Epidemic situation early warning information intelligent management system based on big data

Abstract

The invention discloses an epidemic situation early warning information intelligent management system based on big data, and relates to the technical field of data management. The method comprises the following steps: acquiring physical sign data of different users, performing preprocessing and distributed storage on the physical sign data, acquiring trajectory data of the different users, constructing a space-time association diagram of the different users, constructing a normal physical sign map and an abnormal physical sign map to acquire physical sign evaluation coefficients of the different users, dividing the space-time association diagram into a plurality of sub-regions, the prevention and control level of each sub-region and the track evaluation coefficients and comprehensive evaluation coefficients of different users are obtained, the comprehensive evaluation coefficients of different users are compared with a preset evaluation threshold value to obtain key prevention and control users, and early warning information is generated for feedback; whether the user is infected or not can be evaluated from different dimensions, and a proactive prevention and control mechanism can be formed.

Description

Technical Field

[0001] The present invention relates to the technical field of data management, and in particular to an intelligent management system for epidemic warning information based on big data. Background Art

[0002] By using big data technology to collect, store, analyze and mine multi-source data related to the epidemic, and predicting and warning the development trend of the epidemic through intelligent algorithms, the effective management and release of warning information can be realized, which can help public health departments, medical institutions and schools to timely master the epidemic situation, make scientific decisions and take effective prevention and control measures;

[0003] When there is a tuberculosis epidemic in the campus area, there is a lack of a technical means to timely monitor the physical conditions of students. Most of the existing prevention and control methods stay at physical sign monitoring, and fail to comprehensively evaluate other students in combination with the whereabouts of infected students, resulting in the inability to accurately judge whether students are at risk of being infected. In view of the deficiencies of the existing technology, the present invention provides an intelligent management system for epidemic warning information based on big data. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent management system for epidemic warning information based on big data.

[0005] The purpose of the present invention can be achieved by the following technical solutions: An intelligent management system for epidemic warning information based on big data, including the following modules:

[0006] A data acquisition module, which is used to obtain the physical sign data of different users, preprocess and distribute the storage thereof respectively, obtain the trajectory data of different users, and construct a spatio-temporal association graph of different users;

[0007] A physical sign evaluation module, which is used to construct a normal physical sign map and an abnormal physical sign map according to the physical sign data of different users, and compare the normal physical sign map and the abnormal physical sign map to obtain the physical sign evaluation coefficients of different users;

[0008] A trajectory evaluation module, which is used to divide several sub-regions in the spatio-temporal association graph, obtain the prevention and control levels of each sub-region, and use the trajectory data combined with the prevention and control levels to obtain the trajectory evaluation coefficients and comprehensive evaluation coefficients of different users;

[0009] A data feedback module, which is used to compare the comprehensive evaluation coefficients of different users with a preset evaluation threshold respectively to obtain key prevention and control users, and generate corresponding warning information for feedback.

[0010] Furthermore, the process of obtaining the physical sign data of different users, preprocessing and distributing the storage thereof respectively includes:

[0011] Set the collection period. When a collection period is reached, obtain the physical sign data of different users once. The physical sign data includes body temperature, respiratory rate, pulse, blood oxygen saturation, weight change, cough condition, chest auscultation condition, and vaccination condition.

[0012] Divide users into diseased users and non-diseased users, and label their physical sign data as diseased data and non-diseased data. The preprocessing includes outlier processing, missing value processing, and duplicate value processing.

[0013] Set up several databases, and upload the preprocessed diseased data and non-diseased data to the databases corresponding to their users for storage.

[0014] Further, the process of obtaining the trajectory data of different users and constructing the spatio-temporal association graph of different users includes:

[0015] Obtain the building distribution information and the monitoring distribution information. The building distribution information refers to the distribution positions of each building in the area, and the monitoring distribution information refers to the distribution positions of each monitoring device in the area. Use GIS technology to construct the corresponding GIS distribution map according to the building distribution information, and upload the monitoring distribution information to the GIS distribution map for synchronization.

[0016] Use face recognition technology to perform face recognition and positioning on users based on each monitoring device. When a single monitoring device recognizes a single user, generate a trajectory point for the single user at the distribution position of the single monitoring device in the GIS distribution map.

[0017] The trajectory data refers to all the trajectory points generated for a single user by different monitoring devices within a single collection period. For diseased users, fix their trajectory lines within different collection periods in the GIS distribution map. For non-diseased users, fix their trajectory lines within the current collection period in the GIS distribution map. Use the GIS distribution map at this time as the spatio-temporal association graph of different users.

[0018] Further, the process of constructing the normal physical sign atlas and the abnormal physical sign atlas according to the physical sign data of different users includes:

[0019] Convert the physical sign data of a single user into a unified format and semantics, establish the association relationships between various entities in the physical sign data, and use ontology to define various entities and their association relationships in the physical sign data to obtain the semantic structure of the knowledge graph.

[0020] Use natural language processing technology to extract key information from the text data, reason about the key information to fill in the blanks in the knowledge graph, and combine the semantic structure to construct the data structure of the knowledge graph, and fill the key information into the knowledge graph.

[0021] For diseased users, the knowledge graph corresponding to their diseased data is used as the abnormal physical sign graph, and for non-diseased users, the knowledge graph corresponding to their non-diseased data is used as the normal physical sign graph.

[0022] Furthermore, the process of comparing the normal physical sign graph and the abnormal physical sign graph to obtain the physical sign evaluation coefficients of different users includes:

[0023] Set corresponding normal ranges for all quantifiable data in the normal physical sign graph and the abnormal physical sign graph. The normal range refers to the numerical range in which the physical sign data is in normal conditions. Compare the quantifiable data of a single user with its normal range respectively to obtain normal nodes and abnormal nodes, and obtain the physical sign evaluation coefficient P of this single non-diseased user in the current collection period t ;

[0024]

[0025] Record the number of abnormal nodes of a single non-diseased user in the current collection period as Y a , and record the average value of the number of abnormal nodes of all diseased users in the current collection period as Y b .

[0026] Furthermore, the process of dividing several sub-regions in the spatio-temporal correlation graph and obtaining the prevention and control levels of each sub-region includes:

[0027] Divide the entire region into several sub-regions in the spatio-temporal correlation graph, obtain the number of abnormal trajectory points of the trajectory points corresponding to the diseased users in each sub-region, set the range of the number of trajectory points, and compare the number of abnormal trajectory points of each sub-region with the range of the number of trajectory points to obtain the prevention and control levels of each sub-region. The prevention and control levels are divided into first-level prevention and control, second-level prevention and control, and third-level prevention and control.

[0028] Furthermore, the process of using trajectory data combined with the prevention and control level to obtain the trajectory evaluation coefficient and comprehensive evaluation coefficient of different users includes:

[0029] Obtain the number of times the trajectory points generated by the trajectory data of a single non-diseased user in the current collection period appear in sub-regions with different prevention and control levels, which are the number of first-level coincidence points C1, the number of second-level coincidence points C2, and the number of third-level coincidence points C3 respectively;

[0030] Set corresponding weight values for the number of coincidence points at different prevention and control levels, which are recorded as q1, q2, and q3 respectively, and obtain the trajectory evaluation coefficient P of this single non-diseased user g ;

[0031] P g= q1C1 + q2C2 + q3C3;

[0032] Obtain the corresponding comprehensive evaluation coefficient P according to the physical sign evaluation coefficient and trajectory evaluation coefficient of the single non-diseased user in the current collection period z ;

[0033] P z = q t P t + q g P g ;

[0034] q t 、q g are respectively the preset weight values of the physical sign evaluation coefficient and the trajectory evaluation coefficient, and obtain the comprehensive evaluation coefficients of each non-diseased user in the current collection period respectively.

[0035] Furthermore, the process of comparing the comprehensive evaluation coefficients of different users with the preset evaluation threshold respectively to obtain the key prevention and control users and generating the corresponding warning information for feedback includes:

[0036] Set the evaluation threshold, compare the comprehensive evaluation coefficients of each non-diseased user in the current collection period with the evaluation threshold respectively to obtain the key prevention and control users, generate warning information, and feedback the warning information to the relevant personnel.

[0037] Compared with the prior art, the beneficial effects of the present invention are:

[0038] The present invention divides users into diseased users and non-diseased users, obtains diseased data and non-diseased data respectively, constructs abnormal physical sign maps and normal physical sign maps of diseased users and non-diseased users, obtains the number of abnormal nodes in the physical sign maps of different users, and uses the comparison situation of the normal physical sign map and the abnormal physical sign map to obtain the physical sign evaluation coefficients of different users, which is beneficial to realizing the first evaluation of non-diseased users on physical sign data;

[0039] By obtaining the trajectory data of different users and constructing the spatio-temporal association map of different users, it is beneficial to more intuitively display the places where different users have appeared together, divide several sub-regions in the spatio-temporal association map, and obtain the prevention and control levels of each sub-region according to the trajectory data of diseased users, and obtain the trajectory evaluation coefficients of each non-diseased user on this basis, which can realize the second evaluation. By combining the two evaluation results to obtain the corresponding comprehensive evaluation coefficient, it is possible to evaluate whether a user has the possibility of being infected from different dimensions, including their physical signs and whereabouts, and obtain potential key prevention and control users, which is beneficial to forming a forward-looking prevention and control mechanism. Description of the Drawings

[0040] Figure 1 This is the schematic diagram of the present invention. Detailed implementation manners

[0041] As Figure 1 shown, an intelligent management system for epidemic warning information based on big data includes the following modules:

[0042] A data collection module, which is used to obtain the physical sign data of different users, preprocess and distribute the storage of them respectively, obtain the trajectory data of different users, and construct a spatio-temporal association graph of different users;

[0043] A physical sign evaluation module, which is used to construct a normal physical sign map and an abnormal physical sign map according to the physical sign data of different users, and compare the normal physical sign map and the abnormal physical sign map to obtain the physical sign evaluation coefficients of different users;

[0044] A trajectory evaluation module, which is used to divide several sub-regions in the spatio-temporal association graph, obtain the prevention and control levels of each sub-region, and use the trajectory data combined with the prevention and control levels to obtain the trajectory evaluation coefficients and comprehensive evaluation coefficients of different users;

[0045] A data feedback module, which is used to compare the comprehensive evaluation coefficients of different users with the preset evaluation thresholds respectively to obtain the key prevention and control users, and generate corresponding warning information for feedback.

[0046] It should be further noted that in the specific implementation process, the process of obtaining the physical sign data of different users and preprocessing and distributing the storage of them respectively includes:

[0047] In the embodiment of the present invention, taking the prevention and control of tuberculosis epidemic among students in the campus area as an example, but the technical solution of the present invention is also applicable to other areas and other users. Set the collection period, and when a collection period is reached, obtain the physical sign data of different users respectively once;

[0048] The physical sign data refers to the relevant data collected for students to prevent and control tuberculosis epidemic, including body temperature, respiratory rate, pulse, blood oxygen saturation, weight change, cough condition, chest auscultation condition, vaccination condition, etc.;

[0049] Divide users into diseased users and non-diseased users, and mark their corresponding physical sign data as diseased data and non-diseased data respectively. Preprocess the diseased data and non-diseased data respectively, including outlier processing, missing value processing, duplicate value processing, etc.;

[0050] Treat each college as an independent unit for processing, set a corresponding database for each college respectively, and upload the preprocessed diseased data and non-diseased data to the databases of the colleges where their users are located for storage.

[0051] It should be further noted that in the specific implementation process, the process of obtaining the trajectory data of different users and constructing the spatio-temporal association graph of different users includes:

[0052] Obtain the building distribution information within the campus area. The building distribution information refers to the distribution positions of each building within the campus area, and use GIS technology to construct a GIS distribution map within the campus area based on the collected building distribution information;

[0053] Obtain the monitoring distribution information within the campus area. The monitoring distribution information refers to the distribution positions of each monitoring device within the campus area, and upload the obtained monitoring distribution information to the constructed GIS distribution map for synchronization;

[0054] Use face recognition technology to perform face recognition and positioning of users based on each monitoring device. When a single monitoring device recognizes a single user, a trajectory point is generated for the single user at the distribution position of the single monitoring device in the GIS distribution map;

[0055] The trajectory data refers to all the trajectory points generated for a single user by different monitoring devices within a single collection period, and their corresponding timestamps. Connect the trajectory points of the single user within a single collection period in chronological order to obtain the corresponding trajectory line;

[0056] For the diseased users, fix their trajectory lines in different collection periods in the GIS distribution map. For the non-diseased users, fix their trajectory lines in the current collection period in the GIS distribution map, and mark the GIS distribution map at this time as the spatio-temporal association graph of different users.

[0057] It should be further noted that in the specific implementation process, the process of constructing the normal physical sign atlas and the abnormal physical sign atlas according to the physical sign data of different users includes:

[0058] Convert the physical sign data of a single user into a unified format and semantics for easy representation in the knowledge graph, and establish the association relationships between various entities in the physical sign data. Use ontology to define various entities and their association relationships in the physical sign data to obtain the semantic structure of the knowledge graph;

[0059] Use natural language processing technology to extract key information from the text data, reason about the key information to fill in the blanks in the knowledge graph, and combine the obtained semantic structure to construct the data structure of the knowledge graph, and fill the obtained key information into the knowledge graph;

[0060] In the constructed knowledge graph, the physical sign data of the single user is processed by dividing it into nodes and edges. The node is the basic unit in the knowledge graph, representing the entity in the physical sign data. The edge is the relationship connecting the nodes, used to represent the association relationship between the nodes, and the edge describes the semantic structure between the nodes.

[0061] For the diseased user, the knowledge graph corresponding to his diseased data is marked as an abnormal physical sign graph. For the non-diseased user, the knowledge graph corresponding to his non-diseased data is marked as a normal physical sign graph.

[0062] It should be further noted that in the specific implementation process, the process of comparing the normal physical sign graph and the abnormal physical sign graph to obtain the physical sign evaluation coefficients of different users includes:

[0063] Set the corresponding normal ranges for all quantifiable data in the constructed normal physical sign graph and abnormal physical sign graph. Among them, the quantifiable data refers to all physical sign data that can be represented by numbers, and the normal range refers to the numerical range in which the physical sign data is in the normal situation.

[0064] Compare the quantifiable data of the single user with its normal range respectively. If it is within the normal range (including equality), then take the position of the quantifiable data in the corresponding physical sign graph as a normal node. If it is outside the normal range (excluding equality), then take the position of the quantifiable data in the corresponding physical sign graph as an abnormal node.

[0065] Record the number of abnormal nodes of a single non-diseased user in the current collection period as Y a , and record the average value of the number of abnormal nodes of all diseased users in the current collection period as Y b , and obtain the physical sign evaluation coefficient of this single non-diseased user in the current collection period, denoted as P t ;

[0066]

[0067] Adopt the same method to obtain the physical sign evaluation coefficients of each non-diseased user in the current collection period, and it is possible to evaluate the non-diseased users in terms of physical sign data according to the difference in the number of abnormal nodes between the non-diseased users and the diseased users.

[0068] It should be further noted that in the specific implementation process, the process of dividing several sub-regions in the spatio-temporal association graph and obtaining the prevention and control levels of each sub-region includes:

[0069] Divide the entire campus area in the spatio-temporal association graph to divide it into several sub-regions with equal areas, obtain the number of trajectory points corresponding to diseased users in each sub-region, and mark it as the number of abnormal trajectory points, denoted as Gs ;

[0070] Set the range of the number of trajectory points as [G min , G max . Compare the number of abnormal trajectory points in each sub-region with the range of the number of trajectory points respectively. If G s < G min , then mark it as first-level prevention and control. If G s > G max , then mark it as third-level prevention and control. If G min ≤ G s ≤ G max , then mark it as second-level prevention and control. The prevention and control levels are divided into first-level prevention and control, second-level prevention and control, and third-level prevention and control.

[0071] It should be further noted that in the specific implementation process, the process of obtaining the trajectory evaluation coefficient and the comprehensive evaluation coefficient of different users by using the trajectory data in combination with the prevention and control level includes:

[0072] Taking a single user as an example, obtain the number of times the trajectory points generated by the trajectory data of the single user in the current collection period appear in the sub-regions of different prevention and control levels, and record them as the number of first-level coincidence points C1, the number of second-level coincidence points C2, and the number of third-level coincidence points C3 respectively;

[0073] Set the corresponding weight values for the number of coincidence points at different prevention and control levels, and record them as q1, q2, q3 respectively, and obtain the trajectory evaluation coefficient of the single user, denoted as P g ;

[0074] P g = q1C1 + q2C2 + q3C3;

[0075] According to the physical sign evaluation coefficient P t and the trajectory evaluation coefficient P g of the single user in the current collection period, obtain its corresponding comprehensive evaluation coefficient, denoted as P z ;

[0076] P z = q t P t + q g P g ;

[0077] Among them, q t , q g are the weight values corresponding to the physical sign evaluation coefficient and the trajectory evaluation coefficient respectively. Use the same method to obtain the comprehensive evaluation coefficients of each non-diseased user in the current collection period respectively.

[0078] It should be further noted that in the specific implementation process, the process of comparing the comprehensive evaluation coefficients of different users with the preset evaluation threshold respectively to obtain the key prevention and control users and generating corresponding warning information for feedback includes:

[0079] Set the evaluation threshold P0, and compare the comprehensive evaluation coefficients of each non-diseased user in the current collection period with the evaluation threshold respectively. If P z ≥P0, mark it as a key prevention and control user, generate the corresponding warning information, and feedback the generated warning information to the relevant personnel to prompt the relevant personnel to conduct a physical examination on it in time. If P z <P0, do not perform any other operations on it.

[0080] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An intelligent management system for epidemic warning information based on big data, characterized in that, It includes the following modules: The data acquisition module is used to obtain the physical sign data of different users, preprocess and distribute and store them respectively, obtain the trajectory data of different users, and construct the spatio-temporal association graph of different users; The physical sign evaluation module is used to construct a normal physical sign atlas and an abnormal physical sign atlas according to the physical sign data of different users, and compare the normal physical sign atlas and the abnormal physical sign atlas to obtain the physical sign evaluation coefficients of different users; The trajectory evaluation module is used to divide several sub-regions in the spatio-temporal association graph, obtain the prevention and control levels of each sub-region, and use the trajectory data combined with the prevention and control levels to obtain the trajectory evaluation coefficients and comprehensive evaluation coefficients of different users; The data feedback module is used to compare the comprehensive evaluation coefficients of different users with the preset evaluation thresholds respectively to obtain the key prevention and control users, and generate corresponding warning information for feedback.

2. An intelligent management system for epidemic warning information based on big data according to claim 1, characterized in that, The process of obtaining the physical sign data, preprocessing and distributing and storing it includes: Set the acquisition period. When a acquisition period is reached, obtain the physical sign data of different users once. The physical sign data includes body temperature, respiratory rate, pulse, blood oxygen saturation, weight change, cough condition, chest auscultation condition, and vaccination condition; Divide users into diseased users and non-diseased users, and mark their physical sign data as diseased data and non-diseased data. The preprocessing includes outlier processing, missing value processing, and duplicate value processing; Set several databases, and upload the preprocessed diseased data and non-diseased data to the databases corresponding to their users for storage.

3. An intelligent management system for epidemic warning information based on big data according to claim 2, characterized in that, The process of obtaining the trajectory data and constructing the spatio-temporal association graph of different users includes: Obtain the building distribution information and the monitoring distribution information. The building distribution information refers to the distribution positions of each building in the area, and the monitoring distribution information refers to the distribution positions of each monitoring device in the area. Use GIS technology to construct the corresponding GIS distribution map according to the building distribution information, and upload the monitoring distribution information to the GIS distribution map for synchronization; Use face recognition technology to perform face recognition and positioning of users based on each monitoring device. When a single monitoring device recognizes a single user, generate a trajectory point for the single user at the distribution position of the single monitoring device in the GIS distribution map; The trajectory data refers to all the trajectory points generated for a single user by different monitoring devices within a single acquisition period. For diseased users, fix their trajectory lines within different acquisition periods in the GIS distribution map. For non-diseased users, fix their trajectory lines within the current acquisition period in the GIS distribution map. Use the GIS distribution map at this time as the spatio-temporal association graph of different users.

4. An intelligent management system for epidemic warning information based on big data according to claim 3, characterized in that, The process of constructing the normal physical sign atlas and the abnormal physical sign atlas includes: Convert the physical sign data of a single user into a unified format and semantics, establish the association relationships between various entities in the physical sign data, and use ontology to define various entities and their association relationships in the physical sign data to obtain the semantic structure of the knowledge graph; Using natural language processing technology to extract text data to extract key information, reasoning on the key information to fill in the blanks in the knowledge graph, and constructing the data structure of the knowledge graph in combination with the semantic structure, and filling the key information into the knowledge graph; For diseased users, the knowledge graph corresponding to their diseased data is used as the abnormal sign graph, and for non-diseased users, the knowledge graph corresponding to their non-diseased data is used as the normal sign graph.

5. An intelligent management system for epidemic warning information based on big data according to claim 4, characterized in that, The process of obtaining the physical sign evaluation coefficients of different users includes: Set corresponding normal ranges for all quantifiable data in the normal physical sign atlas and the abnormal physical sign atlas. The normal range refers to the numerical range in which the physical sign data is located under normal circumstances. Compare the quantifiable data of a single user with its normal range respectively to obtain normal nodes and abnormal nodes, and obtain the physical sign evaluation coefficient P of this single non-diseased user in the current collection period t ; Let the number of abnormal nodes of a single non-diseased user in the current collection period be denoted as Y a Let the mean value of the number of abnormal nodes of all diseased users in the current collection period be denoted as Y b .

6. The intelligent management system for epidemic warning information based on big data according to claim 5, wherein, The process of obtaining the prevention and control levels of each sub-region includes: In the spatio-temporal association graph, the entire region is divided into several sub-regions, the number of abnormal trajectory points of the trajectory points corresponding to the diseased users in each sub-region is obtained, a range of the number of trajectory points is set, and the number of abnormal trajectory points in each sub-region is compared with the range of the number of trajectory points to obtain the prevention and control levels of each sub-region. The prevention and control levels are divided into first-level prevention and control, second-level prevention and control, and third-level prevention and control.

7. An intelligent management system for epidemic warning information based on big data according to claim 6, characterized in that, The process of obtaining the trajectory evaluation coefficients and comprehensive evaluation coefficients of different users includes: Obtaining the number of times the trajectory points generated from the trajectory data of a single non-diseased user in the current collection period appear in sub-regions with different prevention and control levels, namely the number of first-level coincidence points C1, the number of second-level coincidence points C2, and the number of third-level coincidence points C3; Set corresponding weight values for the number of coincidence points under different prevention and control levels, denoted as q1, q2, and q3 respectively, and obtain the trajectory evaluation coefficient P of this single non-diseased user g ; P g = q1C1 + q2C2 + q3C3; Obtain the corresponding comprehensive evaluation coefficient P according to the physical sign evaluation coefficient and trajectory evaluation coefficient of the single non-diseased user in the current collection period z ; P z = q t P t + q g P g ; q t and q g are the preset weight values of the physical sign evaluation coefficient and the trajectory evaluation coefficient respectively, and the comprehensive evaluation coefficients of each non-diseased user in the current collection period are obtained respectively.

8. An intelligent management system for epidemic warning information based on big data according to claim 7, characterized in that, The process of obtaining key prevention and control users and generating warning information for feedback includes: Setting an evaluation threshold, comparing the comprehensive evaluation coefficients of each non-diseased user in the current collection period with the evaluation threshold respectively to obtain key prevention and control users, generating warning information, and feeding back the warning information to relevant personnel.

Images