word information, and highlight the directed relationship of the corresponding confirmed keyword nodes in the directed association rule graph of the keywords; or, receive the pointer action information of the manager user in the directed association rule graph of the keywords, confirm the keyword node pointed to by the pointer according to the pointer action information, and highlight the directed relationship of the corresponding confirmed keyword nodes in the directed association rule graph of the keywords.

[0017] Furthermore, in the process of determining the location area where each problem-reflecting data occurs, if the responder of the problem-reflecting data has filled in the location area, directly determine the location area; if the responder of the problem-reflecting data reflects the problem in a direct way of reflecting the problem and has not filled in the determined location area, then determine the location area by means of information extraction.

[0018] Furthermore, the method of determining the location area by the information extraction method uses the method of information matching to determine the location area; the specific process of using the method of information matching to determine the location area is as follows:

[0019] Based on the expression order of the problem-reflecting data, match each word in the word segmentation result with the address feature word. If a word in the word segmentation result can directly match the address feature word completely, or if multiple consecutive words in the word segmentation result can match the address feature word completely, then use the corresponding address feature word as the location information word, and determine the geographical location and spatial range corresponding to the problem reflection based on the location information word.

[0020] Furthermore, the cleaning process of the data cleaning module includes the processing step of removing stop words.

[0021] A management device based on digital government big data mining and analysis. The device includes a processor and a memory. At least one instruction is stored in the memory, and the at least one instruction is loaded and run by the processor to run the management system based on digital government big data mining and analysis.

[0022] A decision support system based on digital government big data mining and analysis management, including:

[0023] Institution and personnel database: Stores the institutional organizational relationships of all departments in each jurisdiction area and the corresponding job responsibilities.

[0024] Requirement information receiving module: Used to obtain the keywords and the directed association rule graph obtained by the directed association rule graph generation module of the management system based on digital government big data mining and analysis described in any one of claims 1 to 4; at the same time, provide keyword options for the user, and provide the corresponding directed association rule graph according to the keyword options;

[0025] First support and suggestion module: Based on the keywords and the corresponding directed association rule graph confirmed by the user, first identify the keywords corresponding to the keyword nodes, and judge whether the word belongs to the word related to "problem" of the reflected problem or the word related to "location" of the reflected problem; then based on the word related to "location" of the reflected problem and the word related to "problem", perform jurisdiction area, institutional organization and job responsibility matching in the institution and personnel database.

[0026] Further, in the process of performing jurisdiction area, institutional organization and job responsibility matching in the institution and personnel database based on the word related to "location" of the reflected problem and the word related to "problem", for the word related to "location" of the reflected problem, according to the reflected problem data corresponding to the keyword reflected by the node belonging to "problem", determine the ID corresponding to the reflected problem data, and correspondingly obtain the location area where the reflected problem data determined in the location area determination module occurs. According to the occurring location area, give the jurisdiction matching recommendation corresponding to the solution that can be implemented.

[0027] Further, the system further includes a second support and suggestion module: Based on the keywords and the corresponding directed association rule graph confirmed by the user, extract the theme and scenario information, obtain all the data of the corresponding complaint cases according to the theme and scenario information determined by the scenario and theme determination module, analyze the data and give the information about the future occurrence of similar problems. The information about the future occurrence includes whether it will occur in the future and / or the expected occurrence location, so as to be used as potential decision support suggestions.

[0028] Further, the process of the second support and suggestion module analyzing the data and giving the information about the future occurrence of similar problems is implemented by using a Bayesian network.

[0029] A decision support device based on digital government big data mining, analysis and management. The device includes a processor and a memory. At least one instruction is stored in the memory, and the at least one instruction is loaded and run by the processor to run the decision support system based on digital government big data mining, analysis and management.

[0030] Beneficial effects:

[0031] The management system based on digital government big data mining and analysis of the present invention performs word segmentation and cleaning on complaint data, and then performs clustering and keyword extraction based on the cleaned data. During this process, the keyword selection index is determined and keywords are extracted based on the number of the word x in the text corresponding to the problem-reflecting data where the word x is located, the total number of texts corresponding to the problem-reflecting data, and the number of texts corresponding to the problem-reflecting data where the word x is located. This can effectively solve the problem that the existing keyword extraction submerges common problem texts and weakens serious problems. The present invention can effectively "reflect" the importance degree of feature words. Therefore, the present invention can effectively extract problems with a large influence area and a more serious influence degree. At the same time, because the management system based on digital government big data mining and analysis of the present invention can accurately and effectively determine the complaint problems faced, a decision support system based on digital government big data mining, analysis and management can effectively provide support for decision-making, can improve the information processing ability of the "problem feedback / complaint" function / system / module of the management platform, and can further enable targeted and efficient matching and response to these problems, thereby providing a basis for improving the management efficiency and management level of managers. Description of the drawings

[0032] Figure 1 It is a screenshot of real complaint case data collected on the urban grid management platform of a certain district in a certain city from July 2022 to July 2023.

[0033] Figure 2 It is a part-of-speech distribution diagram of the complaint case data in a certain district of the management platform.

[0034] Figure 3 It is a feature word table and part-of-speech bar chart of the complaint case data in a certain district of the management platform.

[0035] Figure 4 It is a keyword cloud diagram of the complaint case data in a certain district of the management platform.

[0036] Figure 5 It is an association rule diagram of the complaint cases in a certain district of the management platform.

[0037] Figure 6 It is the node data of the association rule diagram of the keyword (Xipan Community) in the management platform.

[0038] Figure 7 The directed relationship graph for the keyword (Xipan Community).

[0039] Figure 8 The directed association rule graph for the complaint case keywords of Xipan Community in a certain area of the management platform.

[0040] Figure 9 The word cloud graph of the 12345 complaint cases in a certain city from January to August 2023.

[0041] Figure 10 For Figure 9 The bar graph of the number of keywords with the most frequent occurrences in the word cloud graph in

[0042] Figure 11 The directed association rule graph determined based on the 12345 complaint cases.

[0043] Figure 12 The Bayesian network prediction graph for pipelines.

[0044] Figure 13 The Bayesian network prediction graph corresponding to the complaint reasons for gas.

[0045] Figure 14 The Bayesian network prediction graph for gas occurrences in a certain area within a time range. Detailed implementation manner

[0046] The present invention proposes a management system based on big data mining and analysis of digital government, as well as a problem-solving decision support system. The system is an embedded system / subsystem or functional unit of the urban grid management platform, providing a foundation for the urban management platform, guiding urban management services through big data analysis, assisting in building a smart city, and improving the urban management level. For the urban grid management platform, information collection is the starting point of all tasks of the urban governance supervision and command platform, and also the foundation of the system described in the present invention. In terms of the urban grid management platform, the main body of information collection is the general public and / or urban management supervisors.

[0047] If it is a citizen, they can report problems at the user registration entrance of the urban management platform or in the form of a tourist on the platform or hotline. For example, the public can report urban management problems found through the WeChat official account or the citizen hotline. In some cases, the platform or center generates a case file based on the reported situation, and the on-site inspection urban management supervisors can be sent a verification task and decide whether to file a case based on the feedback verification result. If it is an urban management supervisor, they hold a mobile phone and conduct inspections within the jurisdiction according to the predetermined rules. Once a problem is found, information is collected and reported using the APP. After these problems are processed by government functional departments or enterprises, in some cases, the urban management supervisors can conduct on-site verification and feedback the verification results to the system.

[0048] The urban management platform belongs to the category of the municipal digital urban management system. Therefore, the present invention can be used as an embedded system / subsystem of a specially developed urban management platform and can be connected to the provincial digital urban management system for business as needed. The provincial digital urban management platform can also be used as a data source for information collection.

[0049] It should be noted that: when the present invention is developed as an urban management platform, the present invention is actually a functional unit in the urban management platform, or an embedded system / subsystem of a specially developed urban management platform. The corresponding information collection is stored in the problem reflection database. According to actual needs, if it is necessary to confirm the problem and generate a case file, a separate database can also be designed to store the confirmed problem reflection data, denoted as the confirmed problem reflection database, or the confirmed problem reflection data can also be marked and stored in the problem reflection database. For the convenience of expression, the database where the problem reflection data to be processed is located is uniformly referred to as the problem reflection database.

[0050] At the same time, the present invention also designs a basic information database, which stores information such as administrative geographical factors and corresponding regional boundaries, natural geographical factors and corresponding regional boundaries, streets and corresponding segmented regional boundaries, and infrastructure names and corresponding geographical locations and boundaries within the administrative region.

[0051] The following is an explanation in combination with specific implementation manners, and the specific implementation manners are described by taking the urban grid management platform of a certain city as an example. Specific implementation manner 1:

[0053] This implementation manner is a management system based on digital government big data mining and analysis, including:

[0054] Data acquisition and word segmentation module: According to the information of the time period to be analyzed, obtain the problem-reflecting data from the management platform database, and each piece of problem-reflecting data corresponds to a problem data ID; then perform word segmentation and part-of-speech tagging on the obtained problem-reflecting data.

[0055] This embodiment analyzes the real complaint case data collected on the urban grid management platform of a certain district in a certain city from July 2022 to July 2023. The real complaint case data collected on the urban grid management platform of a certain district in a certain city from July 2022 to July 2023 is as Figure 1 shown. For data analysis, the word segmentation and part-of-speech tagging results of the case data reflecting problems are as Figure 2 shown, and the basic data situation is shown in Table 1.

[0056] Table 1

[0057] Number of characters 341,117 characters Size 794.18 KB Number of valid items 8,241 items Total number of words 101,867 words Number of characteristic words 8,250 words Average sentence length 12.36 words Word density 8.10％

[0058] Data cleaning module: used to clean the word segmentation results; the cleaning process includes the processing step of removing stop words, and the stop words can be set according to actual needs and can be extended / deleted.

[0059] The setting of removing stop words is not only to reduce the calculation amount, but also to cooperate with the processing of the feature word extraction module.

[0060] Location area determination module: Based on the cleaned data, taking each piece of problem-reflecting data as a unit, determine the location area where each piece of problem-reflecting data occurs, and "bind" the location area to the problem-reflecting data, that is, establish a mapping relationship between the problem-reflecting data and the location area corresponding to the problem, and store it.

[0061] In the process of determining the location area where each piece of problem-reflecting data occurs, if the responder of the problem-reflecting data has filled in the location area, directly determine the location area; if the responder of the problem-reflecting data reflects the problem in a direct way of reflecting the problem and does not fill in the determined location area, the location area is determined by means of information extraction.

[0062] The method of determining the location area by means of information extraction can adopt the method of semantic recognition to determine the location area; it can also adopt the method of information matching to determine the location area.

[0063] Considering that government service platforms such as urban grid management platforms have the characteristics of relatively short information and relatively fixed formats, in some embodiments, the method of information matching is adopted to determine the location area, and the specific process is as follows:

[0064] Based on the expression order of the problem - reflecting data, each word in the word - segmentation result is matched with the address feature words. If a single word in the word - segmentation result can directly and completely match an address feature word, or if multiple consecutive words in the word - segmentation result can completely match an address feature word, then the corresponding address feature word is used as the location information word. Since each address feature word has a corresponding regional boundary or geographical location and boundary, the geographical location and spatial scope corresponding to the problem - reflecting data can be determined based on the location information word.

[0065] It should be noted that: even the problem reflected by a single problem - data ID may involve two or even three jurisdictions, that is, the determined location areas may be two or even three. For example, if a sewage pollution problem occurs at the junction of two or even three districts, then the location area where the problem - reflecting data occurs may be two or even three.

[0066] Address feature words refer to the administrative geographical factors, natural geographical factors, streets and corresponding segmented areas, and names corresponding to infrastructure in the basic information database;

[0067] Basic information database: stores basic information such as administrative geographical factors and corresponding regional boundaries, natural geographical factors and corresponding regional boundaries, streets and corresponding segmented area boundaries, and infrastructure and corresponding geographical locations and boundaries within the administrative region; among them,

[0068] Administrative geographical factors include districts, counties, towns, villages, etc.;

[0069] Natural geographical factors include rivers, lakes, mountains, etc.;

[0070] The infrastructure mentioned above includes communities (including hotels, guesthouses, restaurants, etc.), landmarks, etc.

[0071] A community not only includes the address range of the entire community, but also is subdivided into the addresses and ranges of door (building) plates.

[0072] The landmark address includes the following:

[0073] a) Memorial sites and buildings with geographical name significance, including buildings, squares, sports facilities, park green spaces, memorial sites, scenic spots and historical sites, etc.;

[0074] b) Units and courtyards with geographical name significance, including hospitals, schools, units, etc.;

[0075] c) Transportation facilities with geographical name significance, including bridges, road roundabouts, transportation stations, etc.

[0076] In this embodiment, the geographic information data (administrative geographical factors, natural geographical factors, streets and lanes, and infrastructure, etc.) in the basic information database should be designed based on the national "1:500 1:1000 1:2000 Topographic Map Element Classification and Code" and "1:500 1:1000 1:2000 Topographic Map Schema" and in combination with the requirements of the Ministry of Housing and Urban-Rural Development and relevant provincial standards. The information in the basic information database is coded and stored according to the unit grid:

[0077] Unit grid data refers to the basic jurisdictional responsibility units for urban management, divided according to specific rules. This data can include geographic unit data for human activities, natural geographic unit data for undeveloped mountains or rivers, sheet-like geographic unit data for residential or commercial communities in urban areas, and strip-like geographic unit data for special jurisdictions such as highways and railways. The unit grid is the unit for spatially locating urban management components and events and dividing jurisdictional responsibilities, serving as one of the benchmarks for locating urban management objects. Unit grids must cover every inch of a city's land, with no blind spots. The formation of unit grid data involves the generation of geographic spatial division information data and grid coding data.

[0078] The process of unit grid division is in accordance with the requirements of the national industry standard "Unit Grid Division and Coding Rules for Urban Comprehensive Supervision Information System" (CT / T213-2005) issued by the Ministry of Construction, combined with local actual conditions and the differentiated needs of various departments, to divide the geographic unit grid and draw a unit grid atlas.

[0079] In terms of the principles of unit grid division, the following principles are mainly followed in combination with the actual situation of a certain city:

[0080] 1. Principle of territorial management

[0081] The division and coding of unit grids follow the current administrative jurisdiction hierarchy, and are divided into three levels: street office, community, and unit grid; the largest boundary of the unit grid is the community boundary, and generally it should not cross the community boundary.

[0082] 2. Geographical layout principles

[0083] The area is divided according to the layout of natural units such as roads, residential land, public utility land, public green space, squares, bridges, open spaces, rivers, hills, and lakes in the city; the area is divided based on the outer boundary line of the community; the center line of the road is the dividing line; if it is a mountain, the ridge line is the dividing line.

[0084] 3. Status quo management principle

[0085] In order to take into account the integrity of buildings and management components and facilitate effective management, the boundaries of unit grids should not pass through buildings and management components.

[0086] 4. Principle of Load Balancing

[0087] When dividing the grid, the manageable unit scope of the workload of a single supervisor should be considered, taking into account the differences in the density of urban components and the frequency of incident occurrences. For grids located in the built-up area and patrolled on foot, the grid boundary and area should be determined based on the total length of the patrolled roads within the grid being about 8 kilometers. For grids in non-built-up areas or patrolled by vehicle, the area range of the unit grid can be expanded to about 2 million square meters.

[0088] 5. Principle of Convenient Access

[0089] For the implementation of convenient management, the management path should be made as convenient as possible, and the easy accessibility of walking or using transportation means should be considered. Principle of Unified Opening The design and division of unit grids should not only be applicable to urban management work, but also consider providing a compatible operation platform for the expansion and application in multiple fields such as public security management, medical first aid, special surveys, and urban emergency response. Each unit grid is both a geographical unit for citizens' daily production and life and a geographical unit for the decision-making and analysis applications of large-scale urban management. Principle of Seamless Splicing The division of unit grids should neither have gaps nor overlaps. Principle of Relative Stability The division of unit grids should be kept relatively stable.

[0090] The unit grids corresponding to the information in the basic information database are designed according to the requirements of geographical coding surveys. Each code of the unit grid coding is unique in terms of time and space definition. When the unit grid changes, its original code should not be occupied. Since a certain city is a developing urban area, the boundaries and quantities of unit grids may be appropriately adjusted with the continuous expansion of the built-up area. Therefore, when coding the unit grids, situations such as possible grid splitting and grid merging in the future should be considered, and code numbers should be reserved for the possibly newly added grids.

[0091] The sequence codes of unit grids are all coded in accordance with the principle of from left to right and from top to bottom in space. The unit grid atlas is based on the topographic map and carried out according to the division principle. After division, grid codes are attached to each unit grid and printed into a map. Each unit grid has an attribute description in the atlas.

[0092] The matching method in the location area determination module can not only ensure the accurate matching of the area reflected by administrative geographical factors, but also effectively avoid the address information matching errors caused by word segmentation. Especially for communities in infrastructure, it can effectively avoid incorrect matching. For example, when segmenting the Riverside Community, it will be segmented into "Riverside" + "Community". If it is located and matched by "Riverside" or "Community", errors may occur. In addition, this method can more effectively avoid the incorrect matching of address feature words composed of verbs + nouns. For example, for communities such as Guanhu Community, Guanhu International, and Guajiang International, when segmented, they will be split into "Guan" + other nouns. If only the noun part is used for matching during the matching process, errors may occur.

[0093] Keyword extraction module: Based on the cleaned data, cluster according to nouns to obtain feature words; then, based on the clustered feature words, extract keywords.

[0094] Taking the complaint case data of a certain district of the "Whistleblowing Report" urban grid management platform in a certain city from July 2022 to July 2023 as an example for illustration, this data set is called the case data set. After analysis, it is found that there are more nouns in the cleaned data, and the combination of "organization name + nominal verb + noun" accounts for 70.47% of the total. Therefore, in this invention, nouns are used as keywords for analysis, and there will be less ambiguity; while custom words are not analyzed as a key point because of their strong personalization and chaotic standards. The basic situation of the real complaint case data collected on the urban grid management platform of a certain district of the "Whistleblowing Report" in a certain city from July 2022 to July 2023 is as Figure 3 shown, and finally 8,250 feature words are sorted out in the feature word list.

[0095] In some embodiments, the existing TF-IDF (term frequency–inverse document frequency) is commonly used to extract keywords. However, after manual verification, it is found that the effect of this keyword extraction method has a relatively large gap with the content that the problem data actually reflects. After research and analysis, it is found that the reason why this method is not ideal is mainly because the way the problem data actually reflects is quite different from the previous IDF statistical method. Because IDF is the inverse document frequency index, Where N represents the total number of documents, and DF(x) represents the number of documents containing the word x. Research has found that if the number of documents containing a keyword is smaller, or if the number of documents containing a keyword is relatively fixed while the total number of documents is larger, then the corresponding IDF value will be larger. However, for government affairs processing platforms / systems, especially those that provide feedback on municipal issues, the more similar issues are, the greater the impact of the issue on people's lives and the more serious the issue. Consequently, the number of issues (which can be understood as the corresponding number of documents) will increase. Using the IDF method may cause common issues to be overwhelmed to a certain extent, which is inappropriate. The fundamental reason for this is that IDF itself is a measure of the general importance of a word. Its specific formula is designed to determine the importance of certain words for a specific document. For documents, the higher the word frequency, the better. Documents often contain multiple "的" (de) and "了" (le), and almost every document contains these words. To remove the influence of these unimportant, high-frequency words, some scholars use TF-IDF to extract keywords. Although the present invention can eliminate the need to remove stop words, using TF-IDF to extract keywords is not recommended. However, this will still cause the problem of drowning out common documents (which are actually texts reflecting problems) to a certain extent.

[0096] After analyzing the real complaint case data collected on the urban grid management platform of a certain district in a certain city from July 2022 to July 2023, the present invention found that the text reflecting the problem is generally not too long, among which 3-5 sentences account for the largest proportion of the problem data, and the number of words contained in each sentence is not too large. The average sentence length is 12.36 words, which is about 10 after removing stop words. On this basis, the present invention uses keyword selection indicators to select keywords, and the keyword selection indicators are as follows:

[0097]

[0098] Among them, n x represents the number of words x in the text corresponding to the problem data, N represents the total number of texts corresponding to the problem data, T x Represents the number of texts corresponding to the problem data reflected by word x. λ is a proportional coefficient used to adjust the keyword selection index. It can be determined based on the actual problem reflection dataset. For example, if some cities have an overpopulated population or are too large, the data reflecting the problem will increase accordingly, and the problems reflected will cover a wider range. As a result, the relative proportion of some problems will become smaller. When the keyword selection index is relatively small, it is not conducive to the overall data expression. Therefore, it needs to be adjusted through the proportional coefficient. For the case dataset described in this embodiment, the proportional coefficient here is 2.

[0099] Based on the foregoing complaint case data, through the research and analysis of the corresponding text of each problem - reflecting data and the word count in the text, it is found that the number of words in the text corresponding to each problem - reflecting data is generally 3 - 10, and basically does not exceed 15. Therefore, the present invention uses to limit the expression amount of word x in the text corresponding to the problem - reflecting data, which is equivalent to the average expression amount of the word frequency in the problem - reflecting data and is used to measure the importance of the word in the problem - reflecting data. In the formula, using n x in the form of - 5 can control the expression amount of words with 0 - 15 words in the corresponding text within 0 - 1, thus equivalent to a probability - like processing. At the same time, it makes the words with 3 - 7 words limited to a more "tendency" - like linear expression, makes the words corresponding to the word counts 1 - 10 near both ends "tendency" - compressed expression, so as to play a certain restrictive role, and for the words corresponding to the word counts 11 - 15 near both ends, "to a greater extent" compressed expression. It is used to measure the recurrence times of the problem - reflecting data corresponding to the words reflecting a certain problem in all the data, that is, the importance degree and influence range of some problems. Therefore, the keyword selection index I can effectively "reflect" the importance degree of the feature words.

[0100] Based on each noun after noun clustering, sort them according to the selection index corresponding to the word, and extract keywords in the sorting order.

[0101] Based on the previous noun clustering results, after extracting keywords, the keywords with the most occurrences are market, riverside, property management, disturbing residents, rights protection, noise, etc. The keyword cloud of the case data set is as Figure 4 shown, and an obvious clustering effect can be seen. The key complaint source communities, main problems, etc. are all clear at a glance.

[0102] Directed association rule graph generation module: Use the keywords as nodes and obtain the directed association rule graph of the keywords by using the directed hypergraph spanning tree algorithm.

[0103] For the case data set described in this embodiment, regard the keywords as nodes, connect the keywords that appear in the same case, and mark the direction of the edges according to the attributes of the left - adjacent word and the right - adjacent word, that is, complete the transformation from the keyword data to the directed hypergraph, and obtain the directed association rule graph of the keywords. The association rule graph of the complaint cases in a certain district of the management platform is as Figure 5 shown. In a certain district, the most complained communities are Cultural Community, Bridge Community and Riverside Community. The most complained department is the Market Supervision Bureau. The most concentrated problems in the matter category are rights protection problems, disturbing residents problems, noise problems and property management problems. Figure 6 It is the node data of the association rule graph of a keyword (Riverside Community) for the management platform, and the directed relationship graph is as Figure 7As shown, the strongest association rule is the noise, nuisance, and property management issues complained about in Xipan Community. Taking Xipan as an example, we continue with data mining. The basic situation of the data is shown in Table 2.

[0104] Table 2

[0105]

[0106] Figure 8 It is a directed association rule graph of the complaint case keywords of Xipan Community in a certain area of the management platform. As can be seen from the graph, the strongest association rule is: the noise, nuisance, and property management issues complained about in Xipan Community. Subsequently, by asking the business personnel, we found that the real situation is as follows: 1. The noise problem is because Xipan Community is adjacent to Wanda Plaza, and there are three dance teams practicing square dance with high-volume speakers every night regardless of weather. 2. Xipan Community has three areas, namely A, B, and C. Among them, B and C are old areas. Area A has completed new shantytown renovation, and the occupancy rate is not high enough. The coverage of property fee collection is not complete, and the property company is located in Area B for office, resulting in a slow response to some demands in Area A. The complainants are mainly the owners of Area A, and the problems of the property company's inaction they reported are mainly concentrated in aspects such as community environmental hygiene, decoration construction, random placement of sundries, and poor management of the garden. 3. The nuisance problems mentioned in Xipan Community are concentrated in urban appearance and municipal administration. This is because Xipan Community is adjacent to Wanda and the pedestrian street, and there are frequent cases of merchants occupying the road for business. In addition, the large fumes from roadside barbecues are also the problems they mainly complained to the urban management about; the fact of road maintenance problems is also supported, and there are also problems such as the lack of manhole covers on the main road near the community, resulting in too much noise when vehicles pass over the manhole covers at midnight. In short, through communication with business experts, it shows that our data mining results are basically consistent with the real situation.

[0107] Based on the platform data of the "Internet + Government Affairs Service" type, the system of the present invention can directly use the platform database, or establish a data sharing library, build a basic information library according to the real business scenario, generate high-dimensional big data based on the platform data / shared database data and the self-built thesaurus, and conduct data mining, which can effectively discover the problems reflected by the masses in urban management, provide data support for solving urban management problems, can effectively improve the efficiency of urban emergency response or problem handling, and thus improve the urban management level.

[0108] The management system based on digital government big data mining and analysis described in this embodiment, as an embedded system / subsystem or functional unit of the urban grid management platform, can effectively provide a data analysis foundation for the entire management platform. It is one of the governance models and policy tools with the widest influence in urban management and grass-roots governance, enabling the overall platform to solve grass-roots problems in urban governance, respond to the changing expectations of the public, and relieve the increasing urban governance pressure. It plays an important role and has a significant impact in work such as urban facility management and maintenance, community safety, public services, community administration, employment assistance, grass-roots public security, and information collection and maintenance. According to statistics, as of the end of 2022, grid management across the country has covered more than 270 prefecture-level cities in 31 provinces, with a coverage rate of over 90%. There are more than 2.6 million grids divided in villages (communities) across the country, and more than 4.3 million grid workers. Therefore, the system of the present invention can be used as a separate embedded system or as a subsystem of the urban management platform system to jointly undertake urban management operations with the platform, indicating the broad application prospects of the present invention.

[0109] Keyword query and display module: It is used to receive the query word information input by the manager user, query the keyword information that best matches the query word information, and highlight the directed relationship of the corresponding confirmed keyword nodes in the directed association rule graph of keywords; or, receive the pointer action information of the manager user in the directed association rule graph of keywords, confirm the keyword node pointed to by the pointer according to the pointer action information, and highlight the directed relationship of the corresponding confirmed keyword nodes in the directed association rule graph of keywords.

[0110] During the process of confirming the keyword node pointed to by the pointer according to the pointer information action, collect the dragging information of the pointer (the dragging action and direction of the pointer); in the directed association rule graph of keywords, according to the dragging information of the pointer (the dragging action and direction of the pointer), scroll the directed association rule graph of keywords correspondingly according to the connection relationship of the nodes, and accept the finally confirmed keyword node pointed to by the pointer. Specific embodiment two:

[0112] This embodiment is a management device based on digital government big data mining and analysis. The device includes a processor and a memory. It should be understood that any device including a processor and a memory described in the present invention may also include other units and modules for display, interaction, processing, control, etc. through signals or instructions, as well as other functions.

[0113] At least one instruction is stored in the memory, and the at least one instruction is loaded and run by the processor to execute the management system based on digital government big data mining and analysis.

[0114] It should be understood that the instructions include computer program products, software, or computerized methods corresponding to any method described in the present invention; the instructions can be used to program a computer system or other electronic devices. The memory may include a readable medium on which the instructions are stored, which may include, but is not limited to, magnetic storage media, optical storage media; magneto-optical storage media include read-only memory ROM, random access memory RAM, erasable programmable memory (e.g., EPROM and EEPROM), and flash memory layers, or other types of media suitable for storing electronic instructions. Specific Embodiment Three:

[0116] This embodiment is a decision support system based on digital government big data mining and analysis management, which is the decision support system corresponding to the management system based on digital government big data mining and analysis described in Specific Embodiment One, that is, decision support suggestions are generated based on the information provided by the directed association rule graph generation module of the management system based on digital government big data mining and analysis.

[0117] A decision support system based on digital government big data mining and analysis management described in this embodiment includes:

[0118] Institution and Personnel Database: Stores the institutional organization relationships of all departments in each jurisdiction area and the corresponding job responsibilities, specifically including: department table, department relationship table, position table, role table, employee table, employee-role relationship table, department-employee table, employee-position relationship table, role relationship table. It is used to describe the composition of departments and personnel, and can describe the relationships and changes among departments, personnel, positions, and roles, such as the transfer of positions and personnel, and determines the permissions of employees through roles, serving for information authorization. The institution and personnel data can be independent of process management and at the same time provide a basis for process management.

[0119] Requirement Information Receiving Module: Used to obtain the keywords and directed association rule graph obtained by the directed association rule graph generation module of the management system based on digital government big data mining and analysis; at the same time, provides keyword options for users and provides corresponding directed association rule graphs according to the keyword options.

[0120] First Support Suggestion Module: Based on the keywords confirmed by the user and the corresponding directed association rule graph, first identify the keywords corresponding to the keyword nodes, and determine whether the words belong to the words related to "problem" of the reflected problem or the words related to "location" of the reflected problem; then, based on the words related to "location" of the reflected problem and the words related to "problem", perform jurisdiction area, institutional organization, and job responsibility matching in the institution and personnel database;

[0121] For example, for the words related to "location" that reflect problems, according to the problem-reflecting data corresponding to the keywords reflected by the nodes belonging to "problems", determine the ID corresponding to the problem-reflecting data, and correspondingly obtain the location area where the problem-reflecting data determined in the location area determination module occurs. According to the occurring location area, give a recommended match of the jurisdiction corresponding to the solution that can be implemented. It should be noted that even for the problems reflected by the corresponding problem data ID, the location area where the problem-reflecting data determined by the location area determination module may be two or even three. Therefore, the recommended results of the jurisdiction may also be two or even three, and different jurisdictions and the departments of different jurisdictions are matched simultaneously. At the same time, for the words related to "problems" that reflect the problems, such as "noise", "environment", etc., match the corresponding departments that solve the corresponding problems, and the corresponding persons in charge who give suggestions according to the job responsibilities.

[0122] When the recommended result of the jurisdiction is more than one, in the task allocation system used in the management platform to solve problems, the corresponding solution tasks (after approval) are allocated to the corresponding districts, and the corresponding districts are handed over to the corresponding solution departments for solution. The task allocation system can also be other embedded systems / subsystems or functional units on the management platform other than the system described in the present invention. Since the present invention does not focus on task allocation, the present invention will not be described in detail.

[0123] The second support and suggestion module: Based on the keywords confirmed by the user and the corresponding directed association rule graph, extract the theme and scenario information, obtain all the data of the corresponding complaint cases according to the theme and scenario information determined by the scenario and theme determination module, analyze the data and give the information about the future occurrence of similar problems. The information about the future occurrence includes whether it will occur in the future and / or the expected occurrence location, so as to serve as potential decision support suggestions.

[0124] The process of analyzing the data and giving the information about the future occurrence of similar problems can adopt methods such as Bayesian networks and neural networks. In this embodiment, taking the Bayesian network as an example to give potential decision support suggestions can not only retrospectively deal with the situation once the problem occurs, but also provide suggestions for future resource scheduling to reduce the impact of the corresponding problem occurrence. Even relevant resources can be arranged to conduct pre-inspections in advance to reduce the possibility of the corresponding problem occurrence.

[0125] To illustrate the effectiveness of a management system based on digital government big data mining and analysis and a decision support system based on digital government big data mining and analysis management, especially the accuracy, effectiveness, and general applicability of keyword extraction in a management system based on digital government big data mining and analysis (which is the basis for the effectiveness of decision support suggestions), the present invention is illustrated with the data of Implementation 2. In Implementation 2, the number of 12345 complaint cases in a certain city from January to August 2023 was approximately 910,000. After effective data cleaning, 115,734 total words were retained, and 14,030 common feature words. Using nouns as keywords, a word cloud is as follows Figure 9 shown, and the number of keywords with the most frequent occurrences in the word cloud diagram is as follows Figure 10 shown. The directed association rule graph obtained based on a management system based on digital government big data mining and analysis is as follows Figure 11 shown.

[0126] In this embodiment, decision support is given for pipeline lines. The Bayesian network is used to analyze and predict the cases of complaints related to pipeline lines in a certain area, and decision support suggestions are given. The structure learning of the Bayesian network is to determine the optimal network structure according to data samples or prior knowledge, or a combination of both. The research difficulty lies in finding a model with the highest coincidence degree with the data. If there are redundant variable relationship connections (directed arcs) in the model, it will increase the probability parameters that need to be learned, and at the same time, it will also reveal the causal relationships of variables in the wrong domain. This embodiment selects a hybrid modeling method: it can not only use experience or expert knowledge to provide subjective guidance, but also obtain the hidden causal relationships between variables through data-driven, so as to establish a relatively perfect Bayesian network model. The hybrid modeling method is to integrate information such as variable relationships, rules, and orders in domain knowledge into the data learning algorithm. Since this method can determine some relationships of parameters and a simple model structure before data learning, and remove meaningless causal relationships, this can greatly reduce the search space of the optimal model, thereby improving the learning efficiency and the fitting accuracy of the optimal model and the data.

[0127] Then, the constructed Bayesian network is used for prediction, and the past data is used for prediction testing. The results of predicting similar problems using the Bayesian network are as follows Figure 12 , Figure 13 and Figure 14 shown. Figure 12 This is the Bayesian network prediction diagram for pipeline lines. For the prediction of pipeline lines, the types of a certain area are mainly the four shown in the figure (the actual problem types of pipeline lines are at least 11). Figure 13The associated diagram corresponding to the reasons for gas complaints. Gas complaints mainly revolve around reasons such as routine construction and renovation, and are concentrated in Xingda Road, Xinfa Community, Bile Street, Xuanxi Community, Yonghe Street, Haxi Street, and near Xuanxi Community. After manually screening and comparing with the data of complaint cases, it is found that the prediction results are very consistent with the actual situation. Figure 14 The associated diagram of gas incidents in a certain area within a time range. After manually screening and comparing with the data of complaint cases, it is found that the prediction results are consistent with the actual situation. Based on the corresponding situation of complaint analysis, similar complaints may still occur, and manual judgment also indicates a high probability event. For example, it is found that the gas pipeline in Building B06 of Xinfa Community in a certain district is rotten. However, due to the proximity of the gas pipeline in the stairwell to the telecommunications pipeline and the heating pipeline, construction cannot be carried out, and the gas renovation has stagnated. In August 2023, during the construction process, there were obstructions from individual residents and individual on-site construction workers, and the construction was not successful. In December 2023, there were still complaints about the unfinished construction. This problem has not been solved, and it is predicted that there will still be related complaints in 2024. Specific implementation method four:

[0129] This implementation method is a decision support device based on digital government big data mining, analysis, and management. The device includes a processor and a memory. It should be understood that any device including a processor and a memory described in the present invention, the device may also include other units and modules for display, interaction, processing, control, etc. through signals or instructions and other functions;

[0130] At least one instruction is stored in the memory, and the at least one instruction is loaded and run by the processor to execute a decision support system based on digital government big data mining, analysis, and management.

[0131] It should be understood that the instructions include any computer program products, software, or computerized methods corresponding to the methods described in the present invention; the instructions can be used to program a computer system or other electronic devices. The memory may include a readable medium on which instructions are stored, which may include but are not limited to magnetic storage media, optical storage media; magneto-optical storage media include read-only memory ROM, random access memory RAM, erasable programmable memory (e.g., EPROM and EEPROM), and flash memory layers, or other types of media suitable for storing electronic instructions.

[0132] The above examples of the present invention are only used to illustrate in detail the calculation model and calculation process of the present invention, rather than to limit the implementation methods of the present invention. For those of ordinary skill in the art, other different forms of changes or modifications can be made based on the above description. It is impossible to list all the implementation methods here. Any obvious changes or modifications derived from the technical solutions of the present invention still fall within the protection scope of the present invention.

Claims

1. A management system based on digital government big data mining and analysis, characterized in that, including: Data acquisition and word segmentation module: According to the information of the time period to be analyzed, obtain the problem-reflecting data from the management platform database. Each piece of problem-reflecting data corresponds to a problem data ID; then perform word segmentation and part-of-speech tagging on the obtained problem-reflecting data. Data cleaning module: used to clean the data of the word segmentation result. Location area determination module: Based on the cleaned data, taking each piece of problem-reflecting data as a unit, determine the location area where each piece of problem-reflecting data occurs, and "bind" the location area to the problem-reflecting data, that is, establish a mapping relationship between the problem-reflecting data and the location area corresponding to the problem, and store it. Basic information database: Store the administrative geographical factors and corresponding regional boundaries within the administrative region, natural geographical factors and corresponding regional boundaries, streets and corresponding segmented regional boundaries, and infrastructure and corresponding geographical locations. Keyword extraction module: Based on the cleaned data, cluster according to nouns to obtain feature words; then based on the clustered feature words, extract keywords; in the process of extracting keywords, use keyword selection indicators to select keywords. The keyword selection indicators are as follows: where n x represents the number of the word x in the text corresponding to the problem - reflecting data where the word x is located, N represents the total number of texts corresponding to the problem - reflecting data, T x represents the number of texts corresponding to the problem - reflecting data where the word x is located; λ is a proportionality coefficient. Directed association rule graph generation module: Use the keywords as nodes and use the directed hypergraph spanning tree algorithm to obtain the directed association rule graph of the keywords. Keyword query and display module: used to receive the query word information input by the manager user, query the keyword information that best matches it according to the query word information, and highlight the directed relationship of the corresponding confirmed keyword nodes in the directed association rule graph of the keywords; or, receive the pointer action information of the manager user in the directed association rule graph of the keywords, confirm the keyword node pointed to by the pointer according to the pointer action information, and highlight the directed relationship of the corresponding confirmed keyword nodes in the directed association rule graph of the keywords.

2. The management system based on digital government big data mining and analysis according to claim 1, characterized in that, In the process of determining the location area where each piece of problem-reflecting data occurs, if the responder of the problem-reflecting data has filled in the location area, directly determine the location area. If the responder of the problem-reflecting data reflects the problem in a direct way of reflecting the problem and does not fill in the determined location area, the location area is determined by means of information extraction.

3. A management system based on digital government big data mining and analysis according to claim 2, characterized in that, The method of determining the location area by means of information extraction uses the method of information matching to determine the location area; the specific process of using the method of information matching to determine the location area is as follows: Based on the expression order of the problem-reflecting data, match each word of the word segmentation result with the address feature word. If a word of the word segmentation result can directly match the address feature word completely, or if multiple consecutive words of the word segmentation result can match the address feature word completely, then use the corresponding address feature word as the location information word, and determine the geographical location and spatial range corresponding to the problem-reflecting problem based on the location information word.

4. A management system based on digital government big data mining and analysis according to claim 3, characterized in that, The cleaning process of the data cleaning module includes the processing step of removing stop words.

5. A management device based on digital government big data mining and analysis, characterized in that, The device includes a processor and a memory. At least one instruction is stored in the memory, and the at least one instruction is loaded and run by the processor for a management system based on digital government big data mining and analysis as described in any one of claims 1 to 4.

6. A decision support system based on digital government big data mining, analysis and management, characterized in that, including: Institution and Personnel Database: Stores the institutional organizational relationships of all departments in each jurisdiction area and the corresponding job responsibilities. Requirement Information Receiving Module: Used to obtain the keywords and the directed association rule graph obtained by the directed association rule graph generation module of the management system based on digital government big data mining and analysis described in any one of claims 1 to 4; at the same time, provide keyword options for the user and provide the corresponding directed association rule graph according to the keyword options. First Support and Recommendation Module: Based on the keywords and the corresponding directed association rule graph confirmed by the user, first identify the keywords corresponding to the keyword nodes to determine whether the words belong to the words related to "problems" or the words related to "locations" of the problems reflected; then, based on the words related to "locations" and the words related to "problems" of the reflected problems, perform jurisdiction area, institutional organization, and job responsibility matching in the Institution and Personnel Database.

7. A decision support system based on digital government big data mining analysis and management according to claim 6, characterized in that, During the process of performing jurisdiction area, institutional organization, and job responsibility matching in the Institution and Personnel Database based on the words related to "locations" and the words related to "problems" of the reflected problems, for the words related to "locations" of the reflected problems, determine the ID corresponding to the reflected problem data according to the reflected problem data corresponding to the keywords reflected by the nodes belonging to "problems", and correspondingly obtain the location area where the reflected problem data determined in the Location Area Determination Module occurs. According to the location area where it occurs, give the jurisdiction matching recommendation corresponding to the solution that can be implemented.

8. A decision support system based on digital government big data mining, analysis and management according to claim 6 or 7, characterized in that, The system further includes a Second Support and Recommendation Module: Based on the keywords and the corresponding directed association rule graph confirmed by the user, extract the theme and scenario information, obtain all the data of the corresponding complaint cases according to the theme and scenario information determined by the Scenario and Theme Determination Module, analyze the data and give the information on the future occurrence of similar problems. The information on the future occurrence includes whether it will occur in the future and / or the expected occurrence location, so as to serve as potential decision support recommendations.

9. A decision support system based on digital government big data mining analysis and management according to claim 8, characterized in that, The process of the Second Support and Recommendation Module analyzing the data and giving the information on the future occurrence of similar problems is implemented using a Bayesian network.

10. A decision support device based on digital government big data mining, analysis and management, characterized in that, The device includes a processor and a memory. At least one instruction is stored in the memory, and the at least one instruction is loaded and run by the processor to execute a decision support system for digital government big data mining and analysis management described in claim 8.

Images