WO2023098445A1 - 食品安全突发事件的应急处置推荐方法及系统 - Google Patents

食品安全突发事件的应急处置推荐方法及系统 Download PDF

Info

Publication number
WO2023098445A1
WO2023098445A1 PCT/CN2022/131071 CN2022131071W WO2023098445A1 WO 2023098445 A1 WO2023098445 A1 WO 2023098445A1 CN 2022131071 W CN2022131071 W CN 2022131071W WO 2023098445 A1 WO2023098445 A1 WO 2023098445A1
Authority
WO
WIPO (PCT)
Prior art keywords
emergency
model
information
database
hazard factor
Prior art date
Application number
PCT/CN2022/131071
Other languages
English (en)
French (fr)
Inventor
韩小敏
李孟寒
李凤琴
徐进
闫韶飞
Original Assignee
国家食品安全风险评估中心
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 国家食品安全风险评估中心 filed Critical 国家食品安全风险评估中心
Publication of WO2023098445A1 publication Critical patent/WO2023098445A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9536—Search customisation based on social or collaborative filtering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2457—Query processing with adaptation to user needs
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/28—Databases characterised by their database models, e.g. relational or object models
    • G06F16/284—Relational databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/26—Government or public services
    • G06Q50/265—Personal security, identity or safety

Definitions

  • the invention belongs to the technical field of food safety, and in particular relates to a method and system for recommending emergency response to food safety emergencies.
  • the data set for food safety emergencies is not perfect, including
  • the source of the occurrence that is, the hazard factors, the categories of the hazard factors, and the emergencies caused by the specific hazard factors have not yet had a detailed associated data set; for example, the biological hazard factors include bacteria, fungi, viruses and natural toxins, and among them There are also various subtypes, and the emergencies they bring also have obvious differences to a certain extent.
  • the development of data clustering algorithms Most of the food safety emergencies are handled through the clustering of patients’ symptoms, and the selected data features are relatively limited.
  • the present invention intends to provide a method and system for recommending emergency response to food safety emergencies, so as to solve the problem that users cannot obtain emergency methods in a simple way at the first time when facing food safety emergencies.
  • the technical problem to be solved by the invention is realized through the following technical solutions:
  • the present invention provides a method for recommending emergency response to food safety emergencies, including:
  • the matching degree of the poisoning hazard factor is greater than a preset matching degree threshold, obtaining preliminary poisoning hazard factor information from the hazard factor database;
  • the recommended score of the standard emergency plan is less than the preset score threshold, then use a retrospective model to match and analyze the description information of the user emergency with the data in the emergency disposal case database to obtain the first case matching result;
  • the matching result of the first case is a successful matching case
  • the information of the target poisoning hazard factor and the treatment plan of the target case are obtained, and the information of the target poisoning hazard factor includes the name of the target poisoning hazard factor and the name of the target poisoning hazard.
  • Hazard factor category
  • the target poisoning hazard factor information and the data in the expert information database are subjected to decision-making analysis to obtain target expert recommendation information.
  • the step of obtaining preliminary poisoning hazard factor information from the hazard factor database further includes:
  • the matching degree of the poisoning hazard factor is less than the preset matching degree threshold, using the retrospective model to perform matching analysis on the description information of the user emergency and the data in the emergency treatment case database to obtain a second case matching result;
  • the matching result of the second case is an unsuccessful matching case, a prompt message that cannot provide help is fed back to the user.
  • the recommended score of the standard emergency plan is lower than the preset score threshold, then use the retrospective model to match the description information of the user emergency with the data in the emergency treatment case database to obtain the first case matching
  • the resulting steps also include:
  • the step of acquiring description information of user emergencies includes:
  • the description text of the emergency event is analyzed and key features are extracted to obtain the description information of the user emergency event.
  • the step further includes:
  • the data exploratory analysis model is used to carry out variable analysis and importance analysis on the hazard factor prediction model, the standard evaluation model, the retrospective model and the expert-aided decision-making model to determine the corresponding model output characteristics.
  • the standard evaluation model is created using a multiple nonlinear regression analysis method.
  • the expert-aided decision-making model is created using a collaborative filtering recommendation method.
  • the present invention also provides an emergency response recommendation system for food safety emergencies, including:
  • An event acquisition module configured to acquire description information of user emergencies
  • the hazard factor prediction module is configured to use the hazard factor prediction model to perform similarity analysis between the description information of the user emergency and the data in the hazard factor database to obtain the matching degree of poisoning hazard factors;
  • the hazard information acquisition module is configured to obtain preliminary poisoning hazard factor information from the hazard factor database if the matching degree of the poisoning hazard factor is greater than a preset matching degree threshold;
  • the evaluation and analysis module is configured to use the standard evaluation model to evaluate and analyze the preliminary poisoning hazard information and the data in the standard emergency plan database to obtain the recommended score of the standard emergency plan;
  • the case matching module is configured to use a retrospective model to match and analyze the description information of the user's emergency with the data in the emergency disposal case database to obtain the first case if the recommended score of the standard emergency plan is less than the preset score threshold matching result;
  • the target hazard information acquisition module is configured to acquire the target poisoning hazard information and the target case treatment plan according to the first case matching result if the matching result of the first case is a successful matching case, and the target poisoning hazard information Include the name of the target poisoning hazard factor and the category of the target poisoning hazard factor;
  • the expert recommendation module is configured to use the expert-assisted decision-making model to conduct decision analysis on the target poisoning hazard factor information and the data in the expert information database to obtain target expert recommendation information.
  • the present invention also provides an electronic device including: a processor and a memory, where computer-readable instructions are stored on the memory, and when the computer-readable instructions are executed by the processor, the above-mentioned Recommended methods for emergency response to food safety emergencies.
  • the present invention also provides a computer-readable storage medium, on which a computer program is stored, which is characterized in that, when the computer program is executed by a processor, the above-mentioned emergency response to food safety emergencies is realized. Disposal recommended method.
  • the emergency response recommendation method for food safety emergencies of the present invention is formed by associating the constructed hazard factor database, emergency response case database, and standard emergency plan database into a comprehensive database, as the data set of the recommended method, and then through the hazard factor prediction model, standard
  • the evaluation model, backtracking model, and expert-aided decision-making model constitute the recommendation algorithm for emergency plans, making immediate decisions on the characteristics of emergency-related data submitted by users and recommending emergency plans. Experts perform remote visual diagnosis and decision-making. Therefore, the recommended method for emergency response to food safety emergencies of the present invention solves the problem that users cannot obtain emergency methods in a simple manner in the first instance of emergencies caused by food safety issues in public places and family environments .
  • Fig. 1 is a schematic flow chart of some embodiments of the recommended method for emergency response to food safety emergencies of the present invention
  • Fig. 2 is a schematic flow chart of other embodiments of the recommended method for emergency response to food safety emergencies of the present invention
  • Fig. 3 is the data flow flow diagram of the emergency response recommendation method of the food safety emergency according to the embodiment of the present invention.
  • Fig. 4 is a schematic diagram of the association of each model of the recommended method for emergency response to food safety emergencies according to an embodiment of the present invention
  • Fig. 5 is the framework diagram of database design of the recommended method for emergency response to food safety emergencies implemented by the present invention
  • FIG. 6 is a flow chart of model building for the recommended method for emergency response to food safety emergencies according to an embodiment of the present invention
  • Fig. 7 is a structural block diagram of some embodiments of the recommendation system for emergency response to food safety emergencies according to the present invention.
  • an embodiment of the present invention provides a recommended method for emergency response to food safety emergencies, including:
  • Step 100 Obtain description information of user emergencies
  • the description information of the emergency event of the user is obtained through literal text or voice uploaded by the user.
  • Step 200 Use the hazard factor prediction model to perform similarity analysis between the description information of the user emergency and the data in the hazard factor database to obtain the matching degree of poisoning hazard factors;
  • the pre-trained hazard factor prediction model is used to analyze the similarity between the description information of user emergencies and the data in the hazard factor database, and the model outputs the matching degree results of each hazard factor to obtain possible poisoning according to the matching degree Hazard factor.
  • the hazard factor database in this embodiment includes data of biological hazard factors, data of radioactive hazard factors, and data of chemical hazard factors.
  • Step 300 If the matching degree of the poisoning hazard factor is greater than the preset matching degree threshold, obtain preliminary poisoning hazard factor information from the hazard factor database;
  • possible poisoning hazard factors are screened according to a preset matching degree threshold to obtain preliminary poisoning hazard factor information.
  • the preset matching degree threshold is an empirical value obtained from training experiments, and is not specifically limited here.
  • Step 400 Using the standard evaluation model to evaluate and analyze the preliminary poisoning hazard factor information and the data in the standard emergency plan database to obtain the recommended score of the standard emergency plan;
  • the data in the standard emergency plan database is the standard emergency method specification data compiled by the country/industry.
  • the initial poisoning hazard factor information and the data in the standard emergency plan database can be evaluated and analyzed to obtain each standard.
  • the recommendation score of the emergency method, and the emergency method with a high score is recommended first.
  • Step 500 If the recommended score of the standard emergency plan is less than the preset score threshold, use the retrospective model to match and analyze the description information of the user's emergency with the data in the emergency disposal case database to obtain the first case matching result;
  • the preset score threshold is an empirical value obtained from a training experiment, and is not specifically limited here.
  • Step 600 If the matching result of the first case is a successful matching case, then according to the matching result of the first case, obtain the information of the target poisoning hazard factor and the treatment plan of the target case, the information of the target poisoning hazard factor includes the name of the target poisoning hazard factor and the category of the target poisoning hazard factor ;
  • Step 700 Use the expert-aided decision-making model to conduct decision-making analysis on the target poisoning hazard factor information and the data in the expert information database to obtain target expert recommendation information.
  • the pre-trained expert-aided decision-making model conducts decision-making analysis on the target poisoning hazard factor information and the data in the expert information database, thereby outputting recommended expert information.
  • the recommended method for emergency response to food safety emergencies in the embodiment of the present invention combines the constructed hazard factor database, emergency response case database, and standard emergency plan database into a comprehensive database, as a data set for the recommended method, and then through the hazard factor prediction model , standard evaluation model, backtracking model, and expert-aided decision-making model to form a recommendation algorithm for emergency plans. It makes immediate decisions on the characteristics of emergency-related data submitted by users and recommends emergency plans. At the same time, according to the specific decision-making results, Recommend specific experts for remote visual diagnosis and decision-making.
  • the recommended method for emergency response to food safety emergencies in the embodiment of the present invention solves the problem that users cannot obtain emergency methods in a simple way at the first time for emergencies caused by food safety issues in public places and family environments .
  • the recommended method for emergency response to food safety emergencies of the present invention further includes:
  • Step 500 If the matching degree of the poisoning hazard factor is less than the preset matching degree threshold, use the retrospective model to match and analyze the description information of the user's emergency with the data in the emergency treatment case database to obtain the second case matching result;
  • the hazard factor prediction model cannot be used to find possible poisoning hazard factors in the hazard factor database, so it is necessary to re-match and analyze the description information of user emergencies with the data in the emergency disposal case database to find a matching emergency disposal case.
  • Step 900 If the matching result of the second case is an unsuccessful matching case, feed back a prompt message that no help can be provided to the user.
  • the recommended method for emergency response to food safety emergencies of the present invention further includes:
  • Step 800 If the recommended score of the standard contingency plan is greater than the preset score threshold, then obtain the target standard contingency plan from the standard contingency plan database;
  • the recommended standard emergency plan is obtained from the standard emergency plan database, and the preliminary poisoning hazard factor information is used as the target poison hazard factor information.
  • step 700 use the expert-assisted decision-making model to conduct decision-making analysis on the preliminary poisoning hazard factor information and the data in the expert information database to obtain target expert recommendation information.
  • Fig. 3 shows the data flow flow chart of the method for recommending emergency response to food safety emergencies according to an embodiment of the present invention.
  • the description information of the emergency incident by the user i.e., the patient
  • the model outputs the category of poisoning hazard factors and the names of hazard factors whose matching degree is higher than the preset matching degree threshold (for example, 85%), and then enters the standard evaluation model to obtain the recommended score of emergency measures. If the matching degree is lower than the preset matching degree threshold (for example, 85%), the initial event description information is further used as an input of the retrospective model to obtain a case matching result.
  • the category and name of the hazard factors are used as the input of the auxiliary decision-making model, and the recommended experts are output for remote visual diagnosis and decision-making. For example, if the user enters symptoms: "abdominal pain, joint cramps, stomach discomfort", age: 18-22 years old, food: undercooked beef and other key characteristics, the hazard factor prediction model can accurately predict the type of hazard factor And the similarity score of the corresponding category of hazard factors, according to the ranking of the scores as the recommendation basis, here the top three hazard factors are extracted as recommendations, such as: Salmonella, Escherichia coli, and Clostridium perfringens. Then query the hazard factor database to obtain the corresponding emergency method, and evaluate the emergency method according to the standard scheme scoring model. Ultimately a standard emergency approach is recommended. These include specific experts obtained through hazard factor query associations.
  • step 100 in the recommended method for emergency response to food safety emergencies of the present invention includes:
  • the existing semantic analysis method can be used to analyze the description text of the emergency, and then extract key features according to the analysis result, and then obtain the description information of the user emergency.
  • Symptoms caused by Clostridium perfringens acute gastroenteritis type: more than 90% of patients are mainly acute gastroenteritis such as abdominal distension, abdominal pain, and diarrhea. Abdominal pain can be manifested as severe abdominal cramps. Diarrhea is generally loose or watery stools, and sometimes mucus or bloody stools may appear. The feces have a rancid smell, and a lot of gas is produced, and diarrhea occurs several times to more than 10 times. Abdominal distension is common, nausea and vomiting are rare, body temperature is normal or low-grade fever is common, and symptoms of infection such as headache, dizziness, weakness, and soreness are rare. We extract symptoms through the word segmentation model, and the result is:
  • a list of symptoms transforming unstructured data into structured data.
  • the recommended method for emergency response to food safety emergencies of the present invention also includes before step 100:
  • the data exploratory analysis model is used to carry out variable analysis and importance analysis on the hazard factor prediction model, standard evaluation model, retrospective model and expert-aided decision-making model to determine the corresponding model output characteristics.
  • the original data set is provided in the form of documents and pictures. It is necessary to first construct feature word segmentation through semantic word segmentation, design the form, and finally enter it into the MySQL database; most of the data fields are text content of string type. It is necessary to truncate the word segmentation again to form a word segmentation array, and then standardize the coding through Onehot label; for the selection of the output features of the model, it is necessary to perform univariate analysis, multivariate analysis, and importance analysis for each feature. Therefore, the main tasks of the data exploratory analysis model are: 1) process optimization and data analysis report on the food safety data set. Transform unstructured data into structured data, and simultaneously perform univariate and multivariate analysis on features.
  • the random forest method is used to create a hazard factor prediction model in the recommended method for emergency response to food safety emergencies in the embodiment of the present invention.
  • a decision tree method is used to create a retrospective model.
  • a multiple nonlinear regression analysis method is used to create a standard evaluation model.
  • a collaborative filtering recommendation method is used to create an expert-assisted decision-making model.
  • the hazard factor prediction model, the retrospective model, the standard evaluation model, and the expert-aided decision-making model can be created using other machine learning methods, which are not described here.
  • Fig. 5 shows a database design frame diagram of the recommended method for emergency response to food safety emergencies implemented by the present invention.
  • a comprehensive database is built, that is, a regionalized and localized food safety emergency and emergency response standard database is built, and a MySQL database management system is created to realize functions such as input and retrieval of self-owned data sets; and Connect with other databases of the food safety emergency response system.
  • the main contents include the following: 1) Creation of the database of biological hazard factors 2) Creation of the database of radioactive hazard factors 3) Creation of the database of chemical hazard factors 4) Creation of the emergency case database (ie emergency disposal case database) 5) 6) Creation of expert information database 7) Creation and query of comprehensive database 8) Import, download, preview, edit and other functions of database management terminal.
  • the specific environment configuration is as follows:
  • WEB server CPU 2*E5-2630 V4 2.2GHz 10 cores; 128GB memory; 4*900GB 10K HDD hard disk; Raid card.
  • File server CPU 2*E5-2630 V4 2.2GHz 10 cores CPU, 128GB memory, 2*900GB 10K HDD+6*8TB 7.2K HDD hard disk; Raid card.
  • CPU P3 or above.
  • Memory 256M or more.
  • Hard disk more than 20G.
  • Resolution 1024*768 pixels are recommended
  • Operating system choose one of Microsoft windows2000 professional, Microsoft window2000 server, Microsoft windowXP Professional, Windows 7 Professional. Browser: Firefox, Chrome or above.
  • an integrated design is carried out from the data segment to the model end and the final remote visualization end, and the final emergency response recommendation for food safety emergencies is used as an app software for users to use.
  • the entire process is carried out online in real time, from user input to query results and the final remote diagnosis of connected experts.
  • the user can simply output the necessary characteristic data required by the system, and then get the corresponding hazard factor category, hazard factor name, emergency response method and expert recommendation.
  • This embodiment can recommend timely emergency measures to the user to avoid unnecessary losses.
  • Fig. 6 shows the model construction flow chart of the recommended method for emergency response to food safety emergencies in the embodiment of the present invention
  • the construction process of the hazard factor prediction model, the retrospective model, the standard evaluation model and the expert-aided decision-making model are basically the same, the general process To: first extract data from the corresponding database, perform necessary data filling and data segmentation processing, then conduct importance analysis and variable selection for the corresponding model data set, determine the model training set and test set, and then carry out model development. After the model development is completed, use the test set to optimize, adjust the model parameters, and finally evaluate the model performance. When the model performance meets the design requirements, it will be deployed online.
  • GridSearchCV can be used to optimize the model parameters
  • the SHAP visualization tool can be used to explain the above machine learning model and the influence of each feature variable on the prediction result.
  • the embodiment of the present invention also provides an emergency response recommendation system 1 for food safety emergencies, including:
  • the event obtaining module 10 is configured to obtain description information of user emergencies
  • the hazard factor prediction module 20 is configured to use the hazard factor prediction model to perform similarity analysis between the description information of the user emergency and the data in the hazard factor database to obtain the matching degree of poisoning hazard factors;
  • the hazard information acquisition module 30 is configured to obtain preliminary poisoning hazard factor information from the hazard factor database if the matching degree of the poisoning hazard factor is greater than the preset matching degree threshold;
  • the evaluation analysis module 40 is configured to evaluate and analyze the preliminary poisoning hazard factor information and the data in the standard emergency plan database by using the standard evaluation model to obtain the recommended score of the standard emergency plan;
  • the case matching module 50 is configured to, if the recommended score of the standard emergency plan is less than the preset score threshold, use the retrospective model to match and analyze the description information of the user's emergency with the data in the emergency disposal case database to obtain the first case matching result;
  • the target hazard information acquisition module 60 is configured to obtain the target poisoning hazard information and the target case treatment plan according to the first case matching result if the first case matching result is a successful matching case, the target poisoning hazard information includes the target poisoning hazard factor Name and target poisoning hazard category;
  • the expert recommendation module 70 is configured to use the expert-assisted decision-making model to perform decision analysis on the target poisoning hazard factor information and the data in the expert information database to obtain target expert recommendation information.
  • the embodiment of the present invention also provides an electronic device, including: a processor and a memory, where computer-readable instructions are stored, and when the computer-readable instructions are executed by the processor, the food safety described in the above-mentioned embodiments is realized. Recommended methods for emergency response to emergencies.
  • the above-mentioned memory and processor can be general-purpose memory and processor, which are not specifically limited here.
  • the processor runs the computer-readable instructions stored in the memory, it can execute the food safety emergencies described in the above-mentioned embodiments. Recommended methods for emergency response.
  • the embodiments of the present invention also provide a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the method for recommending emergency response to food safety emergencies in the above-mentioned embodiments is implemented.
  • spatially relative terms may be used here, such as “on !, “over !, “on the surface of !, “above”, etc., to describe The spatial positional relationship between one device or feature shown and other devices or features. It will be understood that the spatially relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation depicted in the figures. For example, if the device in the figures is turned over, devices described as “above” or “above” other devices or configurations would then be oriented “beneath” or “above” the other devices or configurations. under other devices or configurations”. Thus, the exemplary term “above” can encompass both an orientation of “above” and “beneath”. The device may be oriented in different ways, rotated 90 degrees or at other orientations, and the spatially relative descriptions used herein interpreted accordingly.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Databases & Information Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Development Economics (AREA)
  • Operations Research (AREA)
  • Computational Linguistics (AREA)
  • Quality & Reliability (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Computer Security & Cryptography (AREA)
  • Educational Administration (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

涉及食品安全突发事件的应急处置推荐方法及系统，通过将构建的危害因子数据库、应急处置案例数据库和标准应急方案数据库关联成为综合数据库，作为推荐方法的数据集，再通过危害因子预测模型、标准评估模型、回溯模型以及专家辅助决策模型组成应急方案的推荐算法，对用户提交的突发事件相关数据特征进行第一时间的决策和给出应急方案的推荐，同时根据具体的决策结果，推荐具体的专家进行远程可视化诊断和决策，解决了在公共场所和家庭环境下由食品安全问题所引发突发事件时用户在第一时间无法通过简单的方式获取应急方法的问题。

Description

食品安全突发事件的应急处置推荐方法及系统
相关申请的交叉引用
本申请要求享有于2021年11月30日提交的名称为“食品安全突发事件的应急处置推荐方法及系统”的中国专利申请CN 202111473542.7的优先权，上述申请的全部内容通过引用并入本文中。
技术领域
本发明属于食品安全技术领域，尤其涉及一种食品安全突发事件的应急处置推荐方法及系统。
背景技术
食品安全突发事件一直是当下卫生部门极为重视的问题，针对食品安全突发事件的应急处置目前市场上大多数通过发现问题之后到医院寻求诊治或者通过网络在线问答医生询问的方式进行处理。这些方式在对于突发事件的应急方法上无法给到及时的处理，突发事件的患者由于未能在第一时间获取当下应急方法和措施，导致更坏的结果甚至对生命安全造成威胁。目前要创建一个食品突发事件应急方法的推荐系统，有以下几个问题尚未解决：1、数据库的缺乏，现阶段对于食品安全突发事件的整理数据集尚不完善，其中包括导致食品安全事件发生的根源，即危害因子，其危害因子的类别以及具体的危害因子所导致的突发事件尚未有详细关联数据集；如生物危害因子中包含细菌、真菌、病毒和天然毒素，而在这当中还存在各种不同的亚型，其所带来的突发事件也在一定程度上有明显的差异性。2、数据聚类算法的开发，针对食品安全突发事件的处理方式大多是通过患者的症状进行聚类，选取的数据特征比较局限，常见的有梅奥诊所的症状-疾病的监督模型、根据症状和其他特征诊断某人患该病的概率以及通过症状推荐响应的药方等等。在实际情况下，虽然机器学习模型和自然语言模型已经得到广泛的应用，但是在针对具体的项目中，对于数据挖掘过程中特征的选择至关重要，直接影响着预测或推荐的准确性和泛化能力。这也是目前尚未开发出一种针对食品安全突发事件应急方法推荐的准确模 型的主要原因之一。3、专家远程可视化应急方法推荐，在前两者问题没有解决的情况下，对于如何选择合适的专家进行远程可视化诊断与决策也是一个无法解决的问题。准确的专家推荐结果依赖于危害因子类别的聚类结果以及具体危害因子的名称。从而推荐该领域具体的专家，在第一时间针对情形严重者给予远程可视化应急方法推荐。
因此，如何在面对食品安全突发事件时，用户在第一时间通过简单的方式获取应急方法是当下亟待解决的问题。
发明内容
本发明意在提供一种食品安全突发事件的应急处置推荐方法及系统，以解决当前在面对食品安全突发事件时，用户在第一时间无法通过简单的方式获取应急方法的问题，本发明要解决的技术问题通过以下技术方案来实现：
一方面，本发明提供了一种食品安全突发事件的应急处置推荐方法，包括：
获取用户突发事件的描述信息；
利用危害因子预测模型将所述用户突发事件的描述信息与危害因子数据库中数据进行相似性分析得到中毒危害因子匹配度；
如果所述中毒危害因子匹配度大于预设匹配度阈值，则从所述危害因子数据库中获取初步中毒危害因子信息；
利用标准评估模型将所述初步中毒危害因子信息与标准应急方案数据库中数据进行评估分析得到标准应急方案的推荐得分；
如果所述标准应急方案的推荐得分小于预设得分阈值，则利用回溯模型将所述用户突发事件的描述信息与应急处置案例数据库中数据进行匹配分析得到第一案例匹配结果；
如果所述第一案例匹配结果为成功匹配案例，则根据所述第一案例匹配结果获取目标中毒危害因子信息和目标案例处置方案，所述目标中毒危害因子信息包括目标中毒危害因子名称和目标中毒危害因子类别；
利用专家辅助决策模型将所述目标中毒危害因子信息与专家信息数据库中数据进行决策分析得到目标专家推荐信息。
优选地，所述如果所述中毒危害因子匹配度大于预设匹配度阈值，则从所述危害因子数据库中获取初步中毒危害因子信息的步骤还包括：
如果所述中毒危害因子匹配度小于预设匹配度阈值，则利用所述回溯模型将所述用户突发事件的描述信息与所述应急处置案例数据库中数据进行匹配分析得到第二案例匹配结果；
如果所述第二案例匹配结果为未成功匹配案例，则向用户反馈无法提供帮助的提示信息。
优选地，所述如果所述标准应急方案的推荐得分低于预设得分阈值，则利用回溯模型将所述用户突发事件的描述信息与应急处置案例数据库中数据进行匹配分析得到第一案例匹配结果的步骤还包括：
如果所述标准应急方案的推荐得分大于预设得分阈值，则从所述标准应急方案数据库中获得目标标准应急方案；
利用所述专家辅助决策模型将所述初步中毒危害因子信息与所述专家信息数据库中数据进行决策分析得到目标专家推荐信息。
优选地，所述获取用户突发事件的描述信息的步骤包括：
获取用户输入的突发事件的描述文本；
对所述突发事件的描述文本进行分析并提取关键特征，以得到所述用户突发事件的描述信息。
优选地，所述获取用户突发事件的描述信息的步骤之前还包括：
利用数据探索性分析模型分别对所述危害因子预测模型、所述标准评估模型、所述回溯模型和所述专家辅助决策模型训练所需的数据集进行标准化处理；
利用数据探索性分析模型分别对所述危害因子预测模型、所述标准评估模型、所述回溯模型和所述专家辅助决策模型进行变量分析及重要性分析以确定相应的模型输出特征。
优选地，采用多元非线性回归分析方法创建所述标准评估模型。
优选地，采用协同过滤推荐方法创建所述专家辅助决策模型。
另一方面，本发明还提供了一种食品安全突发事件的应急处置推荐系统，包括：
事件获取模块，被配置为获取用户突发事件的描述信息；
危害因子预测模块，被配置为利用危害因子预测模型将所述用户突发事件的描述信息与危害因子数据库中数据进行相似性分析得到中毒危害因子匹配度；
危害信息获取模块，被配置为如果所述中毒危害因子匹配度大于预设匹配度阈值，则从所述危害因子数据库中获取初步中毒危害因子信息；
评估分析模块，被配置为利用标准评估模型将所述初步中毒危害因子信息与标准应急方案数据库中数据进行评估分析得到标准应急方案的推荐得分；
案例匹配模块，被配置为如果所述标准应急方案的推荐得分小于预设得分阈值，则利用回溯模型将所述用户突发事件的描述信息与应急处置案例数据库中数据进行匹配分析得到第一案例匹配结果；
目标危害信息获取模块，被配置为如果所述第一案例匹配结果为成功匹配案例，则根据所述第一案例匹配结果获取目标中毒危害因子信息和目标案例处置方案，所述目标中毒危害因子信息包括目标中毒危害因子名称和目标中毒危害因子类别；
以及，专家推荐模块，被配置为利用专家辅助决策模型将所述目标中毒危害因子信息与专家信息数据库中数据进行决策分析得到目标专家推荐信息。
再一方面，本发明还提供了一种电子设备包括：处理器和存储器，所述存储器上存储有计算机可读指令，所述计算机可读指令被所述处理器执行时实现如上述所述的食品安全突发事件的应急处置推荐方法。
又一方面，本发明还提供了一种计算机可读存储介质，其上存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现如上述所述的食品安全突发事件的应急处置推荐方法。
本发明的食品安全突发事件的应急处置推荐方法通过将构建的危害因子数据库、应急处置案例数据库和标准应急方案数据库关联成为综合数据库，作为推荐方法的数据集，再通过危害因子预测模型、标准评估模型、回溯模型以及专家辅助决策模型组成应急方案的推荐算法，对用户提交的突发事件相关数据特征进行第一时间的决策和给出应急方案的推荐，同时根据具体的决策结果， 推荐具体的专家进行远程可视化诊断和决策。因此，本发明的食品安全突发事件的应急处置推荐方法解决了在公共场所和家庭环境下由食品安全问题所引发的突发事件，用户在第一时间无法通过简单的方式获取应急方法的问题。
附图说明
图1为本发明食品安全突发事件的应急处置推荐方法的一些实施例的流程示意图；
图2为本发明食品安全突发事件的应急处置推荐方法的另一些实施例的流程示意图；
图3为本发明实施例的食品安全突发事件的应急处置推荐方法的数据流流程图；
图4为本发明实施例的食品安全突发事件的应急处置推荐方法的各模型关联示意图；
图5为本发明实施的食品安全突发事件的应急处置推荐方法的数据库设计框架图；
图6为本发明实施例的食品安全突发事件的应急处置推荐方法的模型构建流程图；
图7为本发明食品安全突发事件的应急处置推荐系统的一些实施例的结构框图。
具体实施方式
需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互组合。下面将参考附图并结合实施例来详细说明本发明。
参见图1所示，本发明实施例提供了一种食品安全突发事件的应急处置推荐方法，包括：
步骤100：获取用户突发事件的描述信息；
本步骤中通过用户上传的文字性文本或者语音等获取用户突发事件的描述信息。
步骤200：利用危害因子预测模型将用户突发事件的描述信息与危害因子数据库中数据进行相似性分析得到中毒危害因子匹配度；
本步骤中通过预先训练好的危害因子预测模型对将用户突发事件的描述信息与危害因子数据库中数据进行相似性分析，模型输出各个危害因子的匹配度结果，以根据匹配度获取可能的中毒危害因子。需要说明的是，本实施例中危害因子数据库包括生物性危害因子数据、放射性危害因子数据以及化学性危害因子数据。
步骤300：如果中毒危害因子匹配度大于预设匹配度阈值，则从危害因子数据库中获取初步中毒危害因子信息；
本步骤中根据预设匹配度阈值筛选可能的中毒危害因子以得到初步中毒危害因子信息。需要说明的是，预设匹配度阈值是根据训练试验得到的经验数值，在此不做具体限定。
步骤400：利用标准评估模型将初步中毒危害因子信息与标准应急方案数据库中数据进行评估分析得到标准应急方案的推荐得分；
本步骤中标准应急方案数据库中数据为国家/行业整理的标准应急方法规范数据，通过预先训练好的标准评估模型将初步中毒危害因子信息与标准应急方案数据库中数据进行评估分析可以获得每个标准应急方法的推荐得分，得分高的应急方法优先推荐。
步骤500：如果标准应急方案的推荐得分小于预设得分阈值，则利用回溯模型将用户突发事件的描述信息与应急处置案例数据库中数据进行匹配分析得到第一案例匹配结果；
本步骤中如果在标准应急方案数据库中无法获得推荐的应急方案，则需要重新将用户突发事件的描述信息与应急处置案例数据库中数据进行匹配分析，寻找相匹配的应急处置案例。需要说明的是，预设得分阈值是根据训练试验得到的经验数值，在此不做具体限定。
步骤600：如果第一案例匹配结果为成功匹配案例，则根据第一案例匹配结果获取目标中毒危害因子信息和目标案例处置方案，目标中毒危害因子信息包括目标中毒危害因子名称和目标中毒危害因子类别；
步骤700：利用专家辅助决策模型将目标中毒危害因子信息与专家信息数据库中数据进行决策分析得到目标专家推荐信息。
本步骤中预先训练好的专家辅助决策模型对目标中毒危害因子信息与专家信息数据库中数据进行决策分析，从而输出推荐的专家信息。
本发明实施例的食品安全突发事件的应急处置推荐方法通过将构建的危害因子数据库、应急处置案例数据库和标准应急方案数据库关联成为综合数据库，作为推荐方法的数据集，再通过危害因子预测模型、标准评估模型、回溯模型以及专家辅助决策模型组成应急方案的推荐算法，对用户提交的突发事件相关数据特征进行第一时间的决策和给出应急方案的推荐，同时根据具体的决策结果，推荐具体的专家进行远程可视化诊断和决策。本发明实施例的食品安全突发事件的应急处置推荐方法解决了在公共场所和家庭环境下由食品安全问题所引发的突发事件，用户在第一时间无法通过简单的方式获取应急方法的问题。
在一些实施例中，参见图2所示，本发明的食品安全突发事件的应急处置推荐方法还包括：
步骤500：如果中毒危害因子匹配度小于预设匹配度阈值，则利用回溯模型将用户突发事件的描述信息与应急处置案例数据库中数据进行匹配分析得到第二案例匹配结果；
本步骤中利用危害因子预测模型无法在危害因子数据库中找到可能的中毒危害因子，则重新需要重新将用户突发事件的描述信息与应急处置案例数据库中数据进行匹配分析，寻找相匹配的应急处置案例。
步骤900：如果第二案例匹配结果为未成功匹配案例，则向用户反馈无法提供帮助的提示信息。
本步骤中，如果在应急处置案例数据库中也无法找到匹配的案子，只能及时反馈用户无法提供帮助的信息，用户可以选择直接去医院就医。
在一些实施例中，参见图2所示，本发明的食品安全突发事件的应急处置推荐方法还包括：
步骤800：如果标准应急方案的推荐得分大于预设得分阈值，则从标准应 急方案数据库中获得目标标准应急方案；
本步骤中当标准应急方案的推荐得分大于预设得分阈值，则从标准应急方案数据库中获得推荐的标准应急方案，并将初步中毒危害因子信息作为目标毒危害因子信息。
之后转入步骤700：利用专家辅助决策模型将初步中毒危害因子信息与专家信息数据库中数据进行决策分析得到目标专家推荐信息。
图3示出了本发明实施例的食品安全突发事件的应急处置推荐方法的数据流流程图，具体的，通过用户(即患者)对于突发事件的描述信息，作为危害因子预测模型的输入，模型输出中毒危害因子的类别，以及匹配度高于预设匹配度阈值(例如85％)的危害因子名称，之后进入标准评估模型，获取应急方法的推荐得分。如果匹配度低于预设匹配度阈值(例如85％)，则进一步将起始的事件描述信息作为回溯模型的输入，得到案例匹配结果。最终将危害因子的类别和名称作为辅助决策模型的输入，输出推荐的专家，用于远程可视化诊断与决策。如用户输入症状：“腹痛、关节绞痛、胃部不适”，年龄：18～22岁，食用食物：未煮熟的牛肉等关键特征，通过危害因子预测模型能够准确的预测出危害因子的种类以及对应类别危害因子的相似性得分，根据得分排序作为推荐依据，这里抽取排名前三的危害因子作为推荐，如：沙门氏菌、大肠杆菌、产气荚膜梭菌。进而查询危害因子数据库获取对应的应急方法，根据标准方案评分模型进行应急方法的评价。最终得到标准的应急方法推荐。其中包括通过危害因子查询关联得到的具体专家。
在一些实施例中，本发明的食品安全突发事件的应急处置推荐方法中步骤100包括：
获取用户输入的突发事件的描述文本；
本步骤中对于用户输入的语音则需要先将语音转换成描述的文字文本。
对突发事件的描述文本进行分析并提取关键特征，以得到用户突发事件的描述信息。
本步骤中可以采用现有的语义分析方法对突发事件的描述文本，再根据分析结果提取关键特征，进而得到用户突发事件的描述信息。如产气荚膜梭菌所 导致的症状：急性胃肠炎型：90％以上的病人以腹胀、腹痛、腹泻等急性胃肠炎为主。腹痛可表现为剧烈腹绞痛，腹泻一般为稀便或水样便，有时会出现黏液便或血便。粪便有腐臭气味，并有大量气体产生，腹泻数次至10余次。常有腹胀，少有恶心、呕吐，一般体温正常或有低热，少有头痛、头晕、浑身无力、酸痛等感染的症状。我们通过分词模型进行症状提取，结果为：
1.腹胀
2.腹痛
3.腹泻
4.恶心
5.呕吐
6.低热
7.头痛
8.头晕
9.浑身无力
10.发冷恶寒
11.虚脱
12.昏迷
的症状列表，将非结构化数据转化为结构化数据。
在一些实施例中，本发明的食品安全突发事件的应急处置推荐方法中步骤100之前还包括：
利用数据探索性分析模型分别对危害因子预测模型、标准评估模型、回溯模型和专家辅助决策模型训练所需的数据集进行标准化处理；
利用数据探索性分析模型分别对危害因子预测模型、标准评估模型、回溯模型和专家辅助决策模型进行变量分析及重要性分析以确定相应的模型输出特征。
具体的，原始数据集是以文档和图片的方式进行提供的，需要首先通过语义分词构建特征分词，进行表单的设计，最后录入MySQL数据库中；对于数据的字段大多数属于字符串类型文本内容，需要再次进行截断分词，构成分词 数组，进而通过Onehot label进行编码标准化；对于模型的输出特征的选择上，需要针对各个特征进行单变量分析、多变量分析以及重要性分析。因而，数据探索性分析模型主要工作是：1)流程优化以及在食品安全数据集上的数据分析报告。将非结构化数据转化为结构化数据，同时对特征进行单变量分析和多变量分析。2)针对危害因子预测模型的数据集进行重要性分析及变量选择，确定模型的训练集和测试集，构建模型以及对模型的训练及优化；3)针对回溯模型的数据集进行重要性分析及变量选择，确定模型的训练集和测试集，构建模型以及对模型的训练及优化。4)针对标准评估模型的数据集进行重要性分析及变量选择，确定模型的训练集和测试集，创建模型以及对模型的训练及优化。6)针对专家辅助决策模型的数据集进行重要性分析及变量选择，确定模型的训练集和测试集，构建模型以及对模型的训练及优化。各模型关联关系具体参见图4所示。
可选地，本发明实施例的食品安全突发事件的应急处置推荐方法中采用随机森林方法创建危害因子预测模型。
可选地，本发明实施例的食品安全突发事件的应急处置推荐方法中采用决策树方法创建回溯模型。
可选地，本发明实施例的食品安全突发事件的应急处置推荐方法中采用多元非线性回归分析方法创建标准评估模型。
可选地，本发明实施例的采用协同过滤推荐方法创建专家辅助决策模型。
需要说明的是，本实施例中危害因子预测模型、回溯模型、标准评估模型以及专家辅助决策模型分别可以采用其他机器学习方法创建，在此不一一例举说明。
图5示出了本发明实施的食品安全突发事件的应急处置推荐方法的数据库设计框架图。本实施例中建设了一个综合数据库，即构建一个区域化、本地化的食品安全突发事件和应急处置标准数据库，并创建MySQL数据库管理系统，实现自有数据集的录入、检索等功能；与食品安全突发事件应急系统其他数据库进行对接。主要内容包括以下：1)生物性危害因子数据库的创建2)放射性危害因子数据库的创建3)化学性危害因子数据库的创建4)突发事件应急案 例数据库(即应急处置案例数据库)的创建5)标准应急方法规范数据库(即标准应急方案数据库)的创建6)专家信息数据库的创建7)综合数据库的创建与查询8)数据库的管理端导入、下载、预览、编辑等功能。
具体环境配置如下：
1、服务端硬件环境：
WEB服务器：CPU 2*E5-2630 V4 2.2GHz 10 cores；128GB内存；4*900GB 10K HDD硬盘；Raid卡。
文件服务器：CPU 2*E5-2630 V4 2.2GHz 10 cores CPU，128GB内存，2*900GB 10K HDD+6*8TB 7.2K HDD硬盘；Raid卡。
2、服务端软件环境：
操作系统：CentOS 7
数据库：MySQL
开发工具包：JDK8
WEB服务器：Tomcat
3、客户端硬件环境：
CPU：P3以上。内存：256M以上。硬盘：20G以上。分辨率：推荐使用1024*768像素
4、客户端软件环境：
操作系统：Microsoft windows2000 professional、Microsoft window2000 server、Microsoft windowXP Professional、Windows 7 Professional任选其一。浏览器：Firefox、Chrome或以上版本。
本实施例从数据段到模型端以及最终的远程可视化端进行了一体化设计，最终的食品安全突发事件的应急处置推荐作为一个app软件供用户进行使用。在用户输入-查询结果以及最终的连线专家远程诊断，整个过程实时在线进行。用户在面对突发事件的过程中，可以通过简单的输出系统所需要的必须特征数据，就能够即使得到对应的危害因子类别、危害因子名称以及响应的应急方法和专家推荐。本实施例能够给用户及时的应急方法推荐，避免不必要的损失。
图6示出了本发明实施例的食品安全突发事件的应急处置推荐方法的模 型构建流程图，危害因子预测模型、回溯模型、标准评估模型以及专家辅助决策模型的构建流程基本相同，大体过程为：首先从相应数据库中提取数据，进行必要的数据填充和数据分割处理，之后针对相应模型的数据集进行重要性分析及变量选择，确定模型的训练集和测试集，然后进行模型开发，待模型开发完成后利用测试集进行优化，调整模型参数，最后评估模型性能，当模型性能满足设计要求后进行在线部署。
本实施例中可使用GridSearchCV进行模型参数调优，使用SHAP可视化工具解释以上机器学习模型，各特征变量对预测结果的影响。
另一方面，参见7所示，本发明实施例还提供了一种食品安全突发事件的应急处置推荐系统1，包括：
事件获取模块10，被配置为获取用户突发事件的描述信息；
危害因子预测模块20，被配置为利用危害因子预测模型将用户突发事件的描述信息与危害因子数据库中数据进行相似性分析得到中毒危害因子匹配度；
危害信息获取模块30，被配置为如果中毒危害因子匹配度大于预设匹配度阈值，则从危害因子数据库中获取初步中毒危害因子信息；
评估分析模块40，被配置为利用标准评估模型将初步中毒危害因子信息与标准应急方案数据库中数据进行评估分析得到标准应急方案的推荐得分；
案例匹配模块50，被配置为如果标准应急方案的推荐得分小于预设得分阈值，则利用回溯模型将用户突发事件的描述信息与应急处置案例数据库中数据进行匹配分析得到第一案例匹配结果；
目标危害信息获取模块60，被配置为如果第一案例匹配结果为成功匹配案例，则根据第一案例匹配结果获取目标中毒危害因子信息和目标案例处置方案，目标中毒危害因子信息包括目标中毒危害因子名称和目标中毒危害因子类别；
以及，专家推荐模块70，被配置为利用专家辅助决策模型将目标中毒危害因子信息与专家信息数据库中数据进行决策分析得到目标专家推荐信息。
上述中食品安全突发事件的应急处置推荐系统各模块的具体细节已经在 对应的食品安全突发事件的应急处置推荐方法中进行了详细的描述，因此此处不再赘述。
再一方面，本发明实施例还提供了一种电子设备，包括：处理器和存储器，存储器上存储有计算机可读指令，计算机可读指令被处理器执行时实现上述实施例所述的食品安全突发事件的应急处置推荐方法。
具体地，上述存储器和处理器能够为通用的存储器和处理器，这里不做具体限定，当处理器运行存储器存储的计算机可读指令时，能够执行上述实施例所述的食品安全突发事件的应急处置推荐方法。
又一方面，本发明实施例还提供了一种计算机可读存储介质，其上存储有计算机程序，计算机程序被处理器执行时实现上述实施例的食品安全突发事件的应急处置推荐方法。
本领域普通技术人员可以理解上述实施例的各种方法中的全部或部分步骤是可以通过程序来指令相关的硬件来完成，该程序可以存储于一计算机可读存储介质中，存储介质可以包括：闪存盘、只读存储器(read-only memory，ROM)、随机存取器(random accessmemory，RAM)、磁盘或光盘等。
应该指出，上述详细说明都是示例性的，旨在对本申请提供进一步的说明。除非另有指明，本文使用的所有技术和科学术语均具有与本申请所属技术领域的普通技术人员的通常理解所相同的含义。
需要注意的是，这里所使用的术语仅是为了描述具体实施方式，而非意图限制根据本申请所述的示例性实施方式。如在这里所使用的，除非上下文另外明确指出，否则单数形式也意图包括复数形式。此外，还应当理解的是，当在本说明书中使用术语“包含”和/或“包括”时，其指明存在特征、步骤、操作、器件、组件和/或它们的组合。
需要说明的是，本申请的说明书和权利要求书及上述附图中的术语“第一”、“第二”等是用于区别类似的对象，而不必用于描述特定的顺序或先后次序。应该理解这样使用的术语在适当情况下可以互换，以便这里描述的本申请的实施方式能够以除了在这里图示或描述的那些以外的顺序实施。
此外，术语“包括”和“具有”以及他们的任何变形，意图在于覆盖不排 他的包含。例如，包含了一系列步骤或单元的过程、方法、系统、产品或设备不必限于清楚地列出的那些步骤或单元，而是可包括没有清楚地列出的或对于这些过程、方法、产品或设备固有的其它步骤或单元。
为了便于描述，在这里可以使用空间相对术语，如“在……之上”、“在……上方”、“在……上表面”、“上面的”等，用来描述如在图中所示的一个器件或特征与其他器件或特征的空间位置关系。应当理解的是，空间相对术语旨在包含除了器件在图中所描述的方位之外的在使用或操作中的不同方位。例如，如果附图中的器件被倒置，则描述为“在其他器件或构造上方”或“在其他器件或构造之上”的器件之后将被定位为“在其他器件或构造下方”或“在其他器件或构造之下”。因而，示例性术语“在……上方”可以包括“在……上方”和“在……下方”两种方位。该器件也可以其他不同方式定位，如旋转90度或处于其他方位，并且对这里所使用的空间相对描述作出相应解释。
在上面详细的说明中，参考了附图，附图形成本文的一部分。在附图中，类似的符号典型地确定类似的部件，除非上下文以其他方式指明。在详细的说明书、附图及权利要求书中所描述的图示说明的实施方案不意味是限制性的。在不脱离本文所呈现的主题的精神或范围下，其他实施方案可以被使用，并且可以作其他改变。
以上所述仅为本发明的优选实施例而已，并不用于限制本发明，对于本领域的技术人员来说，本发明可以有各种更改和变化。凡在本发明的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。

Claims (10)

  1. 一种食品安全突发事件的应急处置推荐方法，其特征在于，包括：
    获取用户突发事件的描述信息；
    利用危害因子预测模型将所述用户突发事件的描述信息与危害因子数据库中数据进行相似性分析得到中毒危害因子匹配度；
    如果所述中毒危害因子匹配度大于预设匹配度阈值，则从所述危害因子数据库中获取初步中毒危害因子信息；
    利用标准评估模型将所述初步中毒危害因子信息与标准应急方案数据库中数据进行评估分析得到标准应急方案的推荐得分；
    如果所述标准应急方案的推荐得分小于预设得分阈值，则利用回溯模型将所述用户突发事件的描述信息与应急处置案例数据库中数据进行匹配分析得到第一案例匹配结果；
    如果所述第一案例匹配结果为成功匹配案例，则根据所述第一案例匹配结果获取目标中毒危害因子信息和目标案例处置方案，所述目标中毒危害因子信息包括目标中毒危害因子名称和目标中毒危害因子类别；
    利用专家辅助决策模型将所述目标中毒危害因子信息与专家信息数据库中数据进行决策分析得到目标专家推荐信息。
  2. 根据权利要求1所述的食品安全突发事件的应急处置推荐方法，其特征在于，所述如果所述中毒危害因子匹配度大于预设匹配度阈值，则从所述危害因子数据库中获取初步中毒危害因子信息的步骤还包括：
    如果所述中毒危害因子匹配度小于预设匹配度阈值，则利用所述回溯模型将所述用户突发事件的描述信息与所述应急处置案例数据库中数据进行匹配分析得到第二案例匹配结果；
    如果所述第二案例匹配结果为未成功匹配案例，则向用户反馈无法提供帮助的提示信息。
  3. 根据权利要求2所述的食品安全突发事件的应急处置推荐方法，其特征在于，所述如果所述标准应急方案的推荐得分低于预设得分阈值，则利用回溯模型将所述用户突发事件的描述信息与应急处置案例数据库中数据进行匹配分析得到第一案例匹配结果的步骤还包括：
    如果所述标准应急方案的推荐得分大于预设得分阈值，则从所述标准应急方案数据库中获得目标标准应急方案；
    利用所述专家辅助决策模型将所述初步中毒危害因子信息与所述专家信息数据库中数据进行决策分析得到目标专家推荐信息。
  4. 根据权利要求3所述的食品安全突发事件的应急处置推荐方法，其特征在于，所述获取用户突发事件的描述信息的步骤包括：
    获取用户输入的突发事件的描述文本；
    对所述突发事件的描述文本进行分析并提取关键特征，以得到所述用户突发事件的描述信息。
  5. 根据权利要求4所述的食品安全突发事件的应急处置推荐方法，其特征在于，所述获取用户突发事件的描述信息的步骤之前还包括：
    利用数据探索性分析模型分别对所述危害因子预测模型、所述标准评估模型、所述回溯模型和所述专家辅助决策模型训练所需的数据集进行标准化处理；
    利用数据探索性分析模型分别对所述危害因子预测模型、所述标准评估模型、所述回溯模型和所述专家辅助决策模型进行变量分析及重要性分析以确定相应的模型输出特征。
  6. 根据权利要求5所述的食品安全突发事件的应急处置推荐方法，其特征在于，采用多元非线性回归分析方法创建所述标准评估模型。
  7. 根据权利要求5所述的食品安全突发事件的应急处置推荐方法，其特征在于，采用协同过滤推荐方法创建所述专家辅助决策模型。
  8. 一种食品安全突发事件的应急处置推荐系统，其特征在于，包括：
    事件获取模块，被配置为获取用户突发事件的描述信息；
    危害因子预测模块，被配置为利用危害因子预测模型将所述用户突发事件的描述信息与危害因子数据库中数据进行相似性分析得到中毒危害因子匹配度；
    危害信息获取模块，被配置为如果所述中毒危害因子匹配度大于预设匹配度阈值，则从所述危害因子数据库中获取初步中毒危害因子信息；
    评估分析模块，被配置为利用标准评估模型将所述初步中毒危害因子信息 与标准应急方案数据库中数据进行评估分析得到标准应急方案的推荐得分；
    案例匹配模块，被配置为如果所述标准应急方案的推荐得分小于预设得分阈值，则利用回溯模型将所述用户突发事件的描述信息与应急处置案例数据库中数据进行匹配分析得到第一案例匹配结果；
    目标危害信息获取模块，被配置为如果所述第一案例匹配结果为成功匹配案例，则根据所述第一案例匹配结果获取目标中毒危害因子信息和目标案例处置方案，所述目标中毒危害因子信息包括目标中毒危害因子名称和目标中毒危害因子类别；
    以及，专家推荐模块，被配置为利用专家辅助决策模型将所述目标中毒危害因子信息与专家信息数据库中数据进行决策分析得到目标专家推荐信息。
  9. 一种电子设备，其特征在于，包括：处理器和存储器，所述存储器上存储有计算机可读指令，所述计算机可读指令被所述处理器执行时实现如权利要求1至7中任一项所述的食品安全突发事件的应急处置推荐方法。
  10. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现如权利要求1至7中任一项所述的食品安全突发事件的应急处置推荐方法。
PCT/CN2022/131071 2021-11-30 2022-11-10 食品安全突发事件的应急处置推荐方法及系统 WO2023098445A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111473542.7A CN113886716B (zh) 2021-11-30 2021-11-30 食品安全突发事件的应急处置推荐方法及系统
CN202111473542.7 2021-11-30

Publications (1)

Publication Number Publication Date
WO2023098445A1 true WO2023098445A1 (zh) 2023-06-08

Family

ID=79016334

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/131071 WO2023098445A1 (zh) 2021-11-30 2022-11-10 食品安全突发事件的应急处置推荐方法及系统

Country Status (2)

Country Link
CN (1) CN113886716B (zh)
WO (1) WO2023098445A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116992160A (zh) * 2023-09-27 2023-11-03 长春中医药大学 一种用于卫生事件的临床护理训练应急预案推荐方法
CN117350288A (zh) * 2023-12-01 2024-01-05 浙商银行股份有限公司 基于案例匹配的网络安全运营辅助决策方法、系统及装置

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113886716B (zh) * 2021-11-30 2022-04-05 国家食品安全风险评估中心 食品安全突发事件的应急处置推荐方法及系统

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004341983A (ja) * 2003-05-19 2004-12-02 Nec Fielding Ltd 住民の健康管理システム，方法，自動販売機およびプログラム
CN104408569A (zh) * 2014-11-28 2015-03-11 东莞中国科学院云计算产业技术创新与育成中心 一种基于预案的多目标辅助决策平台实施方法
CN106202514A (zh) * 2016-07-21 2016-12-07 北京邮电大学 基于Agent的突发事件跨媒体信息的检索方法及系统
CN106504161A (zh) * 2016-10-13 2017-03-15 广州市莱曼信息技术有限公司 一种动物卫生风险管理方法和系统
US20210035679A1 (en) * 2019-07-30 2021-02-04 Experian Health, Inc. Social determinants of health solution
CN113886716A (zh) * 2021-11-30 2022-01-04 国家食品安全风险评估中心 食品安全突发事件的应急处置推荐方法及系统

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CA2541976A1 (en) * 2002-10-08 2004-04-22 Food Security Systems, L.L.C. System and method for identifying a food event, tracking the food product, and assessing risks and costs associated with intervention
KR101499706B1 (ko) * 2012-11-30 2015-03-18 대한민국 온라인을 통한 식품안전 관리 시스템 및 방법
WO2019222471A1 (en) * 2018-05-17 2019-11-21 Ecolab Usa Inc. Food safety risk and sanitation compliance tracking
CN110889635B (zh) * 2019-11-29 2022-10-04 北京金和网络股份有限公司 一种用于对食品安全事件的处理进行应急演练的方法
CN113032493A (zh) * 2019-12-05 2021-06-25 天津科技大学 一种食品安全应急处置流程知识图谱构建方法
CN111738549A (zh) * 2020-05-21 2020-10-02 平安国际智慧城市科技股份有限公司 食品安全风险评估方法、装置、设备及存储介质
CN113052441B (zh) * 2021-03-09 2021-10-08 中国安全生产科学研究院 一种应急食品供应风险分析方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004341983A (ja) * 2003-05-19 2004-12-02 Nec Fielding Ltd 住民の健康管理システム，方法，自動販売機およびプログラム
CN104408569A (zh) * 2014-11-28 2015-03-11 东莞中国科学院云计算产业技术创新与育成中心 一种基于预案的多目标辅助决策平台实施方法
CN106202514A (zh) * 2016-07-21 2016-12-07 北京邮电大学 基于Agent的突发事件跨媒体信息的检索方法及系统
CN106504161A (zh) * 2016-10-13 2017-03-15 广州市莱曼信息技术有限公司 一种动物卫生风险管理方法和系统
US20210035679A1 (en) * 2019-07-30 2021-02-04 Experian Health, Inc. Social determinants of health solution
CN113886716A (zh) * 2021-11-30 2022-01-04 国家食品安全风险评估中心 食品安全突发事件的应急处置推荐方法及系统

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116992160A (zh) * 2023-09-27 2023-11-03 长春中医药大学 一种用于卫生事件的临床护理训练应急预案推荐方法
CN116992160B (zh) * 2023-09-27 2023-12-12 长春中医药大学 一种用于卫生事件的临床护理训练应急预案推荐方法
CN117350288A (zh) * 2023-12-01 2024-01-05 浙商银行股份有限公司 基于案例匹配的网络安全运营辅助决策方法、系统及装置
CN117350288B (zh) * 2023-12-01 2024-05-03 浙商银行股份有限公司 基于案例匹配的网络安全运营辅助决策方法、系统及装置

Also Published As

Publication number Publication date
CN113886716A (zh) 2022-01-04
CN113886716B (zh) 2022-04-05

Similar Documents

Publication Publication Date Title
WO2023098445A1 (zh) 食品安全突发事件的应急处置推荐方法及系统
CN110765257B (zh) 一种知识图谱驱动型的法律智能咨询系统
US10818397B2 (en) Clinical content analytics engine
US11810671B2 (en) System and method for providing health information
CN112131393B (zh) 一种基于bert和相似度算法的医疗知识图谱问答系统构建方法
CN112786194A (zh) 基于人工智能的医学影像导诊导检系统、方法及设备
WO2021139232A1 (zh) 基于医疗知识图谱的分诊方法、装置、设备及存储介质
CN113871003A (zh) 一种基于含因果性医学知识图谱的疾病辅助鉴别诊断系统
US11900063B2 (en) System and method for actionizing comments
CN112148851A (zh) 一种基于知识图谱的医药知识问答系统的构建方法
CN104699730A (zh) 用于识别候选答案之间的关系的方法和系统
US10754882B2 (en) Method of retrieving information from a health report through a machine assisted interrogation process
CN111191048A (zh) 基于知识图谱的急诊问答系统构建方法
CN113569023A (zh) 一种基于知识图谱的中文医药问答系统及方法
CN115394393A (zh) 智能诊疗数据处理方法、装置、电子设备及存储介质
KR101983383B1 (ko) 의료비 지급심사를 위해 보험약관이 적용된 룰 기반 유사질의답변 연계 의료정보 서비스 제공 방법
Bossema et al. Expert quotes and exaggeration in health news: a retrospective quantitative content analysis
US11170172B1 (en) System and method for actionizing comments
Tcheng et al. Achieving Data Liquidity: Lessons Learned from Analysis of 38 Clinical Registries (The Duke-Pew Data Interoperability Project
US20230072155A1 (en) Method and system for incorporating patient information
US20230052022A1 (en) Systems and Methods for Dynamic Charting
Chudasama et al. InterpretME: A tool for interpretations of machine learning models over knowledge graphs
Baghal et al. Agile natural language processing model for pathology knowledge extraction and integration with clinical enterprise data warehouse
TWI773936B (zh) 能根據語音資訊進行內容分析以推薦專家的方法
US20240095455A1 (en) Systems and methods for question-answering using a multi-modal end to end learning system

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22900250

Country of ref document: EP

Kind code of ref document: A1