Emergency plan recommendation system and method for emergencies

Through the emergency plan recommendation system for emergencies, the event matrix matching analysis is used to generate emergency plans, which solves the problem that hospitals find it difficult to provide solutions in a timely manner when facing emergencies, and achieves efficient response to emergencies.

CN117497155BActive Publication Date: 2024-11-19PEKING UNIVERSITY THIRD HOSPITAL (THE THIRD CLINICAL MEDICAL SCHOOL OF PEKING UNIVERSITY)
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202311570978.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-11-23
Publication Date
2024-11-19
Estimated Expiration
2043-11-23

AI Technical Summary

Technical Problem

The hospital faces various emergencies during its daily operations, such as natural disasters and medical disputes, and it is difficult to provide reasonable solutions in a timely manner.

Method used

A emergency response plan recommendation system for emergencies was designed, including a detailed collection module, a detailed analysis module, a matching analysis module and a solution recommendation module. By collecting event details, an event matrix is ​​built, and the event type and level in the preset database are combined. Match the severity and generate a list of emergency plans.

Benefits of technology

Through detailed event matrix matching analysis, reasonable emergency plans can be provided to the hospital in a timely manner, improving the efficiency and accuracy of responding to emergencies.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117497155B_ABST
    Figure CN117497155B_ABST
Patent Text Reader

Abstract

The present invention provides an emergency plan recommendation system and method for emergencies, belonging to the field of intelligent recommendation technology, and the system includes: a detail collection module for acquiring and collecting emergency details; a detail analysis module for analyzing the emergency details and constructing an event matrix; a matching analysis module for matching the event matrix with a preset event type, a preset emergency level, and a preset severity in a preset database; and a plan recommendation module for generating an emergency plan list matching the emergency details according to the matching results and outputting and displaying it. By collecting emergency details and detail analysis, and subsequently matching the type, level, and verification degree, an effective emergency plan is obtained, so that a reasonable solution can be provided to the hospital in a convenient and timely manner.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of intelligent recommendation technology, and in particular to a system and method for recommending emergency plans for emergencies. Background Art

[0002] With the rapid development of science and technology, while humans are eager for social stability and enjoy modern civilization, they are increasingly threatened by disasters and accidents. With the support of modern information technology, an emergency management system with Chinese socialist characteristics has been formed. Because of the transfer and change of social contradictions, the people have put forward higher requirements for emergency management. With the rapid development of urbanization, hidden risk factors are concentratedly exposed, and the risks and challenges borne by emergency management and medical rescue activities are greater than in any previous era. In the process of daily operation, hospitals will encounter various emergencies, such as natural disasters, emergency patients and medical disputes. How to provide hospitals with reasonable solutions in a timely manner based on the details of the emergency has become an urgent problem to be solved.

[0003] Therefore, the present invention provides a system and method for recommending emergency plans for emergencies. Summary of the invention

[0004] The present invention provides an emergency plan recommendation system and method for emergencies, which is used to obtain effective emergency plans by collecting emergency details and detailed analysis, and subsequently matching types, levels, and verification degrees, so as to conveniently and timely provide hospitals with reasonable solutions.

[0005] The present invention provides an emergency plan recommendation system for emergencies, comprising:

[0006] A details collection module is used to obtain and collect details of emergencies;

[0007] A details analysis module, used for analyzing the details of the emergency event to construct an event matrix, wherein the event matrix contains emergency information of different combinations, and the emergency information under each combination is related to the emergency type, the emergency level under the emergency type, and the emergency severity under the emergency type, and each row is a burst vector constructed by the emergency information under one combination, and the column is the number of combinations;

[0008] A matching analysis module, used for matching the event matrix with preset event types, preset emergency levels and preset severity levels in a preset database;

[0009] The solution recommendation module is used to generate a list of emergency solutions that match the details of the emergency event based on the matching results and output them for display.

[0010] Preferably, the details collection module includes:

[0011] A collection unit, used to collect the first event descriptions of the emergency by a number of surrounding persons, and at the same time, collect the second event descriptions of the person who caused the emergency;

[0012] The details determination unit is used to analyze the first event description and the second event description, and combine them with the case information of the person himself to obtain the details of the emergency event.

[0013] Preferably, the details analysis module includes:

[0014] A range determination unit, configured to extract event parameters under different emergency indicators in the emergency event details, and determine the emergency range of each emergency indicator based on the event parameters;

[0015] The matrix construction unit is used to construct an event matrix according to the burst ranges of all burst indicators.

[0016] Preferably, the range determination unit comprises:

[0017] A model analysis subunit, used for inputting the emergency event details into a detail analysis model to obtain a number of key descriptions and a description reference function of each key description;

[0018] A distribution subunit, used for inputting the key description and the description reference role into a prefabricated table in sequence to obtain a description table, and extracting and obtaining parameter distribution under each burst indicator;

[0019] A traversal subunit is used to perform distribution traversal in a historical distribution database according to the parameter distribution to obtain a distribution set matching the parameter distribution, and extract a first distribution with the greatest matching degree with the parameter distribution and a second distribution with the second greatest matching degree from the distribution set;

[0020] a threshold determination subunit, configured to determine a burst threshold corresponding to a burst indicator according to the first distribution and the second distribution, and in combination with the parameter distribution and the event parameter of the parameter distribution;

[0021] A range determination subunit, for determining a dynamic range (D01, D02) of each burst indicator according to the initial dependency of the burst indicator and the current dependency of the burst indicator under the corresponding burst threshold;

[0022]

[0023]

[0024] Wherein, ave1 represents the average dependency coefficient based on the initial dependency relationship under the corresponding burst indicator; ave2 represents the average dependency coefficient based on the current dependency relationship under the corresponding burst indicator; TY represents the burst threshold value of the corresponding burst indicator; represents the dependency coefficient of the i1th dependency factor under the initial dependency relationship; represents the dependency weight of the i1th dependency factor under the initial dependency relationship; n1 represents the total number of dependency factors under the initial dependency relationship; Represents the dependency coefficient of the j1th dependency factor under the current dependency relationship; represents the dependency weight of the j1th dependency factor under the current dependency relationship; n2 represents the total number of dependency factors under the current dependency relationship; min represents the minimum value symbol;

[0025] The curve construction subunit is used to obtain the historical threshold value of each burst indicator under the historical events consistent with all key descriptions from the historical details database, and construct a first curve for each burst indicator and a second curve for each historical event;

[0026] Performing fitting processing on all first curves to obtain a first fitting value of each first curve;

[0027] Performing curve analysis on all second curves to obtain a significance index for each second curve;

[0028] The array construction subunit is used to extract the second curves consistent with the significance index from all the first curves, and randomly lock [M1 / 2] points of each second curve to perform N1 fittings, and combine with the first fitting value to construct a fitting array, wherein M1 represents the number of historical thresholds in the first curve; [ ] represents the integer symbol; and N1 represents the number of fittings;

[0029] A combined analysis subunit is used to combine and analyze the dynamic range, the fitting array and the burst threshold to obtain a burst range (T01, T02) corresponding to the burst indicator;

[0030]

[0031]

[0032] Where d1 represents the first fitting value of the corresponding burst index; is represented as the variance of the remaining values ​​in the fitting array determined when the burst indicator is a significance indicator, except for the first fitting value; Represents the variance of all values ​​in the fitted array determined when the burst indicator is the significance indicator.

[0033] Preferably, the matrix construction unit comprises:

[0034] A first construction subunit is used to screen the maximum value of the burst range of each burst indicator to obtain a first value vector, and input it into the event occurrence model to obtain a first event vector;

[0035] A second construction subunit is used to screen the minimum value of the burst range of each burst indicator to obtain a second value vector, and input it into the event occurrence model to obtain a second event vector;

[0036] The third construction subunit is used to randomly select one value from the burst range corresponding to each burst indicator, and perform N2 random selections to obtain N2 third value vectors, and input them into the event occurrence model to obtain a third event vector;

[0037] The matrix construction subunit is used to construct a corresponding event matrix based on the first event vector, the second event vector and all third event vectors, wherein the event matrix has 3 columns and N2+2 rows.

[0038] Preferably, the matching analysis module includes:

[0039] A matching unit, used to match each row vector in the event matrix with all preset combinations involved in a preset database, and calculate a matching value;

[0040] The retention unit is used to select a preset combination with the largest matching value with the corresponding row vector from all matching values ​​for retention.

[0041] Preferably, the solution recommendation module includes:

[0042] Perform overlap analysis on all retained preset combinations, and select preset combinations with overlap numbers greater than or equal to 2;

[0043] According to the number of overlaps, the recommended solutions and solution arrangements of the corresponding preset combinations are obtained to obtain a list of emergency solutions.

[0044] Preferably, the details determining unit comprises:

[0045] A word segmentation extraction subunit is used to eliminate ambiguity between the first event description and the second event description, and to perform dependency analysis to extract characteristic word segments;

[0046] A set construction subunit is used to calculate the content similarity between different feature segmentations, and to aggregate the feature segmentations within a preset similarity range to form a number of detail feature sets;

[0047] An information mining subunit is used to mine the case information of the person himself, and fill each detail feature set to obtain a corresponding filled set;

[0048] The feature analysis subunit is used to obtain the details of the emergency event based on the set features of each filled set.

[0049] The present invention provides a method for recommending an emergency plan for an emergency, comprising:

[0050] Step 1: Obtain and collect incident details;

[0051] Step 2: Analyze the details of the emergency event and construct an event matrix, wherein the event matrix contains emergency information of different combinations, and the emergency information under each combination is related to the emergency type, the emergency level under the emergency type, and the emergency severity under the emergency type, and each row is a burst vector constructed by the emergency information under a combination, and each column is the number of combinations;

[0052] Step 3: Match the event matrix with the preset event type, preset emergency level and preset severity in the preset database;

[0053] Step 4: Generate a list of emergency plans that match the details of the emergency event based on the matching results and output them for display.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] By collecting emergency details and detailed analysis, and subsequently matching the type, level, and degree of verification, we can obtain effective emergency plans, making it convenient to provide hospitals with reasonable solutions in a timely manner.

[0056] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0057] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0059] Figure 1 A structural diagram of an emergency plan recommendation system for an emergency event in an embodiment of the present invention;

[0060] Figure 2 The present invention is a flowchart of a method for recommending an emergency plan for an emergency event in an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0062] The present invention provides an emergency plan recommendation system for emergencies, such as Figure 1 As shown, including:

[0063] A details collection module is used to obtain and collect details of emergencies;

[0064] A details analysis module, used for analyzing the details of the emergency event to construct an event matrix, wherein the event matrix contains emergency information of different combinations, and the emergency information under each combination is related to the emergency type, the emergency level under the emergency type, and the emergency severity under the emergency type, and each row is a burst vector constructed by the emergency information under one combination, and each column is the number of combinations;

[0065] A matching analysis module, used for matching the event matrix with preset event types, preset emergency levels and preset severity levels in a preset database;

[0066] The solution recommendation module is used to generate a list of emergency solutions that match the details of the emergency event based on the matching results and output them for display.

[0067] In this embodiment, the emergency event details refer to an emergency situation that occurs to a certain person, such as a sudden fall in a fire scene, the severity of the fall, the emergency level, etc.

[0068] In this embodiment, the preset database includes various preset event types, preset emergency levels and preset severity levels, and also includes emergency plans under different types, levels and severity combinations.

[0069] In this embodiment, the emergency plan is, for example, to put out the fire with a fire extinguisher, and when the personnel cannot be moved, dial 120 for emergency evacuation.

[0070] In this embodiment, the emergency plan list includes various emergency plans, and they are sorted in order according to emergency priorities.

[0071] The beneficial effect of the above technical solution is: by collecting emergency details and detailed analysis, and subsequently matching the type, level, and degree of verification, an effective emergency plan can be obtained, so as to provide the hospital with a reasonable solution in a timely manner.

[0072] The present invention provides an emergency plan recommendation system for emergencies, wherein the details collection module comprises:

[0073] A collection unit, used to collect the first event descriptions of the emergency by a number of surrounding persons, and at the same time, collect the second event descriptions of the person who caused the emergency;

[0074] The details determination unit is used to analyze the first event description and the second event description, and combine them with the case information of the person himself to obtain the details of the emergency event.

[0075] In this embodiment, the surrounding personnel refer to the personnel who are near the location of the incident when the incident occurs. The first incident description and the second incident description refer to the specific description of the incident, mainly for obtaining useful information therefrom.

[0076] In this embodiment, the case information refers to the electronic medical record of the person.

[0077] The beneficial effect of the above technical solution is: by obtaining event descriptions from two aspects and combining them with case information, it is easy to obtain the details of the emergency event, providing an effective basis for the subsequent acquisition of emergency plans.

[0078] The present invention provides an emergency plan recommendation system for emergencies, wherein the detail analysis module comprises:

[0079] A range determination unit, configured to extract event parameters under different emergency indicators in the emergency event details, and determine the emergency range of each emergency indicator based on the event parameters;

[0080] The matrix construction unit is used to construct an event matrix according to the burst ranges of all burst indicators.

[0081] For example, when a medical dispute occurs in a hospital, this method can be used to provide a specific response plan for the hospital by obtaining information such as whether the patient and his family members are armed, whether the number of patients and their family members exceeds three, the department where the dispute occurs, whether the medical staff are injured, and whether the medical staff are subject to personal restrictions. Among them, regarding whether the patient and his family members are armed and whether the number of people exceeds three, this method can obtain this information to reasonably assess the degree of harm caused by the medical dispute, so as to accurately judge the corresponding preset degree of this medical dispute under the preset emergency type; regarding the department where the medical dispute occurs, this method can provide different emergency plans for different departments to protect the hospital's valuable equipment; regarding whether the medical staff are injured and whether the medical staff are subject to personal restrictions, in this method, an emergency plan with a professional rescue team or method can be provided based on the current real situation of the medical staff.

[0082] Among them, sudden indicators refer to hazard indicators of different types, and the event parameters involved in different hazard indicators are also different.

[0083] In this embodiment, the burst range refers to a hazard range corresponding to the burst indicator. For example, the hazard range under type 1 is (a1, a2).

[0084] The beneficial effect of the above technical solution is: by obtaining the event parameters of the indicator event, the emergency scope is determined, and then an event matrix is ​​constructed to provide a basis for the subsequent determination of the emergency plan.

[0085] The present invention provides an emergency plan recommendation system for emergencies, wherein the scope determination unit comprises:

[0086] A model analysis subunit, used for inputting the emergency event details into a detail analysis model to obtain a number of key descriptions and a description reference function of each key description;

[0087] A distribution subunit, used for inputting the key description and the description reference role into a prefabricated table in sequence to obtain a description table, and extracting and obtaining parameter distribution under each burst indicator;

[0088] A traversal subunit is used to perform distribution traversal in a historical distribution database according to the parameter distribution to obtain a distribution set matching the parameter distribution, and extract a first distribution with the greatest matching degree with the parameter distribution and a second distribution with the second greatest matching degree from the distribution set;

[0089] a threshold determination subunit, configured to determine a burst threshold corresponding to a burst indicator according to the first distribution and the second distribution, and in combination with the parameter distribution and the event parameter of the parameter distribution;

[0090] A range determination subunit, for determining a dynamic range (D01, D02) of each burst indicator according to the initial dependency of the burst indicator and the current dependency of the burst indicator under the corresponding burst threshold;

[0091]

[0092]

[0093] Wherein, ave1 represents the average dependency coefficient based on the initial dependency relationship under the corresponding burst indicator; ave2 represents the average dependency coefficient based on the current dependency relationship under the corresponding burst indicator; TY represents the burst threshold value of the corresponding burst indicator; represents the dependency coefficient of the i1th dependency factor under the initial dependency relationship; represents the dependency weight of the i1th dependency factor under the initial dependency relationship; n1 represents the total number of dependency factors under the initial dependency relationship; Represents the dependency coefficient of the j1th dependency factor under the current dependency relationship; represents the dependency weight of the j1th dependency factor under the current dependency relationship; n2 represents the total number of dependency factors under the current dependency relationship; min represents the minimum value symbol;

[0094] The curve construction subunit is used to obtain the historical threshold value of each burst indicator under the historical events consistent with all key descriptions from the historical details database, and construct a first curve for each burst indicator and a second curve for each historical event;

[0095] Performing fitting processing on all first curves to obtain a first fitting value of each first curve;

[0096] Performing curve analysis on all second curves to obtain a significance index for each second curve;

[0097] The array construction subunit is used to extract the second curves consistent with the significance index from all the first curves, and randomly lock [M1 / 2] points of each second curve to perform N1 fittings, and combine with the first fitting value to construct a fitting array, wherein M1 represents the number of historical thresholds in the first curve; [ ] represents the integer symbol; and N1 represents the number of fittings;

[0098] A combined analysis subunit is used to combine and analyze the dynamic range, the fitting array and the burst threshold to obtain a burst range (T01, T02) corresponding to the burst indicator;

[0099]

[0100]

[0101] Where d1 represents the first fitting value of the corresponding burst index; is represented as the variance of the remaining values ​​in the fitting array determined when the burst indicator is a significance indicator, except for the first fitting value; Represents the variance of all values ​​in the fitted array determined when the burst indicator is the significance indicator.

[0102] In this embodiment, the detail analysis model is obtained by training a neural network model with samples of event details matching different key descriptions, reference functions and thresholds. Therefore, the key description and description reference function can be directly obtained based on the model.

[0103] In this embodiment, the prefabricated table contains preset positions of descriptions corresponding to different burst indicators. Therefore, by inputting key descriptions and description reference functions into the table, parameter distributions of different burst indicators can be obtained, that is, they can be obtained intuitively through the table.

[0104] In this embodiment, the distribution database contains parameter distributions under different burst indicators, and matching with the parameter distribution is achieved by performing distribution traversal, that is, there are at least two matching distributions for each parameter distribution, which further indicates that the distribution set under each parameter distribution contains at least two matching distributions, which is convenient for subsequent analysis. By matching the current distribution with the historical distribution, it is convenient to reasonably analyze the current distribution and obtain the distribution with the largest and second largest matching degree.

[0105] In this embodiment, the calculation formula of the burst threshold is as follows:

[0106]

[0107] Wherein, d1 represents the threshold corresponding to the first distribution; d2 represents the threshold corresponding to the second distribution; s01 represents the coefficient proportion for the first distribution, and 1-s01 represents the coefficient proportion for the second distribution. It represents the similarity coefficient between the event parameter s1 and all standard parameters s0 under the corresponding burst index involved in the preset table.

[0108] In this embodiment, the dependency relationship between different burst indicators is pre-set and can be directly retrieved and used.

[0109] In this embodiment, the historical details database contains historical events corresponding to different descriptions and thresholds of different indicators existing at the historical time. Therefore, after retrieving the thresholds with consistent descriptions, a curve of the corresponding indicator can be constructed, wherein the first curve is constructed based on the thresholds of the same burst indicator under different historical events, and the second curve is constructed based on the thresholds of different burst indicators involved in the same historical event, wherein the different thresholds, different events, and different indicators involved are calibrated by experts.

[0110] In this embodiment, the significance index refers to an index corresponding to a value with obvious deviation extracted after normalizing the threshold of each burst index involved in the second curve.

[0111] In this embodiment, N1 fittings refer to randomly screening N1 [M1 / 2] points from the second curve, fitting the screened points once each time, and screening N1 times in total to achieve N1 fittings.

[0112] The beneficial effect of the above technical solution is: by analyzing the details based on the model and inputting them into the threshold table, it is easy to obtain the parameter distribution under different indicators, and then obtain two distribution calculation burst thresholds through distribution traversal, and then draw two curves through historical threshold retrieval to comprehensively determine the range, thereby ensuring the reliability of obtaining emergency plans.

[0113] The present invention provides an emergency plan recommendation system for emergencies, wherein the matrix construction unit comprises:

[0114] A first construction subunit is used to screen the maximum value of the burst range of each burst indicator to obtain a first value vector, and input it into the event occurrence model to obtain a first event vector;

[0115] A second construction subunit is used to screen the minimum value of the burst range of each burst indicator to obtain a second value vector, and input it into the event occurrence model to obtain a second event vector;

[0116] The third construction subunit is used to randomly select one value from the burst range corresponding to each burst indicator, and perform N2 random selections to obtain N2 third value vectors, and input them into the event occurrence model to obtain a third event vector;

[0117] The matrix construction subunit is used to construct a corresponding event matrix based on the first event vector, the second event vector and all third event vectors, wherein the event matrix has 3 columns and N2+2 rows.

[0118] In this embodiment, the first vector = [the maximum value of index 1, the maximum value of index 2, ...];

[0119] Second vector = [minimum value of index 1, minimum value of index 2, ...];

[0120] In this embodiment, the event occurrence model is pre-trained. It is obtained by experts training the neural network model with different vectors and event annotations of the vectors as samples, and the event vector can be directly obtained.

[0121] .

[0122] The beneficial effect of the above technical solution is: by obtaining a vector obtained by combining different values ​​in the range and analyzing it through an event occurrence model to obtain an event vector, it is convenient to construct an event matrix, which facilitates the subsequent acquisition of emergency plans.

[0123] The present invention provides an emergency plan recommendation system for emergencies, wherein the matching analysis module comprises:

[0124] A matching unit, used to match each row vector in the event matrix with all preset combinations involved in a preset database, and calculate a matching value;

[0125] The retention unit is used to select a preset combination with the largest matching value with the corresponding row vector from all matching values ​​for retention.

[0126] In this embodiment, the preset database contains preset information of different combinations, which are all pre-set, so the matching value can be calculated through matching, which can be calculated through a similarity function.

[0127] The beneficial effect of the above technical solution is that by comparing each row vector with the preset database, it is easy to obtain the preset combination with the largest screening matching value, which facilitates solution recommendation.

[0128] The present invention provides an emergency plan recommendation system for emergencies, wherein the plan recommendation module comprises:

[0129] Perform overlap analysis on all retained preset combinations, and select preset combinations with overlap numbers greater than or equal to 2;

[0130] According to the number of overlaps, the recommended solutions and solution arrangements of the corresponding preset combinations are obtained to obtain a list of emergency solutions.

[0131] In this embodiment, since there is a matching preset combination for each row vector, for example, row vector 1--combination 01, row vector 2--combination 02, row vector 3--combination 03, row vector 4--combination 01, at this time, the preset combination with the number of overlaps greater than or equal to 2 is combination 01.

[0132] In this embodiment, the emergency plan list includes combination 01. If there are other combinations, they are placed from top to bottom according to the number.

[0133] The beneficial effect of the above technical solution is that by performing overlap analysis on the retained preset combinations, it is convenient to filter according to the number of overlaps to obtain a list of emergency plans.

[0134] The present invention provides an emergency plan recommendation system for emergencies, wherein the details determination unit comprises:

[0135] A word segmentation extraction subunit is used to eliminate ambiguity between the first event description and the second event description, and to perform dependency analysis to extract characteristic word segments;

[0136] A set construction subunit is used to calculate the content similarity between different feature segmentations, and to aggregate the feature segmentations within a preset similarity range to form a number of detail feature sets;

[0137] An information mining subunit is used to mine the case information of the person himself, and fill each detail feature set to obtain a corresponding filled set;

[0138] The feature analysis subunit is used to obtain the details of the emergency event based on the set features of each filled set.

[0139] In this embodiment, the specific process of disambiguation includes: introducing a dictionary, collecting all words of each meaning in the dictionary as a word library, using the emergency details containing the word library as an ambiguity library, counting the keywords of each meaning marked in the emergency details, and calculating the weight of the keywords: , where S n is the weight of keyword n, P(n) is the probability of keyword n ​​appearing in the ambiguous database, and P(n|M) is the probability of keyword n ​​appearing in a word database M. The word sense similarity is weighted according to the weight of the keyword, and the word sense with the highest weighted result is taken as the disambiguation result;

[0140] In this embodiment, the event description is disambiguated to provide a reliable data source for comparing the emergency details with the preset emergency types in this method, thereby providing the hospital with a more accurate and reasonable emergency plan; extracting feature segmentations from the semantically disambiguated information can further improve the accuracy of the content analysis of the emergency details; based on the content similarity between the feature segmentations, the feature segmentations within the preset content similarity range are clustered to form a number of emergency detail feature sets, and feature segmentations with similar contents can be grouped together to reduce the amount of data so as to extract more important feature segmentations from the emergency details.

[0141] In this embodiment, case information mining is to further ensure the integrity of the information.

[0142] The beneficial effects of the above technical solution are: it is convenient to extract effective word segmentation by eliminating ambiguity and avoiding word segmentation redundancy, and it is convenient to form a feature set by performing content similarity analysis and then adjust it in combination with the mining results of case information to ensure the accuracy of the details of the emergency.

[0143] The present invention provides a method for recommending emergency plans for emergencies, such as Figure 2 As shown, including:

[0144] Step 1: Obtain and collect incident details;

[0145] Step 2: Analyze the details of the emergency event and construct an event matrix, wherein the event matrix contains emergency information of different combinations, and the emergency information under each combination is related to the emergency type, the emergency level under the emergency type, and the emergency severity under the emergency type, and each row is a burst vector constructed by the emergency information under a combination, and each column is the number of combinations;

[0146] Step 3: Match the event matrix with the preset event type, preset emergency level and preset severity in the preset database;

[0147] Step 4: Generate a list of emergency plans that match the details of the emergency event based on the matching results and output them for display.

[0148] The beneficial effect of the above technical solution is: by collecting emergency details and detailed analysis, and subsequently matching the type, level, and degree of verification, an effective emergency plan can be obtained, so as to provide the hospital with a reasonable solution in a timely manner.

[0149] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An emergency plan recommendation system for emergencies, characterized in that: include: A details collection module is used to obtain and collect details of emergencies; A details analysis module, used for analyzing the details of the emergency event to construct an event matrix, wherein the event matrix contains emergency information of different combinations, and the emergency information under each combination is related to the emergency type, the emergency level under the emergency type, and the emergency severity under the emergency type, and each row is a burst vector constructed by the emergency information under one combination, and each column is the number of combinations; A matching analysis module, used for matching the event matrix with preset event types, preset emergency levels and preset severity levels in a preset database; A plan recommendation module is used to generate a list of emergency plans that match the details of the emergency event according to the matching results and output them for display; Wherein, the detailed analysis module includes: A range determination unit, configured to extract event parameters under different emergency indicators in the emergency event details, and determine the emergency range of each emergency indicator based on the event parameters; A matrix construction unit, used for constructing an event matrix according to the burst ranges of all burst indicators; Wherein, the range determination unit includes: A model analysis subunit, used for inputting the emergency event details into a detail analysis model to obtain a number of key descriptions and a description reference function of each key description; A distribution subunit, used for inputting the key description and the description reference role into a prefabricated table in sequence to obtain a description table, and extracting and obtaining parameter distribution under each burst indicator; A traversal subunit is used to perform distribution traversal in a historical distribution database according to the parameter distribution to obtain a distribution set matching the parameter distribution, and extract a first distribution with the greatest matching degree with the parameter distribution and a second distribution with the second greatest matching degree from the distribution set; a threshold determination subunit, configured to determine a burst threshold corresponding to a burst indicator according to the first distribution and the second distribution, and in combination with the parameter distribution and the event parameter of the parameter distribution; A range determination subunit, for determining a dynamic range (D01, D02) of each burst indicator according to the initial dependency of the burst indicator and the current dependency of the burst indicator under the corresponding burst threshold; Where ave1 represents the average dependency coefficient based on the initial dependency relationship under the corresponding burst indicator; ave2 represents the average dependency coefficient based on the current dependency relationship under the corresponding burst indicator; TY represents the burst threshold of the corresponding burst indicator; Y1 i1 represents the dependency coefficient of the i1th dependency factor under the initial dependency relationship; Q1 i1 represents the dependency weight of the i1th dependency factor under the initial dependency relationship; n1 represents the total number of dependency factors under the initial dependency relationship; Y2 j1 represents the dependency coefficient of the j1th dependency factor under the current dependency relationship; Q2 j1 represents the dependency weight of the j1th dependency factor under the current dependency relationship; n2 represents the total number of dependency factors under the current dependency relationship; min represents the minimum value symbol; The curve construction subunit is used to obtain the historical threshold value of each burst indicator under the historical events consistent with all key descriptions from the historical details database, and construct a first curve for each burst indicator and a second curve for each historical event; Performing fitting processing on all first curves to obtain a first fitting value of each first curve; Performing curve analysis on all second curves to obtain a significance index for each second curve; An array construction subunit is used to extract a second curve consistent with the significance index from all the first curves, and randomly lock [M1 / 2] points of each second curve to perform N1 fittings, and combine with the first fitting value to construct a fitting array, wherein M1 represents the number of historical thresholds in the first curve; [] represents a rounding symbol; and N1 represents the number of fittings; A combined analysis subunit is used to combine and analyze the dynamic range, the fitting array and the burst threshold to obtain a burst range (T01, T02) corresponding to the burst indicator; Where d1 represents the first fitting value of the corresponding burst index; ε1 2 It is expressed as the variance of the remaining values ​​in the fitting array determined when the burst index is a significant index except the first fitting value; ε2 2 Represents the variance of all values ​​in the fitted array determined when the burst indicator is the significance indicator.

2. The emergency plan recommendation system for emergencies according to claim 1, characterized in that: The detail collection module includes: A collection unit, used to collect the first event descriptions of the emergency by a number of surrounding persons, and at the same time, collect the second event descriptions of the person who caused the emergency; The details determination unit is used to analyze the first event description and the second event description, and combine them with the case information of the person himself to obtain the details of the emergency event.

3. The emergency plan recommendation system for emergencies according to claim 1, characterized in that: The matrix construction unit comprises: A first construction subunit is used to screen the maximum value of the burst range of each burst indicator to obtain a first value vector, and input it into the event occurrence model to obtain a first event vector; A second construction subunit is used to screen the minimum value of the burst range of each burst indicator to obtain a second value vector, and input it into the event occurrence model to obtain a second event vector; The third construction subunit is used to randomly select one value from the burst range corresponding to each burst indicator, and perform N2 random selections to obtain N2 third value vectors, and input them into the event occurrence model to obtain a third event vector; The matrix construction subunit is used to construct a corresponding event matrix based on the first event vector, the second event vector and all third event vectors, wherein the event matrix has 3 columns and N2+2 rows.

4. The emergency plan recommendation system for emergencies according to claim 1, characterized in that: The matching analysis module comprises: A matching unit, used to match each row vector in the event matrix with all preset combinations involved in a preset database, and calculate a matching value; The retention unit is used to select a preset combination with the largest matching value with the corresponding row vector from all matching values ​​for retention.

5. The emergency plan recommendation system for emergencies according to claim 4, characterized in that: The solution recommendation module includes: Perform overlap analysis on all retained preset combinations, and select preset combinations with overlap numbers greater than or equal to 2; According to the number of overlaps, the recommended solutions and solution arrangements of the corresponding preset combinations are obtained to obtain a list of emergency solutions.

6. The emergency plan recommendation system for emergencies according to claim 2, characterized in that: The detail determination unit includes: A word segmentation extraction subunit is used to eliminate ambiguity between the first event description and the second event description, and to perform dependency analysis to extract characteristic word segments; A set construction subunit is used to calculate the content similarity between different feature segmentations, and to aggregate the feature segmentations within a preset similarity range to form a number of detail feature sets; An information mining subunit is used to mine the case information of the person himself, and fill each detail feature set to obtain a corresponding filled set; The feature analysis subunit is used to obtain the details of the emergency event based on the set features of each filled set.

7. A method for recommending emergency plans for emergencies, characterized in that: include: Step 1: Obtain and collect incident details; Step 2: Analyze the details of the emergency event and construct an event matrix, wherein the event matrix contains emergency information of different combinations, and the emergency information under each combination is related to the emergency type, the emergency level under the emergency type, and the emergency severity under the emergency type, and each row is a burst vector constructed by the emergency information under a combination, and each column is the number of combinations; Step 3: Match the event matrix with the preset event type, preset emergency level and preset severity in the preset database; Step 4: Generate a list of emergency plans that match the details of the emergency event according to the matching results and output them for display; Wherein, step 2 includes: Extracting event parameters under different emergency indicators in the emergency event details, and determining the emergency range of each emergency indicator based on the event parameters; Construct an event matrix according to the burst range of all burst indicators; Wherein, determining the burst range of each burst indicator based on the event parameter includes: Inputting the emergency event details into a detail analysis model to obtain a number of key descriptions and a description reference function of each key description; Input the key description and the description reference role into a prefabricated table in sequence to obtain a description table, and extract the parameter distribution under each burst indicator; Perform distribution traversal in a historical distribution database according to the parameter distribution to obtain a distribution set matching the parameter distribution, and extract a first distribution with the greatest matching degree with the parameter distribution and a second distribution with the second greatest matching degree from the distribution set; Determine a burst threshold corresponding to a burst indicator according to the first distribution and the second distribution, and in combination with the parameter distribution and the event parameter of the parameter distribution; Determine the dynamic range (D01, D02) of each burst indicator according to the initial dependency of the burst indicator and the current dependency of the burst indicator under the corresponding burst threshold; Where ave1 represents the average dependency coefficient based on the initial dependency relationship under the corresponding burst indicator; ave2 represents the average dependency coefficient based on the current dependency relationship under the corresponding burst indicator; TY represents the burst threshold of the corresponding burst indicator; Y1 i1 represents the dependency coefficient of the i1th dependency factor under the initial dependency relationship; Q1 i1 represents the dependency weight of the i1th dependency factor under the initial dependency relationship; n1 represents the total number of dependency factors under the initial dependency relationship; Y2 j1 represents the dependency coefficient of the j1th dependency factor under the current dependency relationship; Q2 j1 represents the dependency weight of the j1th dependency factor under the current dependency relationship; n2 represents the total number of dependency factors under the current dependency relationship; min represents the minimum value symbol; Obtaining historical thresholds of each burst indicator under historical events consistent with all key descriptions from a historical detail database, and constructing a first curve for each burst indicator and a second curve for each historical event; Performing fitting processing on all first curves to obtain a first fitting value of each first curve; Performing curve analysis on all second curves to obtain a significance index for each second curve; A second curve consistent with the significance index is extracted from all the first curves, and [M1 / 2] points of each second curve are randomly locked for N1 fittings, and a fitting array is constructed by combining with the first fitting value, wherein M1 represents the number of historical thresholds in the first curve; [] represents the integer symbol; and N1 represents the number of fittings; The dynamic range, the fitting array and the burst threshold are combined and analyzed to obtain a burst range (T01, T02) corresponding to the burst index; Where d1 represents the first fitting value of the corresponding burst index; ε1 2 It is expressed as the variance of the remaining values ​​in the fitting array determined when the burst index is a significant index except the first fitting value; ε2 2 Represents the variance of all values ​​in the fitted array determined when the burst indicator is the significance indicator.