Grassroots society intelligent governance method and system based on big data analysis

Abstract

The invention discloses a grassroots society intelligent governance method and system based on big data analysis, and the method comprises the steps: building a community governance model, collecting the complaint information of a user, and storing the complaint information in the community governance model; historical behavior records of the user in the government affair complaint system in a preset time period are called, and the credibility of the user is verified; the method comprises the following steps: analyzing complaint information with user credibility meeting requirements, performing keyword extraction and classification on complaint text information, and determining core keywords by identifying semantics, domain attributes and context association; for the user who complaints the picture and the video information, the complaint picture and the video information need to conform to a preset shooting specification and watermarks are added; and distributing the complaint work order to the corresponding government affair department according to the government affair responsibility matching degree and the complaint type. According to the invention, through dynamic local updating of the cell governance model, only a single minimum cell governance unit is updated, so that full-amount reconstruction of the model is avoided, resource occupation is minimized, and the grassroots governance response speed and the resource utilization rate are integrally improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of basic social governance, and particularly relates to a basic social intelligent governance method and system based on big data analysis. BACKGROUND

[0002] With the acceleration of urbanization, the basic community governance is facing the challenges of data fragmentation, positioning fuzzification and inefficient updating.

[0003] Community governance involves multi-source data such as spatial information, cultural data and dynamic events. In the traditional governance mode, data is scattered and stored in property, community and government systems, lacking unified correlation mechanism and data integration difficulty, resulting in information island. The existing complaint positioning relies on text description, and the spatial positioning accuracy is insufficient, which cannot be accurately positioned to a specific location, resulting in high deviation rate of work order allocation and low disposal efficiency. The community governance model is mostly static structure, and the whole model needs to be reconstructed when new complaints or facility changes occur, which is time-consuming and resource-consuming, and the model updating cost is high, which is difficult to adapt to the processing demand of high-frequency dynamic events. The authenticity of complaint information lacks hierarchical verification mechanism, and there is a phenomenon of false complaints or malicious harassment, which interferes with the allocation of government resources and reduces the governance credibility.

[0004] In view of the above problems, the application provides a basic social intelligent governance method and system based on big data analysis, which realizes accurate positioning, efficient updating and intelligent application of governance data through dynamic local updating of the model. SUMMARY

[0005] In view of the problems proposed in the above background technology, the purpose of the application is to provide a basic social intelligent governance method and system based on big data analysis to solve the problems in the background technology.

[0006] To achieve the above technical purpose, the technical scheme adopted by the application is as follows: A basic social intelligent governance method based on big data analysis, comprising the following steps: A community governance model is constructed, user complaint information is collected and stored in the community governance model, and the complaint information includes text, picture and video information; The historical behavior records of the user in the government complaint system within a preset period are called, and the credibility of the user is verified. The high-credibility user can be directly verified, the medium-credibility user needs to complete the biological feature comparison and specified action execution verification, and the low-credibility user triggers the enhanced verification; The complaint information of the user with the required credibility is analyzed, the key words of the complaint text information are extracted and classified, the core key words are determined by identifying the semantics, domain attributes and context association, and the complaint picture and video information of the user need to meet the preset shooting specification and add watermark; According to the degree of matching of the government affairs responsibilities and the type of the complaint, the complaint work order is distributed to the corresponding government department for management.

[0007] In a preferred embodiment, the step of constructing the community management model, collecting user complaint information and storing the complaint information into the community management model, the complaint information being picture and video information processing, comprises: Collecting community multi-dimensional data as the original input of the community management model construction, and storing the multi-dimensional data into the community management model after standardization processing; The community management model combines the physical space grid to form a three-level block of the community management model, the three-level block being a first-level block, a second-level block and a third-level block, the first-level block being a complete community boundary and community basic information, the second-level block being a functional area, the sub-area being divided according to the use, and the third-level block being a grid divided in the second-level block, each grid corresponding to a specific location of the community; The first-level block, the second-level block and the third-level block are respectively encoded according to the hierarchical progression, being denoted as first-level block encoding, second-level block encoding and third-level block encoding; The complaint information is updated in combination with the first-level block, the second-level block and the third-level block, a mapping relationship between the complaint location coordinates and the third-level block is constructed based on the spatial index, and is stored into the community management model.

[0008] In a preferred embodiment, the step of updating the complaint information in combination with the first-level block, the second-level block and the third-level block, constructing a mapping relationship between the complaint location coordinates and the third-level block based on the spatial index, and storing into the community management model, comprises: A root node is established based on the community management model, each root node is divided into four quadrants with the center point of the community as the origin, and the coordinates of the center points of the four quadrants are taken as first-level nodes, the first-level nodes being the center points of the first-level blocks; Each first-level node is divided into four quadrants again with the center point of the first-level node as the origin, and the coordinates of the center points of the four quadrants are taken as second-level nodes, the second-level nodes being the center points of the second-level blocks; Each second-level node is divided into four quadrants again with the center point of the second-level node as the origin, and the coordinates of the center points of the four quadrants are taken as third-level nodes, the third-level nodes being the smallest management units of the community, and the third-level nodes being the center points of the third-level blocks; An index table is constructed, and a mapping relationship between the third-level block encoding and the center point coordinates of the third-level nodes is stored based on the community management model; According to the index table, the complaint location coordinates, the third-level block encoding and the center point coordinates of the third-level nodes are matched, and the complaint location coordinates, the third-level block encoding and the center point coordinates of the third-level nodes are recorded as complaint data, and the dynamic local update of the community management model is completed.

[0009] In a preferred embodiment, the step of matching the complaint location coordinates, the tertiary block code, and the tertiary node center point coordinates according to the index table, and marking the complaint location coordinates, the tertiary block code, and the tertiary node center point coordinates as complaint data, to complete the dynamic local update of the community management model, comprises: Querying the complaint data and the index table to obtain the complaint location coordinates, the tertiary node center point coordinates, and the associated tertiary block code in the community management model to be updated, denoted as a target unit; Building a static anchoring base for the target unit, and storing the complaint data in the static anchoring base, wherein the static anchoring base interacts with the target unit and the community management model information respectively; the static anchoring base comprises a read-write control unit, a hash verification unit, and a metadata storage unit, the read-write control unit is integrated with a distributed lock interface for subsequent locking / unlocking of the storage space of the static anchoring base, the hash verification unit generates a unique reference hash code based on the complaint data of the target unit, and the metadata storage unit is used to store the complaint data of the target unit; Establishing a bidirectional pointer anchoring between the target unit and the static anchoring base, the target unit generates a pointer pointing to the static anchoring base through memory address mapping, which is used to read the complaint data in the static anchoring base; the static anchoring base reversely generates a pointer pointing to the community management model through the tertiary block code, which is used to write the complaint data, and the connection is ensured to be effective by comparing the hash verification codes of the two pointers; Configuring a preset number of virtual carriers to grab the complaint data of the corresponding target unit; Applying a distributed read-write lock through the read-write control unit of the static anchoring base, applying a write lock to the complaint data of the target unit, and applying a read lock to the index path associated with the non-target unit, and during the locking period, generating real-time verification codes by the hash verification unit to shield the write operation request of the non-target unit; The static anchoring base is unbound from the data stream of the community management model, and an incremental data snapshot of the target unit is generated, the complaint data is written into the incremental data snapshot by the preset number of virtual carriers, and then the incremental data snapshot is written into the static anchoring base; Releasing the distributed read-write lock applied by the read-write control unit, synchronizing the complaint data to the community management model through the bidirectional pointer of the static anchoring base, updating the index table mapping relationship, releasing the read lock of the non-target unit, and restoring the global read-write permission of the community management model.

[0010] In a preferred embodiment, the step of calling the historical behavior record of the user in the government complaint system in the past preset period and verifying the credibility of the user comprises: Querying the user complaint history, which includes the number of complaints, the type of complaints, and the content of complaints; Querying whether the user has a false complaint or malicious harassment record of bad faith; The credibility score is generated based on the credibility score rule, and the credibility of the user is generated.

[0011] In a preferred embodiment, the medium credibility user needs to complete the steps of biometric feature comparison and designated action execution verification, including: Collecting the full-face image of the medium credibility user through the front camera of the user's mobile device, extracting the facial key area features, and generating a unique feature template; Comparing the full-face image of the medium credibility user with the face data in the government identity information library in 1:1; After the face data comparison is successful, designated action execution verification is performed, and the system randomly selects at least one action from the preset action library and guides the medium credibility user to perform through text.

[0012] In a preferred embodiment, the low credibility user triggers the step of enhanced verification, including: Collecting the full-face image of the low credibility user through the front camera of the user's mobile device, extracting the facial key area features, and generating a unique feature template; Comparing the full-face image of the low credibility user with the face data in the government identity information library in 1:1; After the face data comparison is successful, designated action execution verification is performed, and the system randomly selects at least three actions from the preset action library and guides the medium credibility user to perform through text. Through artificial review of the complaint content of the low credibility user, if there are more than 2 false complaint records, the complaint function is limited.

[0013] In a preferred embodiment, the step of extracting and classifying keywords from the complaint text information, and determining the core keywords by recognizing semantics, domain attributes, and context association, includes: Performing word segmentation on the complaint text information based on a dictionary, and annotating the part of speech; Classifying the complaint text information that has been segmented; Determining the core keywords based on the frequency of the segmented and classified complaint text information.

[0014] In a preferred embodiment, the step of collecting picture information from the complaint information includes: Guiding the user to take picture information according to three-level standards of panorama, medium shot, and close-up, the panorama picture information includes complete scene environment, the medium shot picture information focuses on the problem subject, and the close-up picture information presents detailed features; Respectively determining the shooting distance and horizontal angle of view of the panorama, medium shot, and close-up picture information; Adding time and location watermarks to different positions of the panorama, medium shot, and close-up picture information.

[0015] A grassroots social wisdom governance system based on big data analysis, comprising: A model construction module connected with the data acquisition module, used for constructing a community governance model; A data acquisition module connected with the credibility evaluation module, used for acquiring complaint information; A credibility evaluation module connected with the work order scheduling module, used for evaluating user credibility; A work order scheduling module, used for distributing complaint work orders to corresponding government departments, the complaint work orders including complaint text information, complaint picture and video information, complaint location coordinate corresponding three-level block code and user credibility information.

[0016] The beneficial effects of the present application are: The present application constructs a community governance model containing a first-level block, a second-level block and a third-level block, acquires text, picture and video complaint information and matches them to the smallest community governance unit of the third-level node according to spatial index; generates a credibility score through complaint history and bad faith record, and performs biological feature, dynamic action verification and manual review on users with medium and low credibility; classifies and extracts core keywords from text information, and takes pictures according to three-level standards of panorama, medium shot and close-up and adds time and space watermark, and finally distributes work orders according to the matching degree of government responsibilities.

[0017] The advantage is that the community governance model is dynamically updated locally, only a single smallest community governance unit is updated, the unique third-level node and the corresponding code are accurately positioned through the complaint location coordinate or the coding query index table, the static anchor base is constructed for the target unit, the data tamper-proofing is realized by using bidirectional pointers, hash check and distributed read-write lock, the data transmission rate is reduced by setting incremental data snapshots, the data grabbing efficiency is improved by multiple virtual carriers to grab data, thereby realizing local dynamic update of complaint data, avoiding full reconstruction of the model, minimizing resource occupation, reducing the allocation deviation rate of complaint work orders, and improving the grassroots governance response speed and resource utilization rate. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can also be obtained by those skilled in the art based on these drawings.

[0019] Figure 1 The method flowchart of the present application.

[0020] Figure 2 The system block diagram of the present application. DETAILED DESCRIPTION

[0021] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0022] Embodiment 1, please refer to Figure 1 As shown in the figure, the method for grassroots social wisdom governance based on big data analysis comprises the following steps: S1, constructing a community governance model, collecting user complaint information and storing it in the community governance model, the complaint information including text, picture and video information; S2, calling the historical behavior records of the user in the government affair complaint system within the preset period of time, and verifying the credibility of the user, the high-credibility user can pass the verification directly, the medium-credibility user needs to complete the biological feature comparison and designated action execution verification, and the low-credibility user triggers the enhanced verification; S3, analyzing the complaint information of the user with the required credibility, extracting and classifying the keywords of the complaint text information, determining the core keywords by identifying the semantics, domain attributes and context association; for the user of the complaint picture and video information, the complaint picture and video information need to meet the preset shooting specification and add watermark; S4, distributing the complaint work order to the corresponding government department for governance according to the government responsibility matching degree and the complaint type.

[0023] Based on steps S1-S4, the data in the traditional governance mode is scattered in property, community, government and other systems, lacks unified correlation mechanism and data integration is difficult, resulting in information silos. Existing complaint positioning relies on text description, and the spatial positioning accuracy is insufficient, which cannot be accurate to a specific location, resulting in high deviation rate of work order allocation and low disposal efficiency. The community governance model is mostly static structure, and the whole model needs to be reconstructed when new complaints or facility changes occur, which is time-consuming and resource-intensive. The model updating cost is high, and it is difficult to adapt to the demand of high-frequency dynamic event processing. The authenticity of complaint information lacks hierarchical verification mechanism, and there is a phenomenon of false complaints or malicious harassment, which interferes with the allocation of government resources and reduces the governance credibility. The method realizes intelligent governance through four steps of model construction, information collection, credibility verification and work order allocation. By constructing a community governance model containing primary blocks, secondary blocks and tertiary blocks, text, picture and video complaint information is collected and matched to the smallest community governance unit of the tertiary node according to the spatial index; the credibility score is generated by complaint history and credit record, and biological characteristics, dynamic action verification and manual audit are performed on medium and low credibility users; the text information is divided into categories and the core keywords are extracted, the pictures are taken according to the three standards of panorama, medium shot and close-up and the space-time watermark is added, and finally the work order is allocated according to the matching degree of government responsibilities; after the government department receives the complaint work order, the problem type and complaint location coordinates of the complaint work order are extracted, and the corresponding responsible person is sent according to the responsibility matching degree, then the responsible person matches the nearest front-line staff based on the complaint location coordinates, then the front-line staff conducts on-site investigation and disposal, and at the same time the front-line staff uploads the watermarked on-site photos and construction and maintenance videos to the grassroots social wisdom governance system, and the system automatically pushes the result report to the user after the complaint problem is disposed.

[0024] The embodiment updates the model dynamically and locally, only updates a single smallest community governance unit, thereby avoiding full reconstruction of the model, minimizing resource occupation, and reducing the deviation rate of complaint work order allocation, and improving the response speed and resource utilization of grassroots governance as a whole.

[0025] In one embodiment, the step S1 of constructing a community governance model, collecting user complaint information and storing it in the community governance model, includes: S11, the construction of community governance model is divided into two steps, first, collect multi-dimensional data of community comprehensively, provide original input for community governance model, then store the collected multi-dimensional data in the community governance model after standardization processing; S12, the cell management model combines the physical space grid to form a three-level block of the cell management model, the three-level block is respectively a first-level block, a second-level block and a third-level block, the first-level block is a complete cell boundary, including cell basic information, the cell basic information includes cell name, total land area, total number of households and property company, the second-level block is a functional area, which divides a sub-area according to use, such as high-rise residential area, multi-storey residential area, public supporting area and commercial logistics area, and the third-level block is a grid divided in the second-level block, each grid corresponds to a specific position, such as the area in front of the door of 1# building 2 unit and the entrance of the underground garage of 3# building; S13, the first-level block, the second-level block and the third-level block are respectively encoded according to hierarchical progressive encoding; wherein the first-level block encoding adopts a cell name + fixed encoding format, such as XX cell-001, the second-level block encoding adopts the first-level block encoding-function area name-function area serial number, such as XX cell-001-high-rise residential area-01, and the third-level block encoding adopts the second-level block encoding-third-level block center point horizontal and vertical coordinates, such as XX cell-001-high-rise residential area-01-0305; 0305 represents the 3rd grid on the horizontal axis and the 5th grid on the vertical axis; S14, the complaint information is updated in combination with the first-level block, the second-level block and the third-level block linkage, the mapping relationship between the complaint location coordinates and the third-level block is constructed based on the space index mode, and is stored to the cell management model.

[0026] Based on steps S11-S14, the multi-dimension data collection of the community is divided into spatial data, humanistic data and dynamic event data. The spatial data includes community boundary coordinates and building distribution, which are obtained by aerial photography of a drone. The building distribution includes the location, height and unit number of the building. The humanistic data includes resident information and functional division. The resident information is obtained by interfacing with a property registration system. The resident information includes the number of residents and population structure of the community. The dynamic event data is complaint information. After the collection is completed, the community management model performs standardized processing on the collected data to unify the format and ensure data consistency. The first-level block monthly updates community basic information and merges the second-level block data. The second-level block daily updates complaint information and merges the third-level block data. The third-level block receives picture and video complaint information. The community management model analyzes the complaint location through the watermark position coordinates of the complaint information and matches to the corresponding third-level block. If the third-level block receives text keywords, the community management model matches to the corresponding third-level block through the core keyword list of the complaint information. The first-level block of the three-level block division of the community management model includes the basic information of the community and is monthly updated. The second-level block is divided according to the functional division and daily updates complaint information. The third-level block is a 10m x 10m grid unit corresponding to a specific location and real-time updates complaint information. Through the coding of the first-level block, the second-level block and the third-level block, the spatial position information is embedded in the coding to realize the dynamic mapping of the coding and the complaint location in the complaint information. The complaint information (text keywords, picture watermark position coordinates) is matched to the corresponding third-level block through spatial indexing, such as the text keyword "elevator failure" matched to the third-level block "XX community-01-high-rise residential area-01-0305". The above steps embed spatial position information through three-level block coding, combine the index table to map the complaint location coordinates and the three-level block coding, thereby improving the positioning accuracy and data integration efficiency.

[0027] In one embodiment, the step S14 of combining the first-level block, the second-level block and the third-level block for linked update of the complaint information, constructing the mapping relationship between the complaint location coordinates and the third-level block based on spatial indexing, and storing to the community management model, comprises: S141, establishing a root node. Each root node is divided into four quadrants with the center point of the community as the origin. The coordinates of the center points of the four quadrants are taken as first-level nodes. S142, each first-level node is divided into four quadrants again with the center point of the first-level node as the origin. The coordinates of the center points of the four quadrants are taken as second-level nodes. The four quadrants of the second-level nodes correspond to building groups of the community. S143, each second-level node is divided into four quadrants again with the center point of the second-level node as the origin. The coordinates of the center points of the four quadrants are taken as third-level nodes. The third-level nodes are the smallest management units of the community. S144, construct an index table, store the mapping relationship between the three-level block code and the three-level node center point coordinate based on the cell management model; S145, match the complaint location coordinate, the three-level block code and the three-level node center point coordinate according to the index table, mark the complaint location coordinate, the three-level block code and the three-level node center point coordinate as complaint data, and complete the dynamic local update of the cell management model.

[0028] Based on steps S141-S145, the root node is divided into four quadrants with the cell center point as the origin, the four quadrants are equal quadrants, the four quadrants are divided into northeast, southeast, northwest and southwest, and the center point coordinates of the four quadrants are the first-level nodes. Each first-level node is divided into four quadrants again with the first-level node center point as the origin, and the center point coordinates of the four quadrants are the second-level nodes. The second-level nodes correspond to the cell building group. Each second-level node is divided into four quadrants again with the second-level node center point as the origin, and the center point coordinates of the four quadrants are the third-level nodes. The third-level nodes are defined as the smallest management unit of 10m x 10m, forming an index path of root node-first-level node-second-level node-third-level node. By storing the mapping relationship between the three-level block code and the three-level node center point coordinate, the corresponding three-level node path is queried by inputting the complaint location coordinate, and the three-level block code is matched to complete accurate positioning. Through the three-level node hierarchical division of 10m grid unit and the index table mapping, the complaint location positioning accuracy is improved from the traditional building level to the specific grid level, and the complaint work order distribution deviation rate is further reduced.

[0029] In one embodiment, the step S145 of matching the complaint location coordinate, the three-level block code and the three-level node center point coordinate according to the index table, marking the complaint location coordinate, the three-level block code and the three-level node center point coordinate as complaint data, and completing the dynamic local update of the cell management model, includes: S1451, query the complaint location coordinate, the three-level node center point coordinate and the associated three-level block code to be updated in the cell management model based on the complaint data and the index table, and mark them as target units; S1452, construct a static anchoring base for the target units, and store the complaint data to the static anchoring base. The static anchoring base is a processing carrier for the complaint data, and the static anchoring base interacts with the target units and the cell management model information respectively. The static anchoring base includes a read-write control unit, a hash verification unit and a metadata storage unit. The read-write control unit integrates a distributed lock interface, which is used for subsequent locking / unlocking of the storage space of the static anchoring base. The hash verification unit generates a unique reference hash code based on the complaint data of the target units. The metadata storage unit is used to store the complaint data of the target units, including the complaint location coordinate, the three-level block code, the three-level node center point coordinate and the index path; S1453, establish a bidirectional pointer anchor between the target unit and the static anchor base, the target unit generates a pointer to the static anchor base through memory address mapping to read complaint data in the static anchor base; the static anchor base reversely generates a pointer to the cell management model through three-level block encoding to write complaint data, and ensures the connection effective by comparing the hash check codes of the pointers at both ends; S1454, configure a preset number of virtual carriers to grab complaint data of the target unit; S1455, apply a distributed read-write lock through a read-write control unit of the static anchor base, apply a write lock to complaint data of the target unit, and apply a read lock to the index path associated with a non-target unit, and generate real-time check codes by a hash check unit to shield write operation requests of the non-target unit during the locking period; S1456, the static anchor base releases the data stream binding with the cell management model, generates an incremental data snapshot of the target unit, writes complaint data into the incremental data snapshot by the preset number of virtual carriers, and then writes the incremental data snapshot into the static anchor base; S1457, release the distributed read-write lock applied by the read-write control unit, the static anchor base synchronizes complaint data to the cell management model through the bidirectional pointer, updates the index table mapping relationship, releases the read lock of the non-target unit and restores the global read-write permission of the cell management model.

[0030] The complaint position coordinates and the three-level block code in the user complaint information received in S1451 are matched and queried through an index table of the cell management model, the index table stores a mapping relationship between the three-level block code and the three-level node center point coordinates, and the query result is marked as a target unit according to the above S1451-S1457 steps. The static anchoring base in S1452 is constructed by allocating an independent storage space for the target unit through the metadata storage unit. In the bidirectional pointer anchoring process of the target unit and the static anchoring base, the hash check unit verifies the validity of the bidirectional pointer by calculating the combined hash value (target unit ID + base ID) of the target unit ID and the base ID. If the hash values are inconsistent, the hash check unit determines that the pointer is broken or the data is tampered with, triggering a retry mechanism to ensure the accuracy of the anchoring relationship. The bidirectional pointer in S1453 includes “target unit → static anchoring base pointer”: the target unit generates a pointer pointing to the physical storage address of the static anchoring base through memory address mapping, generates a forward pointer, and is used to read complaint data in the static anchoring base; it also includes “static anchoring base → cell management model pointer”: the static anchoring base reversely generates a pointer pointing to the cell management model by analyzing the three-level block code of the target unit, generates a reverse pointer pointing to the three-level node path of the main model, and is used to write complaint data; and the connection is ensured to be effective by comparing the hash check codes of the two pointers: calculating the combined hash value Hash (target unit ID + static anchoring base ID, such as target unit 1 + static anchoring base 1), comparing whether the hash values of the target unit and the static anchoring base are consistent, and regenerating the pointer if they are inconsistent.

[0031] The virtual carrier in S1454 is configured with virtual carrier 1, virtual carrier 2 and virtual carrier 3, virtual carrier 1 grabs complaint position coordinates, virtual carrier 2 grabs three-level block code index values, and virtual carrier 3 grabs three-level node center point coordinates, through the cooperative work of multiple virtual carriers, the data grabbing efficiency is improved. The write lock in S1455 is applied to the complaint data of the target unit, an exclusive lock is adopted, only updated complaint data is allowed to write, and any other write operation is prevented; the exclusive lock is an exclusive resource locking mechanism, the core feature is: only one process is allowed to hold the lock and perform write operation at the same time, and any read and write operation on the target unit by other processes is prohibited, thereby ensuring the safety of the modification of the complaint data of the target unit.

[0032] The read lock is applied to the association index path of the non-target unit, a shared lock is used to allow reading but prevent writing, and the data of the non-target unit during updating is ensured not to be tampered with; the application process of the read-write lock calls the distributed lock interface through the read-write control unit, generates a lock identifier based on the target unit ID and the index path, the write lock is applied by applying an exclusive lock to the complaint data of the target unit, and any process is prohibited from modifying the complaint data during the locking period; the read lock is applied by applying a shared lock to the index path (such as "root node-first level node-second level node-third level node") of the non-target unit, only reading the index table mapping relationship is allowed, and the write operation request is shielded. The incremental data snapshot in step S1456 is a lightweight data packet containing only complaint data generated after the static anchor base is unbound from the cell management model data stream, avoiding full update of the cell management model, realizing efficient updating of the target unit, and the incremental data snapshot includes complaint position coordinates and three-level block encoding. In step S1457, after the distributed read-write lock is successfully applied and the target unit and the non-target unit lock state are confirmed to be correct, the distributed read-write lock is unbound, and the unbinding process of the distributed read-write lock is: the static anchor base disconnects the data channel with the cell management model, and stops the data stream transmission between the target unit and the cell management model.

[0033] In one embodiment, the step S2 of calling the user's historical behavior record in the government complaint system in a preset period and verifying the user's credibility includes: S21, query the user complaint history, the complaint history including the number of complaints, complaint type and complaint content; S22, query whether the user has a false complaint or malicious harassment bad faith record; S23, generate a credibility score based on the credibility scoring rules to generate the user's credibility.

[0034] As described in steps S21-S23 above, the bad faith record refers to the user's bad faith behavior recorded by the government system due to false complaints, malicious harassment, providing fake evidence, etc. If the user has no bad faith record, the score is 100 points. The user's score is deducted by 20 points for each bad faith behavior. If the user has a single bad faith behavior, the score is 80 points, and the subsequent complaint verification needs to be strengthened. If the user has two bad faith behaviors, the score is 60 points, and the user is directly judged as low credibility, triggering the strengthened verification. The user's credibility score is divided into high, medium and low three categories. The high credibility user score is ≥80 points, which can pass the verification directly. The medium credibility user score is 60 points ≤ medium credibility user score < 80 points, which needs to strengthen the subsequent complaint verification. The low credibility user score is < 60 points, which is directly judged as low credibility, triggering the strengthened verification. The above steps based on the credit image of historical behavior and bad faith record can reduce the false complaint rate and simultaneously reduce the invalid work order processing amount of the government department.

[0035] In one embodiment, the step S2 of the medium credibility user needing to complete the biometric feature comparison and specified action execution verification includes: S24, a trusted user full-face image is collected through the front camera of the user's mobile device, facial key area features are extracted, and a unique feature template is generated; S25, the medium trustworthiness user full-face image is compared with the face data in the government identity information library in a 1:1 manner; S26, after the face data comparison is successful, specified action execution verification is performed, the system randomly selects 1-2 actions from a preset action library, and guides the medium trustworthiness user to execute in the form of text.

[0036] As described in steps S24-S26, the facial key area features of the user include eye area, nose area, mouth area, and contour area, and the actions in the preset action library include nodding action, shaking action, mouth opening action, blinking action, eyebrow lifting action, and head turning action. When nodding, the user is required to swing the head vertically up and down, with a swing amplitude ≥ 30°; when shaking, the user is required to swing the head horizontally left and right, with a swing amplitude ≥ 45°; when opening the mouth, the user's mouth is vertically opened by a distance ≥ 2 cm, which is 1 / 5 of the face width; when blinking, the user's eyes are closed and opened simultaneously for a period of 0.5-1 second; when lifting the eyebrows, the user's eyebrows are moved up by ≥ 5 mm; and when turning the head, the user's face is turned to the left and right by ≥ 30°, and the contralateral ear is exposed. The system randomly selects 1-2 non-repeating actions from the preset action library to generate an action sequence, and guides the user to execute through text instructions such as “please complete the nodding action”. If both actions are verified, it is passed. The setting of this step can reduce malicious complaint interference and reduce the number of invalid complaints.

[0037] In one embodiment, the step S2 of triggering enhanced verification by the low trustworthiness user includes: S27, a low trustworthiness user full-face image is collected through the front camera of the user's mobile device, facial key area features are extracted, and a unique feature template is generated; S28, the low trustworthiness user full-face image is compared with the face data in the government identity information library in a 1:1 manner; S29, after the face data comparison is successful, specified action execution verification is performed, the system randomly selects at least three actions from a preset action library, and guides the medium trustworthiness user to execute in the form of text; S30, the low trustworthiness user complaint content is audited by an artificial, and if there are 2 or more false complaint records, the complaint function is limited.

[0038] As described in steps S24-S26, the low credibility user triggers the enhanced verification, increases the action verification and manual review, randomly selects 3-4 non-repeating actions from the preset dynamic action library to generate a verification sequence, such as shaking head, opening mouth, raising eyebrows and turning head. After the action verification is passed, the system automatically pushes the user complaint information, including text, picture watermark and three-level block coding positioning information, to the government platform for independent judgment by the reviewer. The reviewer judges whether the text keywords and picture watermark position match, and whether the picture meets the shooting specification. If the system finds that the user has ≥2 false complaint records within one year, the system triggers the complaint function restriction, restricts the complaint function through the accumulation of false complaint records, forms a credit cost deterrent, thereby reducing the user's repeated complaint rate, and to some extent, limiting the malicious user's complaint authority and reducing the invalid interference on the property, community grassroots units.

[0039] In one embodiment, the step S3 of extracting and classifying the complaint text information to determine the core keywords based on the semantic, domain attribute and context association includes: S31, based on the dictionary, the complaint text information is segmented and the part of speech is marked; S32, the complaint text information after segmentation is classified; S33, the core keywords are determined based on the frequency of the complaint text information after segmentation and classification.

[0040] As described in steps S31-S33, the professional terms in the grassroots governance field, government special vocabulary and Modern Chinese Dictionary are integrated to build a segmentation dictionary library. The professional terms in the grassroots governance field usually include elevator failure, garbage classification and occupation of the road. The government special vocabulary usually includes work order, settlement and follow-up; the part of speech marking system is defined by Chinese Part-of-Speech Tag Set, which usually includes nouns such as elevator, garbage and street light; verbs include damage, blockage and occupation; adverbs include underground, in the corridor and east of the square. The complaint text information classification includes three categories: public facilities, environmental sanitation and public security prevention and control. The complaint text information of public facilities includes elevators, street lights and fitness equipment; the complaint text information of environmental sanitation includes garbage collection and transportation, sewage discharge and noise pollution; the complaint text information of public security prevention and control includes strangers breaking in, high-altitude throwing and private wire pulling; the core keywords are determined based on the frequency of the complaint text information after segmentation and classification, and stored in the core keyword list, which is matched to the corresponding three-level block coding through the core keyword list of the complaint text information, thereby improving the accuracy of work order allocation and reducing the cost of manual intervention.

[0041] In one embodiment, the step S3 of collecting picture information of the complaint information includes: S34, guiding the user to take pictures according to three levels of standards, panorama, medium shot and close-up, the panorama picture information includes complete scene environment, the medium shot picture information focuses on the subject, and the close-up picture information presents detailed features; S35, respectively determining the shooting distance and horizontal angle of view of the panorama, medium shot and close-up picture information; S36, respectively adding time and location watermarks to the panorama, medium shot and close-up picture information at different positions of the picture.

[0042] As described in steps S34-S36, the panorama picture information is used to completely present the spatial relationship of the complaint scene, including the relative position of the subject and the surrounding fixed reference; the medium shot picture information is used to focus on the complete form of the subject, including the boundary, size, color and relationship with the directly related object, and the subject area ratio is greater than or equal to 60%, solving the problem of blurred subject in traditional single picture; the close-up picture information is used to highlight the microscopic features of the problem, and the detail area ratio is greater than or equal to 80%, providing key visual evidence for subsequent responsibility identification. The horizontal angle of view of the panorama picture information is not less than 120 degrees, the shooting distance is not less than 10 meters, the vertical angle of view covers the top of the scene to the ground, and the watermark position is located at the lower right corner of the picture; the horizontal angle of view of the medium shot picture information is 50-80 degrees, the shooting distance is 3-5 meters, and the picture subject area ratio is not less than 60%, and the watermark position is located at the lower left corner of the picture; the horizontal angle of view of the close-up picture information is 20-30 degrees, the shooting distance is 0.5-2 meters, and the problem detail area ratio is not less than 80%, and the watermark position is located at the top of the picture. The combination of the above three levels of pictures improves the completeness of the complaint scene information, and the remote research and judgment accuracy of the government personnel is improved, thereby reducing the frequency of on-site review.

[0043] A grassroots social wisdom governance system based on big data analysis, comprising: A model construction module connected with the data acquisition module, used for constructing a community governance model; A data acquisition module connected with the credibility evaluation module, used for acquiring complaint information, A credibility evaluation module connected with the work order scheduling module, used for evaluating user credibility, A work order scheduling module for distributing complaint work orders to corresponding government departments.

[0044] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any skilled person in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A grassroots wisdom governance method based on big data analysis, characterized in that: Comprise the following steps: A community management model is constructed, user complaint information is collected and stored in the community management model, and the complaint information includes text, picture and video information; The historical behavior records of the user in the government affair complaint system within a preset time period are called, and the user credibility is verified. High credibility users can be directly verified, medium credibility users need to complete biological feature comparison and specified action execution verification, and low credibility users trigger enhanced verification; The complaint information of the user with the required credibility is analyzed, the key words of the complaint text information are extracted and classified, the core key words are determined by identifying the semantics, domain attributes and context association, and the complaint picture and video information of the user are required to meet the preset shooting specification and add a watermark; According to the government responsibility matching degree and the complaint type, the complaint work order is distributed to the corresponding government department for management. 2.The method of claim 1, wherein the method is characterized in that: The step of constructing the community management model, collecting user complaint information and storing it in the community management model, and processing the complaint information as picture and video information, comprises: Collecting multi-dimensional data of the community as the original input for constructing the community management model, and storing the multi-dimensional data in the community management model after standardization processing; The community management model combines with the physical space grid to form a three-level block of the community management model, the three-level block is a first-level block, a second-level block and a third-level block, the first-level block is a complete community boundary and community basic information, the second-level block is a functional area, which divides a sub-area according to the use, and the third-level block is a grid divided in the second-level block, each grid corresponds to a specific position of the community; The first-level block, the second-level block and the third-level block are encoded according to the hierarchical progression, and are recorded as first-level block encoding, second-level block encoding and third-level block encoding; The complaint information is updated in combination with the first-level block, the second-level block and the third-level block, a mapping relationship between the complaint position coordinates and the third-level block is established based on the spatial index, and is stored in the community management model. 3.The method of claim 2, wherein the method further comprises: The step of updating the complaint information in combination with the first-level block, the second-level block and the third-level block, and establishing the mapping relationship between the complaint position coordinates and the third-level block based on the spatial index, and storing it in the community management model, comprises: A root node is established based on the community management model, each root node is divided into four quadrants with the center point of the community as the origin, and the coordinates of the center points of the four quadrants are taken as first-level nodes, the first-level nodes are the center points of the first-level blocks; Each first-level node is divided into four quadrants again with the center point of the first-level node as the origin, and the coordinates of the center points of the four quadrants are taken as second-level nodes, the second-level nodes are the center points of the second-level blocks; Each second-level node is divided into four quadrants again with the center point of the second-level node as the origin, and the coordinates of the center points of the four quadrants are taken as third-level nodes, the third-level nodes are the smallest management units of the community, and the third-level nodes are the center points of the third-level blocks; An index table is constructed, and the mapping relationship between the third-level block encoding and the center point coordinates of the third-level nodes is stored based on the community management model; According to the index table, the complaint position coordinates, the third-level block encoding and the center point coordinates of the third-level nodes are matched, and the complaint position coordinates, the third-level block encoding and the center point coordinates of the third-level nodes are recorded as complaint data, and the dynamic local update of the community management model is completed.

4. The method of claim 3, wherein the method is characterized by: The step of matching the complaint location coordinates, the three-level block code and the three-level node center point coordinates according to the index table, and marking the complaint location coordinates, the three-level block code and the three-level node center point coordinates as complaint data, and completing the dynamic local update of the community management model, comprises: Querying the complaint data and the index table to obtain the complaint location coordinates, the three-level node center point coordinates and the associated three-level block code in the community management model to be updated, and marking them as target units; A static anchoring base is constructed for the target units, and the complaint data is stored in the static anchoring base, wherein the static anchoring base interacts with the target units and the community management model information respectively; the static anchoring base comprises a read-write control unit, a hash check unit and a metadata storage unit, the read-write control unit is integrated with a distributed lock interface, which is used for subsequent locking / unlocking of the storage space of the static anchoring base, the hash check unit generates a unique reference hash code based on the complaint data of the target units, and the metadata storage unit is used for storing the complaint data of the target units; A bidirectional pointer anchor is established between the target units and the static anchoring base, the target units generate a pointer pointing to the static anchoring base through memory address mapping, which is used to read the complaint data in the static anchoring base; the static anchoring base reversely generates a pointer pointing to the community management model through the three-level block code, which is used to write the complaint data, and the connection is ensured to be effective by comparing the hash check codes of the two pointers; A preset number of virtual carriers are configured to capture the complaint data of the corresponding target units; The read-write control unit of the static anchoring base applies a distributed read-write lock, the complaint data of the target units is subjected to a write lock, and the index paths associated with non-target units are subjected to a read lock, and during the locking period, the hash check unit generates real-time check codes to shield the write operation requests of non-target units; The static anchoring base is unbound from the data stream of the community management model, and an incremental data snapshot of the target units is generated, the complaint data is written into the incremental data snapshot by the preset number of virtual carriers, and then the incremental data snapshot is written into the static anchoring base; The distributed read-write lock applied by the read-write control unit is released, the static anchoring base synchronizes the complaint data to the community management model through the bidirectional pointer, updates the mapping relationship of the index table, releases the read lock of the non-target units and restores the global read-write permission of the community management model.

5. The method of claim 1, wherein the method is characterized by: The step of calling the historical behavior records of the user in the government complaint system in the past preset period and verifying the credibility of the user, comprises: Querying the complaint history of the user, the complaint history including the number of complaints, the type of complaint and the complaint content; Querying whether the user has a false complaint or malicious harassment record of bad faith; Generating a credibility score based on a credibility scoring rule to generate the credibility of the user.

6. The grassroots social wisdom governance method based on big data analysis according to claim 1, characterized in that: The step of the medium credibility user needing to complete biological feature comparison and specified action execution verification, comprises: Capturing the full-face image of the medium credibility user through the front camera of the user's mobile device, extracting the key area features of the face, and generating a unique feature template; 1:1 comparison of the full-face image of the medium credibility user with the face data in the government identity information library; After the face data comparison is successful, specified action execution verification is performed, the system randomly selects at least one action from a preset action library, and guides the medium credibility user to perform the action in the form of text.

7. The method of claim 1, wherein the method is characterized by: The step of triggering the enhanced verification by the low credibility user comprises: Collecting a full-face image of the low credibility user through a front camera of a mobile device of the user, extracting a feature of a key area of the face, and generating a unique feature template; Performing 1:1 comparison between the full-face image of the low credibility user and face data in a government identity information library; After the face data comparison is successful, specified action execution verification is performed, the system randomly selects at least three actions from a preset action library, and guides the medium credibility user to perform the actions in the form of text; Through artificial auditing of the complaint content of the low credibility user, if there are more than 2 false complaint records, the complaint function of the low credibility user is limited.

8. The method of claim 2, wherein the method is characterized by: The step of extracting and classifying the core keywords of the complaint text information by identifying semantics, domain attributes and context association comprises: Carrying out word segmentation on the complaint text information based on a dictionary, and marking the part of speech; Classifying the complaint text information after the word segmentation; Determining the core keywords based on the frequency of the complaint text information after the word segmentation and classification. 9.The method of claim 1, wherein the method further comprises: receiving a request for a social governance solution from a user; and providing the user with a list of social governance solutions based on the request. The step of collecting picture information of the complaint information comprises: Guiding the user to shoot picture information according to three-level standards of panorama, medium shot and close-up, the panorama picture information includes a complete scene environment, the medium shot picture information focuses on the subject, and the close-up picture information presents detailed features; Determining the shooting distance and horizontal angle of view of the panorama, medium shot and close-up picture information respectively; Adding time and location watermarks to the panorama, medium shot and close-up picture information at different positions of the pictures. 10.A grassroots social wisdom governance system based on big data analysis, used to implement the grassroots social wisdom governance method based on big data analysis of any one of claims 1-9, characterized in that, Comprise: A model construction module connected with the data collection module, used for constructing a community management model; A data collection module connected with the credibility evaluation module, used for collecting complaint information; A credibility evaluation module connected with the work order scheduling module, used for evaluating the credibility of the user; A work order scheduling module, used for distributing complaint work orders to corresponding government departments, the complaint work orders including complaint text information, complaint picture and video information, three-level block codes corresponding to complaint location coordinates and user credibility information.

Images