Method, device, computer equipment and storage medium for determining emergency strategy

Abstract

The present application relates to a method, device, computer equipment, storage medium and computer program product for determining an emergency strategy. The method comprises: if a target event is detected, the target event characteristics are determined based on the target type of the target event, and the target event characteristics include common event characteristics and event characteristics corresponding to the target type; business support data is extracted based on the target event, and the covariance matrix is ​​obtained by calculation through the target event characteristics and the business support data; principal component analysis is performed based on the covariance matrix of the target event to obtain the principal component analysis results, and the emergency strategy of the target event is obtained by evaluation based on the principal component analysis results. By adopting this method, the correlation influence between various indicators can be eliminated, the data can be processed through a unified standard, the data quality in the database can be guaranteed, and the accuracy of the determined emergency strategy can be improved on the basis of reliable data.

Description

Technical Field

[0001] The present application relates to the technical field of principal component analysis, and in particular to a method, apparatus, computer equipment, storage medium and computer program product for determining an emergency strategy. Background Art

[0002] As the living standards of our people improve, the public's requirements for food are gradually increasing, and frequent dining out has led to the vigorous development of our catering industry. However, some discordant notes have emerged in this process, such as some emergency food safety and health incidents, which require emergency treatment of sudden health incidents.

[0003] In the related technology, the hazard factor database, emergency disposal case database and standard emergency plan database are integrated into a comprehensive database as the data set of the recommendation method, and then the emergency plan recommendation algorithm is composed of the hazard factor prediction model, standard evaluation model, backtracking model and expert decision-making model to make the first-time decision and give emergency plan recommendations based on the emergency event-related data features submitted by the user. However, the above scheme is only a method of judging its level based on the score threshold, and the analysis result is relatively simple, resulting in low accuracy of the determined emergency strategy. Summary of the invention

[0004] Based on this, it is necessary to provide a method, device, computer equipment, computer-readable storage medium and computer program product for determining an emergency strategy that can improve the accuracy of the emergency strategy in response to the above technical problems.

[0005] In a first aspect, the present application provides a method for determining an emergency strategy. The method comprises:

[0006] If a target event is detected, target event features are determined based on the target type of the target event, where the target event features include common event features and event features corresponding to the target type;

[0007] Extracting business support data based on the target event, and calculating through the target event features and the business support data to obtain a covariance matrix;

[0008] A principal component analysis is performed based on the covariance matrix of the target event to obtain a principal component analysis result, and an evaluation is performed based on the principal component analysis result to obtain an emergency response strategy for the target event.

[0009] In this embodiment, the correlation effects between various indicators can be eliminated, and the data can be processed through unified standards to ensure the data quality in the database, thereby improving the accuracy of the emergency strategy determined on the basis of reliable data.

[0010] In one embodiment, extracting business support data based on the target event includes:

[0011] Original data is acquired based on the target event, and the original data is preprocessed to generate business support data.

[0012] In this embodiment, the business support data is determined by the original data, so that the accuracy and comprehensiveness of data collection can be guaranteed.

[0013] In one embodiment, extracting business support data based on the target event includes:

[0014] The similarity between the target event and each sample event included in the business support database is calculated, and the business support data corresponding to the similar events whose similarity meets the preset similarity condition is used as the business support data of the target event.

[0015] In this embodiment, by extracting business support data from the business support database based on similarity, data extraction efficiency can be improved.

[0016] In one embodiment, the calculation by using the target event features and the business support data to obtain a covariance matrix includes:

[0017] In the business support database, determining the characteristic value corresponding to each target event feature, and generating a sample matrix based on each of the target event features and the characteristic value of each of the target event features;

[0018] The sample matrix is ​​standardized to obtain the column mean and standard deviation corresponding to the sample matrix, and a covariance matrix is ​​obtained based on the column mean and standard deviation corresponding to the sample matrix.

[0019] In this embodiment, by performing standardization processing on the data and calculating the covariance, data that is convenient for analysis and calculation can be obtained, providing a reliable data basis for subsequent principal component analysis.

[0020] In one embodiment, the principal component analysis is performed based on the covariance matrix of the target event to obtain the principal component analysis result, including:

[0021] Calculating correlation coefficients and sphericity test values ​​based on the covariance matrix of the target event;

[0022] In the case that the correlation coefficient satisfies the correlation condition and the sphericity test value satisfies the significance condition, a variance explanation table is calculated through the covariance matrix, and the number of each principal component is determined through the variance explanation table;

[0023] The latent variables of each principal component are analyzed through the variance explanation table to obtain a factor loading coefficient table, and the importance corresponding to each target principal component is determined based on the factor loading coefficient table.

[0024] In this embodiment, the accurate importance of each principal component is determined by the principal component analysis algorithm, so that the influencing factors with higher importance can be accurately located, the emergency events can be effectively handled, and the effectiveness of the determined emergency strategy can be ensured.

[0025] In one embodiment, the evaluating based on the principal component analysis result to obtain the emergency response strategy for the target event includes:

[0026] An evaluation is performed based on the principal component analysis result and the requirements of the target event to obtain factors to be improved for the target event, and each factor to be improved is used as an emergency strategy for the target event.

[0027] In this embodiment, the matching emergency strategy is determined by the influencing factors with higher importance, the effectiveness of the determined emergency strategy is guaranteed, the medical rescue capacity assessment is improved, and the medical emergency rescue capacity is improved.

[0028] In one embodiment, the method further comprises:

[0029] Determine the effectiveness of the emergency response strategy based on the target event and the target event;

[0030] When it is determined that the target event and the emergency strategy corresponding to the target event pass the validity check, the target event and the emergency strategy corresponding to the target event are added to the business support database;

[0031] When it is determined that the target event and the emergency strategy corresponding to the target event fail to pass the validity check, it is determined that the target event and the emergency strategy corresponding to the target event are added to the garbage sample data set.

[0032] In this embodiment, by targeting events and emergency strategies corresponding to the target events, it is possible to achieve the effect of accumulating past experience and distinguishing between good and bad sample events, recommending good sample events to be further strengthened, and focusing on avoiding bad sample events.

[0033] In a second aspect, the present application also provides a device for determining an emergency strategy. The device comprises:

[0034] A detection module, configured to determine target event features based on a target type of the target event if a target event is detected, wherein the target event features include common event features and event features corresponding to the target type;

[0035] A first calculation module, configured to extract business support data based on the target event, and perform calculations using the target event features and the business support data to obtain a covariance matrix;

[0036] The analysis module is used to perform principal component analysis based on the covariance matrix of the target event to obtain the principal component analysis result, and to perform evaluation based on the principal component analysis result to obtain the emergency response strategy for the target event.

[0037] In one embodiment, the first computing module is specifically used to:

[0038] Original data is acquired based on the target event, and the original data is preprocessed to generate business support data.

[0039] In one embodiment, the first computing module is specifically used to:

[0040] The similarity between the target event and each sample event included in the business support database is calculated, and the business support data corresponding to the similar events whose similarity meets the preset similarity condition is used as the business support data of the target event.

[0041] In one embodiment, the first calculation module is specifically used to: determine the characteristic value corresponding to each target event feature in the business support database, and generate a sample matrix based on each target event feature and the characteristic value of each target event feature;

[0042] The sample matrix is ​​standardized to obtain the column mean and standard deviation corresponding to the sample matrix, and a covariance matrix is ​​obtained based on the column mean and standard deviation corresponding to the sample matrix.

[0043] In one embodiment, the analysis module is specifically used to: calculate the correlation coefficient and the sphericity test value based on the covariance matrix of the target event;

[0044] In the case that the correlation coefficient satisfies the correlation condition and the sphericity test value satisfies the significance condition, a variance explanation table is calculated through the covariance matrix, and the number of each principal component is determined through the variance explanation table;

[0045] The latent variables of each principal component are analyzed through the variance explanation table to obtain a factor loading coefficient table, and the importance corresponding to each target principal component is determined based on the factor loading coefficient table.

[0046] In one embodiment, the analysis module is specifically used to:

[0047] An evaluation is performed based on the principal component analysis result and the requirements of the target event to obtain factors to be improved for the target event, and each factor to be improved is used as an emergency strategy for the target event.

[0048] In one embodiment, the device further comprises:

[0049] A judgment module, used to judge the effectiveness of the emergency strategy based on the target event and the corresponding target event;

[0050] A first adding module is used to add the target event and the emergency strategy corresponding to the target event to the business support database when it is determined that the target event and the emergency strategy corresponding to the target event pass the validity check;

[0051] The second adding module is used to determine that the target event and the emergency strategy corresponding to the target event are added to the garbage sample data set when it is determined that the target event and the emergency strategy corresponding to the target event have not passed the validity check.

[0052] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0053] If a target event is detected, target event features are determined based on the target type of the target event, where the target event features include common event features and event features corresponding to the target type;

[0054] Extracting business support data based on the target event, and calculating through the target event features and the business support data to obtain a covariance matrix;

[0055] A principal component analysis is performed based on the covariance matrix of the target event to obtain a principal component analysis result, and an evaluation is performed based on the principal component analysis result to obtain an emergency response strategy for the target event.

[0056] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0057] If a target event is detected, target event features are determined based on the target type of the target event, where the target event features include common event features and event features corresponding to the target type;

[0058] Extracting business support data based on the target event, and calculating through the target event features and the business support data to obtain a covariance matrix;

[0059] A principal component analysis is performed based on the covariance matrix of the target event to obtain a principal component analysis result, and an evaluation is performed based on the principal component analysis result to obtain an emergency response strategy for the target event.

[0060] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0061] If a target event is detected, target event features are determined based on the target type of the target event, where the target event features include common event features and event features corresponding to the target type;

[0062] Extracting business support data based on the target event, and calculating through the target event features and the business support data to obtain a covariance matrix;

[0063] A principal component analysis is performed based on the covariance matrix of the target event to obtain a principal component analysis result, and an evaluation is performed based on the principal component analysis result to obtain an emergency response strategy for the target event.

[0064] The above-mentioned method, device, computer equipment, storage medium and computer program product for determining the emergency strategy, wherein the method includes: if a target event is detected, the target event characteristics are determined based on the target type of the target event, and the target event characteristics include common event characteristics and event characteristics corresponding to the target type; business support data is extracted based on the target event, and the covariance matrix is ​​obtained by calculation through the target event characteristics and the business support data; principal component analysis is performed based on the covariance matrix of the target event to obtain the principal component analysis results, and the emergency strategy of the target event is obtained by evaluation based on the principal component analysis results. By adopting this method, the correlation influence between various indicators can be eliminated, the data can be processed through unified standards, the data quality in the database can be guaranteed, and the accuracy of the determined emergency strategy can be improved on the basis of reliable data. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 A flowchart of a method for determining an emergency strategy in an embodiment;

[0066] Figure 2 A schematic diagram of a flow chart of a step of calculating a covariance matrix in one embodiment;

[0067] Figure 3 A schematic diagram of a process for determining the importance of each principal component in an embodiment;

[0068] Figure 4A schematic diagram of a flow chart of a step of updating a database in one embodiment;

[0069] Figure 5 A schematic diagram of the structure of a business model of an emergency strategy in an embodiment;

[0070] Figure 6 A schematic diagram of the structure of a business model of an emergency strategy in another embodiment;

[0071] Figure 7 is a schematic diagram of the structure of a data model in an embodiment;

[0072] Figure 8 A schematic diagram of a flow chart of a database updating step in one embodiment;

[0073] Fig. 9 A curve diagram of an output result in one embodiment;

[0074] Fig.10 is a structural block diagram of a device for determining an emergency strategy in one embodiment;

[0075] Fig.11 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0076] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0077] In one embodiment, Figure 1 As shown, a method for determining an emergency strategy is provided. This embodiment uses the method applied to a terminal as an example for illustration. It is understandable that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. The above-mentioned terminal can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, etc. The server can be implemented as an independent server or a server cluster composed of multiple servers. In this embodiment, the method for determining an emergency strategy includes the following steps:

[0078] Step 102: If the occurrence of a target event is detected, the target event characteristics are determined based on the target type of the target event.

[0079] Among them, the target event features include public event features and event features corresponding to the target type. The target event can be an emergency, such as a public health emergency or a food and drug safety emergency. The public health emergency can be an event closely related to public health. The types of target events can include events of unexplained diseases, major mistakes and occupational poisoning, major transmission events, and other events that seriously affect public health, etc. The target event features are multiple events included in the target event, and the public event features can be event features corresponding to multiple types. In an example, for example, event features can be extracted based on "three dimensions and four levels". Each event is combined with event business and characteristics to extract time features. The three dimensions include the degree of harm, the scope of involvement, and the nature of the event. The four levels can include the classification of food and drug safety emergencies, which are particularly serious at the first level, serious at the second level, larger at the third level, and general at the fourth level.

[0080] Specifically, if the terminal detects that a target event is currently occurring, the terminal can determine the target type of the target event, and based on the correspondence between the preset type and the event characteristics, determine multiple event characteristics corresponding to the target type of the target event, as well as determine the common event characteristics corresponding to each event type, and determine the multiple event characteristics corresponding to the target type and the common event characteristics as the target event characteristics corresponding to the target event.

[0081] In an example, the terminal may perform standardization processing on the extracted target event features respectively to obtain each target event feature after the standardization processing.

[0082] In another example, the terminal can classify the types of emergencies to build a business model for food and drug safety emergencies. In the model, the terminal first starts with the types of public health emergencies, which are divided into unexplained diseases, major errors and occupational poisoning, major transmissions and other events that seriously affect public health; then the three dimensions of hazard level, scope of involvement and nature of the event are combined with the four levels of events to sort out the four types of events, and the event features of the four types of events are extracted from the "three dimensions and four levels" respectively. For each event, time features are extracted based on the event business and characteristics, and feature standardization is completed. The same type of features are unified in name and business meaning; at the same time, the features common to multiple events are extracted and maintained as common features and standardized.

[0083] Step 104: extracting business support data based on the target event, and performing calculations using the target event features and the business support data to obtain a covariance matrix.

[0084] Among them, the business support data can be data related to the business processing of the target event, the stage before the target event occurs, the stage when the target event occurs, and the stage after the target event occurs, that is, the public health emergency can include the pre-stage, the in-stage stage, and the post-stage. Each stage corresponds to different business support data. For example, the business support data corresponding to the pre-stage can include clinical sample retention information, event-frequent area information, etc.; the business support data corresponding to the in-stage stage can include event location information and event patient area information, etc.; the business support data corresponding to the post-stage can include food sampling inspection information and drug use information, etc. The covariance matrix is ​​used to determine the principal component information of the target event, etc.

[0085] Specifically, the terminal can determine the business support data corresponding to the target event based on the target event, and perform standardization and covariance processing through the target event features corresponding to the target event and the business support data corresponding to the target event. After standardization, standardized data is obtained, and a covariance matrix can be obtained based on the standardization processing.

[0086] Step 106, performing principal component analysis based on the covariance matrix of the target event to obtain a principal component analysis result, and performing evaluation based on the principal component analysis result to obtain an emergency response strategy for the target event.

[0087] Among them, principal component analysis (PCA) can determine the more important event features of the target event based on the covariance matrix of the target event. The emergency strategy of the target event is used to handle the target event in an emergency. The emergency strategy of the target event can also be the dimension to be improved corresponding to the target event.

[0088] Specifically, the terminal can perform principal component analysis based on the principal component analysis algorithm and the covariance matrix of the target event to obtain the output result of the principal component analysis algorithm, that is, the principal component analysis result; the terminal can determine the influencing factors with a high degree of correlation with the target event based on multiple results contained in the principal component analysis result, and determine the emergency strategy of the target event based on the influencing factors. In this way, the terminal can perform emergency processing on the target event based on the determined emergency strategy.

[0089] In the above-mentioned method for determining the emergency strategy, if the target event is detected, the target event characteristics are determined based on the target type of the target event, and the target event characteristics include common event characteristics and event characteristics corresponding to the target type. Business support data is extracted based on the target event, and the covariance matrix is ​​obtained by calculation through the target event characteristics and the business support data. Principal component analysis is performed based on the covariance matrix of the target event to obtain the principal component analysis results, and an evaluation is performed based on the principal component analysis results to obtain the emergency strategy for the target event. By adopting this method, the correlation influence between various indicators can be eliminated, the data can be processed through unified standards, the data quality in the database can be guaranteed, and the accuracy of the determined emergency strategy can be improved on the basis of reliable data.

[0090] In one embodiment, the specific processing of step 104 "extracting business support data based on the target event" includes:

[0091] Obtain raw data based on target events, pre-process the raw data, and generate business support data.

[0092] Specifically, the terminal can obtain the original data related to the target event, and perform statistics, analysis and sorting on the original data to obtain business support data. In one example, the original data may include data collected by the event unit, Internet data, and public open data, etc. The business support data obtained after processing may include clinical sample retention information, hospital outpatient visit information, and seasonal case information, etc. The terminal can perform business support processing based on the business support data, for example, it can be used for monitoring, identification, and prevention, etc.

[0093] In this embodiment, the business support data is determined by the original data, so that the accuracy and comprehensiveness of data collection can be guaranteed.

[0094] In one embodiment, the specific processing of step 104 "extracting business support data based on the target event" includes:

[0095] The similarity between the target event and each sample event included in the business support database is calculated, and the business support data corresponding to the similar events whose similarity meets the preset similarity condition is used as the business support data of the target event.

[0096] The similarity may be a cosine similarity calculated based on a cosine similarity algorithm, and the content of the preset similarity condition may be that the similarity is greater than or equal to a preset similarity threshold. The business support database may include multiple events, business support data corresponding to each event, etc., and may also include emergency strategies corresponding to each event, etc.

[0097] Specifically, the terminal can calculate the similarity between the target event and multiple events contained in the business support database through a similarity algorithm, and compare each similarity with a preset similarity threshold, and take events that are greater than or equal to the preset similarity threshold as similar events of the target event; the terminal can extract the business support data corresponding to the similar event in the business support database, and use the business support data as the business support data of the target event.

[0098] In one example, the preset similarity algorithm can be a cosine similarity algorithm, whose logic is to calculate the distance or similarity between certain objects, and can be used to compare the similarity between two events. Its function is to help find the closest event, thereby reducing the time for reviewing similar events or similar cases. The specific calculation process of similarity may include: preprocessing during data aggregation, cleaning, governance and extraction, and the preprocessing mainly includes Chinese word segmentation and removal of stop words. List all the words in the feature values. Calculate the word frequency and write the word frequency vector. Calculate the cosine value between the word frequency vectors corresponding to each event, and use the cosine value as the similarity between the two events.

[0099] In this embodiment, by extracting business support data from the business support database based on similarity, data extraction efficiency can be improved.

[0100] In one embodiment, Figure 2 As shown, the specific processing process of the step "calculating through target event features and business support data to obtain a covariance matrix" includes:

[0101] Step 202: Determine the feature value corresponding to each target event feature in the business support database, and generate a sample matrix based on each target event feature and the feature value of each target event feature.

[0102] Specifically, the terminal may obtain the feature values ​​corresponding to each target event feature in the business support database, and perform splicing based on each target event feature and the feature values ​​of each target event feature to obtain a multi-dimensional sample matrix.

[0103] Step 204 , standardize the sample matrix to obtain the column mean and standard deviation corresponding to the sample matrix, and obtain the covariance matrix based on the column mean and standard deviation corresponding to the sample matrix.

[0104] Specifically, the terminal can perform standardization processing on the sample matrix. For example, for each column in the sample matrix, the terminal can calculate the mean of the eigenvalues ​​of the column and the standard deviation corresponding to the column, obtain the standard value based on the column mean and standard deviation corresponding to each column, perform covariance calculation based on each standard value, and obtain the covariance matrix. In other words, the process of the terminal performing standardization processing to obtain the covariance matrix can be to first perform dimensionality reduction processing on the sample matrix, calculate the mean of the eigenvalues ​​of the column and the standard deviation corresponding to the column, obtain the standard value based on the column mean and standard deviation corresponding to each column, reduce the n-dimensional sample matrix to a one-dimensional standard value matrix, and perform covariance calculation based on each eigenvector contained in the standard value matrix to obtain the covariance matrix, and so on.

[0105] In this embodiment, by performing standardization processing on the data and calculating the covariance, data that is convenient for analysis and calculation can be obtained, providing a reliable data basis for subsequent principal component analysis.

[0106] In one embodiment, Figure 3 As shown in FIG. 1 , the specific processing process of the step “performing principal component analysis based on the covariance matrix of the target event to obtain the principal component analysis result” includes:

[0107] Step 302: Calculate the correlation coefficient and the sphericity test value based on the covariance matrix of the target event.

[0108] Specifically, the terminal may obtain an analysis result by analyzing the covariance matrix, and the analysis result may include a correlation coefficient and a sphericity test value. The correlation coefficient may be a KMO coefficient value, and the sphericity test value may be a Bartlett's test value.

[0109] Step 304: When the correlation coefficient satisfies the correlation condition and the sphericity test value satisfies the significance condition, a variance explanation table is calculated through the covariance matrix, and the number of each principal component is determined through the variance explanation table.

[0110] Specifically, the terminal can determine whether principal component analysis can be performed, that is, whether there is a principal component, through the correlation coefficient and the sphericity test value. The terminal can determine that the correlation condition is met when the correlation coefficient is greater than the preset correlation coefficient threshold, and the terminal can determine the significant value based on the sphericity test value. When the significant value is less than the significant threshold, it can be determined that the significance condition is met. When the terminal determines that the correlation condition and the significance condition are met, the terminal can perform principal component analysis, calculate the variance explanation table through the covariance matrix, and analyze through the variance explanation table to obtain the principal component and the number of principal components.

[0111] In one example, if the KMO test is passed (KMO>0.6), it means that there is a correlation between the item variables, which meets the requirements of principal component analysis. If the Bartlett test is passed: P<0.05, which is significant, principal component analysis can be performed.

[0112] Step 306: Analyze the latent variables of each principal component through the variance explanation table to obtain a factor loading coefficient table, and determine the importance corresponding to each target principal component based on the factor loading coefficient table.

[0113] Specifically, the terminal can calculate the contribution rate and weight ratio of each principal component through the variance explanation table; a scree plot can be generated based on the contribution rate and weight ratio of each principal component in the variance explanation table to display the trend and position of the principal component; based on the contribution rate and weight ratio of each principal component, an importance analysis is performed on the latent variables included in the principal component to obtain the factor load coefficient table of each principal component, and the corresponding importance of each target principal component is determined through the factor load coefficient table, that is, the weight ratio and ranking of each component. The result of the principal component analysis is the weight ratio and ranking of each component.

[0114] In this embodiment, the accurate importance of each principal component is determined by the principal component analysis algorithm, so that the influencing factors with higher importance can be accurately located, the emergency events can be effectively handled, and the effectiveness of the determined emergency strategy can be ensured.

[0115] In one embodiment, the specific processing of the step of "evaluating based on the principal component analysis results to obtain an emergency response strategy for the target event" includes:

[0116] Based on the results of principal component analysis and the needs of the target event, an evaluation is performed to obtain the factors to be improved of the target event, and each factor to be improved is used as an emergency strategy for the target event.

[0117] Specifically, the terminal can determine the degree of match between the needs of the target event and the weight proportion and ranking of each component, determine the main component of the current unmet needs as the factor to be improved, and use the factor to be improved as the emergency strategy for the target event. In one example, combined with the results of basic business data aggregation and analysis, statistics are collected on material reserves, personnel reserves, and the scale of medical facility construction, and compared with the needs of sudden safety events to analyze the dimensions to be supplemented and improved.

[0118] In this embodiment, the matching emergency strategy is determined by the influencing factors with higher importance, the effectiveness of the determined emergency strategy is guaranteed, the medical rescue capacity assessment is improved, and the medical emergency rescue capacity is improved.

[0119] In one embodiment, Figure 4As shown, the method for determining the emergency strategy also includes:

[0120] Step 402: Determine the effectiveness of the emergency response strategy based on the target event and the target event.

[0121] Specifically, the terminal can perform validity verification on the target event and the emergency strategy corresponding to the target event, for example, the credibility and rationality of the emergency strategy can be determined. The terminal can judge the credibility and rationality of the emergency strategy based on the demand data of the target event. When it is determined that the emergency strategy matches the demand of the target event, the terminal can determine that the target event and the emergency strategy corresponding to the target event pass the validity verification; when it is determined that the emergency strategy does not match the target event, the terminal can determine that the target event and the emergency strategy corresponding to the target event do not pass the validity verification.

[0122] In an example, the data generated by the target event (emergency strategy), i.e., the data set, is combined with the business attributes of the event through a model to calculate the result. If the result has high credibility and rationality and the trend is consistent with the actual situation, it can be regarded as the data generated by the event, i.e., the data set, which can be used to improve the capability assessment model, that is, it can be fed back and is valid data, otherwise it is invalid data.

[0123] Step 404: When it is determined that the target event and the emergency response strategy corresponding to the target event pass the validity check, the target event and the emergency response strategy corresponding to the target event are added to the business support database.

[0124] Specifically, when the terminal determines that the target event and the emergency strategy corresponding to the target event have passed the validity verification, that is, the demand data of the target event matches the emergency strategy corresponding to the target event, the terminal can update the business support database based on the target event and the emergency strategy corresponding to the target event. For example, the target event and the emergency strategy corresponding to the target event can be added to the business support database to realize data feedback to the database.

[0125] Step 406 : When it is determined that the target event and the emergency strategy corresponding to the target event fail to pass the validity check, it is determined that the target event and the emergency strategy corresponding to the target event are added to the garbage sample data set.

[0126] Specifically, when the terminal determines that the target event and the emergency strategy corresponding to the target event have not passed the validity check, that is, the demand data of the target event does not match the emergency strategy corresponding to the target event, the terminal can update the garbage sample database based on the target event and the emergency strategy corresponding to the target event. For example, the target event and the emergency strategy corresponding to the target event can be added to the garbage sample database to realize data feedback to the database.

[0127] In this embodiment, through the target event and the emergency strategy corresponding to the target event, it is possible to achieve the effect of accumulating past experience and distinguishing good from bad sample events, recommending good sample events to continue to strengthen, and focusing on avoiding bad sample events.

[0128] The specific implementation process of the above-mentioned method for determining the emergency strategy is described in detail below in conjunction with a specific embodiment:

[0129] The existing technology has problems such as the inability to eliminate the correlation between various indicators. The various indicators in the existing technology are highly correlated. Without classification and grading, it is impossible to achieve a one-to-many correspondence between events and characteristic indicators, resulting in low accuracy of the analysis results. If the correlation between emergencies and characteristic indicators can be found, the specific indicators can be used to prevent and solve food and drug safety problems from the source. The standardization or discretization standards are not unified, which will lead to large deviations in the results and reduced credibility. How to ensure the amount of data and data quality in the standard emergency plan database is also an issue that needs to be verified. It only involves the method of judging its level based on the score threshold. The analysis results are relatively simple and cannot reflect the proportion of the original business data contained in the analysis content to the total data volume. Therefore, the objectivity and rationality of the results need to be studied.

[0130] The embodiment of the present application provides a model for evaluating the medical emergency treatment capacity of sudden food and drug safety incidents based on PCA, such as Figure 5As shown, it includes the business model of sudden food and drug safety incidents (constructed according to the type of sudden incidents), the data model of sudden food and drug safety incidents (constructed according to the stage of sudden incidents), the principal component analysis method, and the model for evaluating the medical emergency treatment capacity of sudden food and drug safety incidents, involving the pre-event, in-event, and post-event stages. The output results of the evaluation model of sudden food and drug safety incidents can also be judged as sample events, and the effective sample time can be used to feed back the business support data (sample library) of each stage, and the useless sample time is judged as a garbage sample event. Among them, the business model of sudden food and drug safety incidents includes the classification of public health emergencies and event characteristics. The types include unexplained diseases, major errors and occupational poisoning, major transmission and other events that seriously affect public health. Event characteristics include diversity of causes, distribution differences, widespread transmission, comprehensive governance, frequency of events, diversity of types, complexity of hazards, and other characteristics of events; based on the classification extraction of emergencies and the extraction of event characteristics, the event feature ranges of various types of public health emergencies are standardized, and the public feature ranges of sudden food and drug safety incidents are standardized, that is, the event feature ranges of unexplained diseases, major errors and occupational poisoning, major transmission and other events that seriously affect public health are standardized; based on sudden food and drug safety incidents, the data set situations in the pre-, mid- and post-event business models are statistically analyzed, and the data set variables are standardized.

[0131] The data model for sudden food and drug safety incidents includes three stages: before, during and after the incident. The before-event stage includes hospital outpatient visit information, seasonal case information, cluster case information, basic information of medical staff, medical supplies reserve information, and other dimensional information, which can be used for monitoring, identification and prevention; the during-event stage includes: event classification information, food provider information, drug provider information, medical supplies reserve information, medical staff reserve information and other dimensional information, which can be used for allocation, reporting and treatment; the after-event stage includes: daily food safety information, food sampling and inspection information, drug use information, other dimensional information, remaining material reserve information and material to be replenished information, which are used for supervision and protection.

[0132] The data output by the business model of sudden food and drug safety incidents is quantitative, and the data output by the data model of sudden food and drug safety incidents is variable. The principal component analysis method is performed based on the above, wherein the specific analysis process includes: standardization, covariance matrix calculation, principal component analysis, calculation of eigenvectors and eigenvalues, formation of eigenvectors, and redrawing of data; the principal component analysis results are input into the model for evaluating the medical emergency treatment capacity of sudden food and drug safety incidents, and the evaluation is performed in the model. The evaluation model may include models corresponding to multiple stages, for example, it may include a pre-medical emergency rescue capacity evaluation model for sudden food and drug safety incidents, a medical emergency rescue capacity evaluation model during sudden food and drug safety incidents, and a post-medical emergency rescue capacity evaluation model for sudden food and drug safety incidents.

[0133] like Figure 6 As shown in the figure, the business model is constructed according to the classification of emergency types. The model is constructed based on the classification of emergency types. First, we start with the types of public health emergencies, which are divided into unexplained diseases, major errors and occupational poisoning, major transmission and other events that seriously affect public health; then we sort out the three dimensions of hazard degree, scope and nature of the event in combination with the four levels of events (especially serious, serious, large and general), and extract event features from the "three dimensions and four levels" of the four types of events. For each event, event features are extracted and standardized in combination with the event business and characteristics, and the same type of features are unified in name and business meaning; at the same time, the common features of multiple events are extracted and maintained as public features and standardized. Standardization of the scope of event characteristics includes standardization of each category. For example, unexplained diseases include diversity of causes, comprehensiveness of governance, complexity of hazards, differences in distribution, widespread dissemination, and other characteristics of events; major errors and occupational poisoning include diversity of types, frequency of events, differences in distribution, comprehensiveness of governance, diversity of causes, and other characteristics of events; major transmissions include: diversity of causes, frequency of events, complexity of hazards, differences in distribution, comprehensiveness of governance, and other characteristics of events; other events that seriously affect public health include: diversity of causes, comprehensiveness of governance, complexity of hazards, diversity of types, widespread dissemination, and other characteristics of events.

[0134] The data model of food and drug safety incidents is constructed according to the stages of the incidents. The logical architecture of the model is shown in the figure below. Figure 7As shown in the figure: This model disassembles the whole process of sudden food and drug safety incidents into three links: before, during and after the incident; at the same time, from the perspective of data, it gathers data from multiple sources such as government affairs, the Internet, and public openness, and performs data cleaning, data governance and data extraction on these data to form business support data with high utilization value and high data quality, and is used for business activity support and decision-making assistance in the three links. Specifically, it includes but is not limited to the original data dimensions shown in the figure, the test data generated by the food and drug safety testing system; such as other chemical pollutants, foodborne pathogenic bacteria contamination, pesticide residues and veterinary drug residues, non-edible substances and food additives, as well as rapid detection of risk substances (such as gutter oil triglycerides), or certain allergens can be used as feature points that can be extracted within the feature range; it also includes relevant information of upstream and downstream practitioners, not limited to the above-mentioned medical staff information. In the pre-event, during-event and post-event stages, due to the different effective feature values ​​that new sample events can provide, their similarities also vary. Newly acquired events are matched for similarity based on the accumulated and optimized business support data and sample events to meet different levels of prevention, monitoring, treatment, resource allocation, supervision and protection scenarios.

[0135] exist Figure 7In the pre-event stage of sudden food and drug safety incidents, the original data include data collected by the unit where the incident occurred, data collected in other dimensions, Internet data, public open data, data from commissions and bureaus, and hospital data; the original data is cleaned, extracted, and governed to obtain business support data, which includes: clinical sample retention information, hospital outpatient treatment information, seasonal case information, information on areas where incidents frequently occur, registered onset time information, basic information on medical staff, basic information on susceptible populations, information on time periods when incidents frequently occur, medical supplies reserve information, cluster case information, and other dimensional information, which can be monitored, identified, and prevented based on business support data; similarly, in the mid-event stage of sudden food and drug safety incidents, the original data include data collected by the unit where the incident occurred, data collected in other dimensions, Internet data, public open data, data from commissions and bureaus, and hospital data; the original data is cleaned, extracted, and governed to obtain business support data The business support data includes: information on drug providers, medical supplies reserves, medical personnel reserves, other dimensional information, food providers, event classification, event location, location of the patient, institution receiving the patient, production and operation information of the institution, disposal plan information and emergency response time statistics. Reporting, treatment and deployment are carried out based on the above business support data. Similarly, in the post-event stage of sudden food and drug safety incidents, the original data includes data collected by the incident unit, data collected in other dimensions, Internet data, public open data, commission and bureau data, and hospital data. The original data is cleaned, extracted and governed to obtain business support data. The business support data includes: food sampling and inspection information, drug use information, information on materials to be replenished, daily food safety information, remaining material reserves information and other dimensional information. Supervision and protection are carried out based on the above business support data.

[0136] Specifically, the algorithm used in the matching process is the cosine similarity algorithm, whose logic is to calculate the distance or similarity between objects, and can be used to compare the similarity between two events. Its function is to help find the closest event, thereby reducing the time for reviewing similar events or similar cases.

[0137] The idea of ​​constructing cosine similarity based on sample event similarity matching is: if the feature vectors of two events are more similar, the content of the events is more similar, so we can start from the feature vectors and calculate their similarity; the specific steps are as follows: preprocessing is performed during data aggregation, cleaning, governance and extraction, and the preprocessing content mainly includes Chinese word segmentation and stop word removal; list all the words in the feature values; calculate the word frequency; write the word frequency vector. The closer the cosine value is to 1, the closer the angle is to 0 degrees, that is, the more similar the two events are. The calculation of the space vector co is based on the linear algebra of space vectors.

[0138] A specific example may be: the first step is to perform word segmentation using influenza as an example. The features of the accumulated influenza sample events are: this / virus / diffusibility / is / strong, and / its / diffusibility / is / covering all age groups; the features of the sample events of emergencies are: this / virus / diffusibility / is / not / weak, and / its / susceptible population / includes children, teenagers, adults and the elderly, etc.; the second step is to calculate the word frequency, that is, to write the frequency of occurrence of each word; the third step is to write the word frequency vector, the former: (1, 1, 2, 1, 1, 1, 0, 0, 0), the latter: (1, 1, 1, 0, 1, 1, 1, 1, 1); the fourth step is to calculate through the formula, and the cosine value of the angle in the calculation result is 0.81, which is very close to 1, so the above two sample events are basically similar.

[0139] Specifically, the process of standardizing the variables of the data set by the terminal may include: statistics, analysis and sorting of the business support data extracted from the original data based on the business and features, comparing the results with the full original data set, calculating the proportion of the original and business data volume to the total data volume, converting the proportion (0%-100%) into a corresponding score, and the score ranges from 0 to 100 points; the specific algorithm is as follows:

[0140] When there are n samples and p indicators, a sample matrix x of size n\times p can be constructed:

[0141]

[0142] The mean and standard deviation are calculated by column to obtain standardized data, where the mean is calculated using the following formula:

[0143]

[0144] The standard deviation is calculated using the following formula:

[0145]

[0146] Normalized data is calculated using the following formula:

[0147]

[0148] The original sample matrix is ​​standardized to obtain a standardized matrix:

[0149]

[0150] Based on the standardized matrix, we get the covariance matrix:

[0151]

[0152]

[0153] The process of principal component analysis may include: the variables used in the PCA algorithm include: {number of sales outlets that have passed drug safety testing, number of sales outlets that have passed food safety testing, number of qualified food supply units, first aid equipment reserves, antibiotic drug reserves, special drug reserves, medical personnel reserves, epidemic prevention and medical supplies reserves, number of qualified drug supply units}; index item: {disease number}; parameters include: number of principal components: {1}.

[0154] First, KMO and Bartlett's tests are performed to determine whether principal component analysis can be performed. For KMO values: 0.8 is very suitable for principal component analysis, 0.7-0.8 is generally suitable, 0.6-0.7 is not very suitable, 0.5-0.6 is poor, and 0.5 is extremely unsuitable. For Bartlett's test, if P is less than 0.05 and the null hypothesis is rejected, it means that principal component analysis can be performed. If the null hypothesis is not rejected, it means that these variables may provide some information independently and are not suitable for principal component analysis.

[0155] By analyzing the variance explanation table and scree plot, the quantitative variance of the principal components is determined; the explanation table mainly looks at the contribution rate of the principal components to the explanation of the variables (which can be understood as how many principal components are needed to express the variable as 100%). If it is too low (less than 60%), the principal component data needs to be adjusted.

[0156] The function of the scree plot is to confirm the number of principal components that need to be selected based on the gradient of the eigenvalue decrease. The combination of the two can be used to confirm or adjust the number of principal components.

[0157] By analyzing the principal component load coefficient and heat map, the importance of the hidden variables in each principal component can be analyzed. Based on the principal component load diagram, multiple principal components are reduced to two or three principal components, and the spatial distribution of the principal components is presented in the form of a quadrant diagram. If two principal components are extracted, the three-dimensional load principal component scatter plot cannot be presented. If one principal component is extracted, the principal component quadrant diagram cannot be displayed. By analyzing the component matrix, the principal component formula and weight are obtained. Output the comprehensive score of the principal component analysis method.

[0158] Specifically, a medical emergency treatment capability assessment model is constructed. The model for assessing the medical emergency treatment capability of sudden food and drug safety incidents can be divided into pre-, mid- and post-models at a fine-grained level. The basis of each model is the business model constructed by classification and stage respectively - the business model of sudden food and drug safety incidents. At the same time, they can be combined as needed. For example, the early and mid-term models are combined into a mid- and early-term capability assessment model, which can provide usage scenarios with wide dimensions, high complexity and wide coverage.

[0159] like Figure 8As shown, it is a logical architecture diagram of conditional feedback of sample events and data sets, which judges the events and data sets generated and accumulated by the model for evaluating the medical emergency treatment capabilities of sudden food and drug safety incidents. Those judged to be valid events can be used to feedback and add business support data, and the measures used to deal with such safety incidents in the past can be recommended as methods for handling new sample events. This part of the application needs to be implemented based on sample event similarity calculation, to achieve the accumulation of previous experience and the identification of good and bad sample events, and recommend that good sample events continue to be strengthened and bad sample events be avoided.

[0160] Specifically, KMO test and Bartlett's test, variance explanation table, scree plot, factor loading coefficient table, factor loading matrix heat map, component matrix table and factor weight analysis can be output on demand to provide support for emergency response process flow and decision analysis.

[0161] In an example, the output result 1 may be as shown in Table 1 below:

[0162] Table 1

[0163]

[0164] Among them, the above table shows the results of the KMO test and the Bartlett sphericity test, which are used to analyze whether principal component analysis can be performed. If the KMO test is passed (KMO>0.6), it means that there is a correlation between the item variables, which meets the requirements of principal component analysis. If the Bartlett test is passed: P<0.05, it is significant, then principal component analysis can be performed. In an example, the results of the KMO test in Table 1 above show that the value of KMO is nan. At the same time, the results of the Bartlett sphericity test show that the significant P value is NaN, which is not significant at the level. If the original hypothesis is accepted, that is, the variables are independent of each other, then the common factor cannot be extracted from them, and the principal component analysis is invalid. It is recommended to adjust the data quality, and the degree is extremely unsuitable.

[0165] The output result 2 can be shown in the following Table 2:

[0166] Table 2

[0167] Element Characteristic root Variance explained (%) Cumulative variance explained (%) 1 5.189 57.651 57.651 2 2.308 25.646 83.297 3 1.134 12.6 95.897 4 0.369 4.103 100 5 100 6 100 7 100 8 100 9 100

[0168] Among them, Table 2 above is the total variance explanation table, which mainly looks at the contribution rate of the principal component to the variable explanation (it can be understood as how many principal components are needed to express the variable as 100%). Generally, it must be expressed to be above 90%, otherwise the factor data must be adjusted. In general, the higher the variance explanation rate, the more important the principal component is, and the higher the weight should be. In the variance explanation table, when the principal component is 4, the characteristic root of the total variance explanation is less than 1, and the contribution rate of the variable explanation reaches 100.0. The above is only for reference. If the characteristic root is less than 1 and the critical value is too large, you can also analyze the specific situation.

[0169] Output 3 can be as follows Fig. 9 As shown in the figure, the scree plot is a plot based on the degree of explanation of each principal component on the data variation. Its function is to confirm the number of principal components to be selected based on the gradient of the eigenvalue decrease. It can be used to confirm or adjust the number of principal components in combination with the variance explanation table. Each principal component is a point, and the number of principal components to be extracted is determined by the trend and position of the "slope tends to be flat" in the figure.

[0170] The output result 4 can be shown in the following Table 3:

[0171] Table 3

[0172]

[0173] Among them, the above table is the factor loading coefficient table, which can analyze the importance of latent variables in each principal component.

[0174] Assuming that n factors are determined in the previous article, the factor loading coefficients of a, b, c, and d in factor i are relatively large, so the principal component i can be summarized and renamed. A factor loading matrix heat map can also be generated, that is, output result 5, which can analyze the importance of the hidden variables in each principal component. At the same time, the hidden variable analysis of each factor can be carried out in combination with specific business.

[0175] The output result 6 can be shown in the following Table 4:

[0176] Table 4

[0177]

[0178] Among them, Table 4 is a component matrix table, which is intended to illustrate the factor score coefficients (principal component loadings) contained in each component, which are used to calculate the component scores and obtain the factor formula. The calculation formula is: linear combination coefficient * (variance explanation rate / cumulative variance explanation rate), and finally normalized to obtain the factor weight score. The linear combination coefficient formula is: factor load coefficient divided by the corresponding characteristic root, that is, the coefficient of the component matrix.

[0179] Model formula: F1 = -0.17952195516586852 × number of sales outlets with qualified drug safety testing + 0.15803662925548737 × number of sales outlets with qualified food safety testing + 0.1796596192902919 × number of qualified food supply units + 0.09093515080040893 × emergency equipment reserves + 0.02934769 0517013334×antibiotic drug reserves+0.04682021557442066×special drug reserves+0.18714258188281263×medical personnel reserves+0.17241759366970372×epidemic prevention medical supplies reserves+0.16485348059843946×qualified number of drug providers. From the above, we can get: F=F1.

[0180] The output result 7 may be shown in the following Table 5:

[0181] Table 5

[0182] name Variance explained (%) Cumulative variance explained (%) Weight (%) Principal component 1 5.189 57.651 100％

[0183] Among them, Table 5 above is the principal component weight analysis based on the load coefficient and other information. The calculation formula is: variance explanation rate / cumulative variance explanation rate after rotation. The weight calculation results of the principal component analysis show that the weight of principal component 1 is 100%,

[0184] The output result 8 may be shown in the following Table 6:

[0185] Table 6

[0186] Ranking Row Index Comprehensive score Principal component 1 1 2 1.365 1.365 2 4 0.294 0.294 3 5 -0.013 -0.013 4 3 -0.235 -0.235 5 1 -1.412 -1.412

[0187] Among them, Table 6 above is a preview result, which only shows the first 5 data in the comprehensive ranking. The comprehensive score is calculated according to the F value and sorted in descending order to obtain the comprehensive score and ranking of each sample, and the result after factor concentration is output at the same time.

[0188] Based on this, the medical treatment capacity assessment report can be generated and analyzed from the following aspects: combined with the results of basic business data aggregation and analysis, statistics on material reserves, personnel reserves and medical facility construction scale, and compared with the needs of sudden safety events, analyze the dimensions to be supplemented and improved, improve the medical treatment capacity assessment, and improve the medical emergency treatment capacity. In an example, it can be first, strengthen the training of grassroots medical personnel, improve the quality of rescue, and pay special attention to the accumulation of diagnosis of event symptoms, including home treatment, rational use of drugs, and treatment with traditional Chinese medicine, especially the identification of early characteristics of events, and strengthen learning, accumulation and mastery, and apply what you have learned. Second, we must give play to regional and spatial advantages. Comprehensively and dynamically grasp the health of residents, achieve "early detection and early treatment" for key groups, and realize non-drug treatment, such as oxygen therapy, nutritional support, fluid infusion and other treatments and guidance. We know that there is no specific antiviral drug, but early detection and use of some antiviral drugs are useful. Therefore, we must strengthen the rational indication of drug use and early identification of events, and use small molecule antiviral drugs in the early stage.

[0189] A method for determining an emergency strategy provided in an embodiment of the present application is a process of constructing a capability assessment model based on constructing a business model for sudden food and drug safety incidents - standardizing data set variables - principal component analysis - constructing a medical emergency treatment capability assessment model - results and analysis, which specifically includes a three-dimensional four-level event feature extraction method and logic; a model construction idea for evaluating the medical emergency treatment capability before, during and after sudden food and drug safety incidents; the application of similarity calculation of the feature vector of the new sample event eigenvalue matching sample event; data set variable standardization processing technology and methods, such as the application idea of ​​constructing a sample matrix and covariance matrix principal component analysis (PCA) in the model of medical emergency treatment capability assessment for sudden food and drug safety incidents; the idea and method of splitting the business links of sudden food and drug safety incidents according to the event type and event development stage; and the rules and logic for the sample events and data sets generated by the model of medical emergency treatment capability assessment for sudden food and drug safety incidents to feed back the business support data.

[0190] It should be understood that, although the steps in the flowcharts involved in the above embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0191] Based on the same inventive concept, the embodiment of the present application also provides an emergency strategy determination device for implementing the emergency strategy determination method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the embodiments of one or more emergency strategy determination devices provided below can refer to the limitations of the emergency strategy determination method above, and will not be repeated here.

[0192] In one embodiment, Fig.10 As shown, a device 1000 for determining an emergency strategy is provided, comprising:

[0193] The detection module 1001 is used to determine the target event characteristics based on the target type of the target event if the target event is detected, and the target event characteristics include common event characteristics and event characteristics corresponding to the target type;

[0194] A first calculation module 1002 is used to extract business support data based on the target event, and perform calculations using the target event features and the business support data to obtain a covariance matrix;

[0195] The analysis module 1003 is used to perform principal component analysis based on the covariance matrix of the target event to obtain the principal component analysis result, and to perform evaluation based on the principal component analysis result to obtain the emergency response strategy for the target event.

[0196] In one embodiment, the first computing module is specifically configured to:

[0197] Obtain raw data based on target events, pre-process the raw data, and generate business support data.

[0198] In one embodiment, the first computing module is specifically configured to:

[0199] The similarity between the target event and each sample event included in the business support database is calculated, and the business support data corresponding to the similar events whose similarity meets the preset similarity condition is used as the business support data of the target event.

[0200] In one embodiment, the first calculation module is specifically used to: determine the characteristic value corresponding to each target event feature in the business support database, and generate a sample matrix based on each target event feature and the characteristic value of each target event feature;

[0201] The sample matrix is ​​standardized to obtain the column mean and standard deviation corresponding to the sample matrix, and the covariance matrix is ​​obtained based on the column mean and standard deviation corresponding to the sample matrix.

[0202] In one embodiment, the analysis module is specifically used to: calculate the correlation coefficient and the sphericity test value based on the covariance matrix of the target event;

[0203] When the correlation coefficient satisfies the correlation condition and the sphericity test value satisfies the significance condition, the variance explanation table is calculated through the covariance matrix, and the number of each principal component is determined through the variance explanation table;

[0204] The latent variables of each principal component are analyzed through the variance explanation table to obtain the factor loading coefficient table, and the importance of each target principal component is determined based on the factor loading coefficient table.

[0205] In one embodiment, the analysis module is specifically used to:

[0206] Based on the results of principal component analysis and the needs of the target event, an evaluation is performed to obtain the factors to be improved of the target event, and each factor to be improved is used as an emergency strategy for the target event.

[0207] In one embodiment, the device further comprises:

[0208] A judgment module is used to judge the effectiveness of the emergency strategy based on the target event and the corresponding target event;

[0209] A first adding module is used to add the target event and the emergency strategy corresponding to the target event to the business support database when it is determined that the target event and the emergency strategy corresponding to the target event pass the validity check;

[0210] The second adding module is used to determine that the target event and the emergency strategy corresponding to the target event are added to the garbage sample data set when it is determined that the target event and the emergency strategy corresponding to the target event have not passed the validity check.

[0211] Each module in the above-mentioned emergency strategy determination device can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each of the above modules.

[0212] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Fig.11 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store event data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for determining an emergency strategy is implemented.

[0213] Those skilled in the art will understand that Fig.11 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0214] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0215] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0216] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0217] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0218] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0219] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0220] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for determining an emergency strategy, characterized in that: The method comprises: If a target event is detected, target event features are determined based on the target type of the target event. The target event features include public event features and event features corresponding to the target type. The target type includes events of unexplained diseases, major mistakes and occupational poisoning, major transmission and other events that seriously affect public health. The target event features are obtained by extracting event features in combination with event services and characteristics. Extracting business support data based on the target event, and calculating through the target event features and the business support data to obtain a covariance matrix, wherein the business support data is data related to business processing of the target event; Performing principal component analysis based on the covariance matrix of the target event to obtain a principal component analysis result, and performing evaluation based on the principal component analysis result to obtain an emergency response strategy for the target event; The evaluating based on the principal component analysis result to obtain the emergency strategy for the target event includes: evaluating based on the principal component analysis result and the needs of the target event to obtain the factors to be improved of the target event, and using each of the factors to be improved as the emergency strategy for the target event.

2. The method according to claim 1, characterized in that The extracting business support data based on the target event includes: Original data is acquired based on the target event, and the original data is preprocessed to generate business support data.

3. The method according to claim 1, characterized in that The extracting business support data based on the target event includes: The similarity between the target event and each sample event included in the business support database is calculated, and the business support data corresponding to the similar events whose similarity meets the preset similarity condition is used as the business support data of the target event.

4. The method according to claim 1, characterized in that: The calculation is performed through the target event features and the business support data to obtain a covariance matrix, including: In the business support database, determining the characteristic value corresponding to each target event feature, and generating a sample matrix based on each target event feature and the characteristic value of each target event feature; The sample matrix is ​​standardized to obtain the column mean and standard deviation corresponding to the sample matrix, and the covariance matrix is ​​obtained based on the column mean and standard deviation corresponding to the sample matrix.

5. The method according to claim 1, characterized in that: The principal component analysis is performed based on the covariance matrix of the target event to obtain the principal component analysis result, including: Calculating correlation coefficients and sphericity test values ​​based on the covariance matrix of the target event; In the case that the correlation coefficient satisfies the correlation condition and the sphericity test value satisfies the significance condition, a variance explanation table is calculated through the covariance matrix, and the number of each principal component is determined through the variance explanation table; The latent variables of each principal component are analyzed through the variance explanation table to obtain a factor loading coefficient table, and the importance corresponding to each target principal component is determined based on the factor loading coefficient table.

6. The method according to claim 4, characterized in that The method further comprises: Determine the effectiveness of the emergency response strategy based on the target event and the target event; When it is determined that the target event and the emergency strategy corresponding to the target event pass the validity check, the target event and the emergency strategy corresponding to the target event are added to the business support database; When it is determined that the target event and the emergency strategy corresponding to the target event fail to pass the validity check, it is determined that the target event and the emergency strategy corresponding to the target event are added to the garbage sample data set.

7. A device for determining an emergency strategy, characterized in that: The device comprises: A detection module, for determining target event features based on the target type of the target event if a target event is detected, wherein the target event features include public event features and event features corresponding to the target type, wherein the target type includes events of unexplained diseases, major errors and occupational poisoning, major transmission and other events that seriously affect public health, and the target event features are obtained by extracting event features in combination with event services and characteristics; A first calculation module is used to extract business support data based on the target event, and calculate through the target event characteristics and the business support data to obtain a covariance matrix, wherein the business support data is data related to business processing of the target event; An analysis module, used for performing principal component analysis based on the covariance matrix of the target event to obtain a principal component analysis result, and performing evaluation based on the principal component analysis result to obtain an emergency response strategy for the target event; The analysis module is specifically used to evaluate based on the principal component analysis results and the needs of the target event, obtain the factors to be improved of the target event, and use each of the factors to be improved as an emergency strategy for the target event.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images