Emergency response method and related equipment

Abstract

The invention discloses an emergency response method and related equipment, and relates to the technical field of data processing, and the emergency response method comprises the steps: capturing global public health data in real time based on a distributed crawler cluster of a Scrapy-Redis framework, processing the global public health data by using a preset BERT model and a preset YOLO model to extract key information and divide risk levels, and based on the key information and the risk level, triggering a differentiated alarm. According to the public health emergency response method and device, the multi-modal data are quickly captured based on the distributed crawler cluster of the Scrapy-Redis framework, and the captured multi-modal data are quickly and accurately processed in combination with the preset BERT model and the preset YOLO model, so that the efficiency of public health emergency response is effectively improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to emergency response methods and related equipment. Background Art

[0002] As the complexity and uncertainty of global public health events increase, the rapid collection and emergency response of public health information has become a key challenge in public health management. Traditional public health information management methods have many problems, such as scattered information, diverse data sources that are difficult to integrate, and untimely information updates. These problems make it impossible to quickly and accurately summarize and analyze public health information, thus affecting the efficiency and effectiveness of public health emergency response.

[0003] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are related technologies. Summary of the invention

[0004] The main purpose of this application is to provide an emergency response method and related equipment, aiming to solve the technical problem of how to improve the efficiency of public health emergency response.

[0005] To achieve the above objectives, the present application proposes an emergency response method, which includes: A distributed crawler cluster based on the Scrapy-Redis framework to capture global public health data in real time; Using a preset BERT model and a preset YOLO model to process the global public health data to extract key information and classify risk levels; Based on the key information and the risk level, a differentiated alarm is triggered.

[0006] In one embodiment, the step of processing the global public health data using a preset BERT model and a preset YOLO model to extract key information and classify risk levels further includes: Use the preset BERT model to identify and classify the text data in the global public health data, and output predefined category labels; Use a preset YOLO model to locate and classify the image data in the global public health data, and output a recognition result with a confidence level higher than a preset confidence threshold; Based on the predefined category labels and the recognition results, key information is extracted and risk levels are divided.

[0007] In one embodiment, the step of crawling global public health data in real time using a distributed crawler cluster based on the Scrapy-Redis framework further includes: In response to the crawling instruction, the master node is called to assign crawling tasks to each working node; Call the working nodes to execute the crawling tasks in parallel, and perform redundant data filtering operations based on the SimHash algorithm and local sensitive hashing to obtain global public health data In one embodiment, the step of calling the master node to assign crawling tasks to each working node in response to the crawling instruction further includes: Calculate the PageRank score of each data source in the global public health data source, and calculate the timeliness weight of each set of data in each data source; Calculate URL priority based on the PageRank score and the timeliness weight; Based on the priority, the master node is called to assign high priority URLs to the worker nodes.

[0008] In one embodiment, before the step of using a preset BERT model and a preset YOLO model to process the global public health data to extract key information and classify risk levels, the step further includes: Based on SimHash and local sensitive hashing, the global public health data is quickly hashed to determine whether there is similar or repeated content; If similar or repeated content exists, a deduplication operation is performed on the global public health data to obtain deduplicated global public health data.

[0009] In one embodiment, before the step of using a preset BERT model and a preset YOLO model to process the global public health data to extract key information and classify risk levels, the step further includes: Acquire sample data, wherein the processing result corresponding to the sample data is a first processing result; Using the current BERT model to perform semantic classification on the sample data, and using the current YOLO model to perform target recognition on the sample data, to obtain a second processing result; Determining whether the first processing result is consistent with the second processing result; If they are inconsistent, adjust the parameters of the current BERT model and the current YOLO model, and based on the current BERT model and the current YOLO model after the adjustment of the parameters, return to the step of using the current BERT model to perform semantic classification on the sample data, and using the current YOLO model to perform target recognition on the sample data to obtain a second processing result, until the first processing result is consistent with the second processing result, and obtain the preset extraction model.

[0010] In addition, to achieve the above objectives, the present application also proposes an emergency response device, which includes: A crawling module, which is used for a distributed crawler cluster based on the Scrapy-Redis framework to crawl global public health data in real time; A data processing module, the data processing module is used to process the global public health data using a preset BERT model and a preset YOLO model to extract key information and classify risk levels; A trigger module is used to trigger a differentiated alarm based on the key information and the risk level.

[0011] In one embodiment, the data processing module includes: A first data processing unit, configured to use a preset BERT model to identify and classify the risks of text data in the global public health data, and output a predefined category label; A second data processing unit is used to use a preset YOLO model to perform target positioning and classification on the image data in the global public health data, and output a recognition result with a confidence level higher than a preset confidence threshold; The second data processing unit is used to extract key information and classify risk levels based on the predefined category label and the recognition result.

[0012] In one embodiment, the capture module includes: A first calling unit, configured to call the master node to assign a crawling task to each working node in response to a crawling instruction; The crawling unit is used to call the working nodes to execute the crawling tasks in parallel, and to perform redundant data filtering operations based on the SimHash algorithm and local sensitive hashing to obtain global public health data.

[0013] In one embodiment, the capture module further includes: The first calculation unit is used to calculate the PageRank score of each data source in the global public health data source, and calculate the timeliness weight of each group of data in each data source; A second calculation unit, configured to calculate a URL priority based on the PageRank score and the timeliness weight; The second calling unit is used to call the main node to allocate the high priority URL to the working node based on the priority.

[0014] In one embodiment, the emergency response further includes a deduplication module, and the deduplication module includes: A first judgment unit, configured to perform fast hash processing on the global public health data based on SimHash and local sensitive hashing to determine whether there is similar or repeated content; A deduplication unit, configured to perform a deduplication operation on the global public health data if there is similar or duplicate content, to obtain deduplicated global public health data.

[0015] In one embodiment, the emergency response further includes a training module, and the training module includes: An acquisition unit, configured to acquire sample data, and a processing result corresponding to the sample data is a first processing result; A third data processing unit, configured to perform semantic classification on the sample data using a current BERT model, and perform object recognition on the sample data using a current YOLO model, to obtain a second processing result; A second determination unit, configured to determine whether the first processing result is consistent with the second processing result; A training unit, configured to, if they are inconsistent, adjust parameters of the current BERT model and the current YOLO model, and based on the current BERT model and the current YOLO model with adjusted parameters, return to the step of performing semantic classification on the sample data using the current BERT model and performing object recognition on the sample data using the current YOLO model to obtain a second processing result, until the first processing result is consistent with the second processing result, to obtain a preset extraction model.

[0016] In addition, to achieve the above object, the present application further provides an emergency response device, and the device includes: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the computer program is configured to implement the steps of the emergency response method as described above.

[0017] In addition, to achieve the above object, the present application further provides a storage medium, the storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium, and when the computer program is executed by a processor, the steps of the emergency response method as described above are implemented.

[0018] In addition, to achieve the above object, the present application further provides a computer program product, the computer program product includes a computer program, and when the computer program is executed by a processor, the steps of the emergency response method as described above are implemented.

[0019] One or more technical solutions proposed by the present application have at least the following technical effects: The present application proposes an emergency response method and related equipment, which relate to the field of data processing technology. Compared with the related technology, public health information cannot be quickly and accurately summarized and analyzed, thereby affecting the efficiency and effectiveness of public health emergency response. In the present application, firstly, a distributed crawler cluster based on the Scrapy-Redis framework is used to crawl global public health data in real time, and then, the global public health data is processed using a preset BERT model and a preset YOLO model to extract key information and divide risk levels. Finally, based on the key information and the risk levels, differentiated alarms are triggered.

[0020] It can be understood that this application uses a distributed crawler cluster based on the Scrapy-Redis framework to quickly capture multimodal data, and combines the preset BERT model and the preset YOLO model to quickly and accurately process the captured multimodal data, effectively improving the efficiency of public health emergency response. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0022] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0023] Figure 1 A flowchart of the first embodiment of the emergency response method of this application is provided; Figure 2 A flowchart of the second embodiment of the emergency response method of this application is provided; Figure 3 A flowchart of the third embodiment of the emergency response method of this application is provided; Figure 4 This is a schematic diagram of the module structure of the emergency response device according to an embodiment of the present application; Figure 5 This is a schematic diagram of the device structure of the hardware operating environment involved in the emergency response method in the embodiment of the present application.

[0024] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0025] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0026] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0027] The main solutions of the embodiments of this application are: In this embodiment, for ease of description, the emergency response device is used as the execution entity for explanation below.

[0028] Due to existing technologies: Traditional public health information management methods have many problems, such as scattered information, diverse and difficult to integrate data sources, and untimely information updates. These problems make it impossible to quickly and accurately summarize and analyze public health information, thus affecting the efficiency and effectiveness of public health emergency responses.

[0029] The present application provides a solution that enables: a distributed crawler cluster based on the Scrapy-Redis framework to capture global public health data in real time, and uses a preset BERT model and a preset YOLO model to process the global public health data to extract key information and divide risk levels, and trigger differentiated alarms based on the key information and the risk levels. The present application uses a distributed crawler cluster based on the Scrapy-Redis framework to quickly capture multimodal data, and combines a preset BERT model and a preset YOLO model to quickly and accurately process the captured multimodal data, effectively improving the efficiency of public health emergency response.

[0030] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, an emergency response device, etc. The following takes the emergency response device as an example to illustrate this embodiment and the following embodiments.

[0031] Based on this, the embodiment of the present application provides an emergency response method, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the emergency response method of the present application.

[0032] In this embodiment, the emergency response method includes steps S100 to S300: Step S100, using a distributed crawler cluster based on the Scrapy-Redis framework to crawl global public health data in real time; It should be noted that Scrapy-Redis is an extension based on the Scrapy framework, designed for distributed crawlers. It uses Redis as a message queue to implement the storage of request queues and deduplication sets, thereby supporting distributed crawling of multiple crawler instances.

[0033] It should be noted that a distributed crawler means that multiple crawler instances share a Redis request queue, which is suitable for large-scale multi-domain crawling.

[0034] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, an emergency response device, etc. The following takes the emergency response device as an example to illustrate this embodiment and the following embodiments.

[0035] Specifically, the step of crawling global public health data in real time using a distributed crawler cluster based on the Scrapy-Redis framework also includes steps S110 to S120: Step S110, in response to the crawling instruction, calling the master node to assign crawling tasks to each working node; It should be noted that the master node is responsible for receiving crawling instructions, assigning tasks to each worker node according to the priority and complexity of the crawling tasks, monitoring the status of the worker nodes, ensuring the smooth execution of tasks, and storing task queues and deduplication sets (using Redis).

[0036] It should be noted that the Worker Node is responsible for obtaining the assigned tasks from the Master Node, executing specific crawling tasks, storing the results in the specified storage system (such as a database or file system), and reporting the task execution status (success, failure, exception, etc.) to the Master Node.

[0037] Specifically, the step of calling the master node to assign crawling tasks to each working node in response to the crawling instruction also includes steps S111 to S113: Step S111, calculating the PageRank score of each data source in the global public health data source, and calculating the timeliness weight of each group of data in each data source; It should be noted that the PageRank algorithm is used to evaluate the importance of each data source in the global public health data source. The specific formula is as follows: PR(A) = (1-d) + d * (PR(T1) / C(T1) + ... + PR(Tn) / C(Tn)) Where d is the damping coefficient (0.85), PR(Ti) is the page weight linked to page A, and C(Ti) is the number of outbound links from page Ti.

[0038] It should be noted that the newer the information, the higher the timeliness weight.

[0039] Step S112, calculating the URL priority based on the PageRank score and the timeliness weight; It should be noted that the formula for calculating URL priority is as follows: Priority = PageRank score * timeliness weight * information importance coefficient Step S113: Based on the priority, the master node is called to allocate the high priority URL to the working node.

[0040] It can be understood that through the above steps, the PageRank scores and timeliness weights of each data source in the global public health data source can be effectively calculated, and based on these indicators, the crawling tasks can be reasonably allocated to improve the efficiency and quality of data crawling.

[0041] Step S120, calling the working nodes to execute the crawling tasks in parallel, and performing redundant data filtering operations based on the SimHash algorithm and local sensitive hashing to obtain global public health data; In this embodiment, first, the master node assigns high-priority URLs to the worker nodes. Each worker node receives the assigned URL list and prepares to perform parallel crawling. Then, the worker node further splits the task into multiple subtasks according to the assigned URL list, and each subtask corresponds to the crawling of a URL.

[0042] It should be noted that the worker node uses multi-threading or multi-process technology to initiate multiple HTTP requests at the same time to crawl the assigned URL. Each worker node can set a thread pool or process pool to dynamically adjust the number of threads or processes based on system resources and task volume. Use thread-safe queues (such as Python's Queue module) to manage the URLs to be crawled to ensure efficient allocation and processing of tasks.

[0043] Furthermore, the emergency response device calculates the data fingerprint based on the SimHash algorithm: First, the emergency response device preprocesses the captured web page content, including removing HTML tags, word segmentation, and stop words. Then, the preprocessed text is converted into a feature vector, such as using the TF-IDF or bag-of-words model. Further, the feature vector is hashed using the SimHash algorithm to generate a 64-bit (or longer) fingerprint.

[0044] It should be noted that the SimHash algorithm maps high-dimensional feature vectors into fingerprints of fixed length through random projection and bit operations while retaining the similarity of the data.

[0045] Finally, the emergency response device filters redundant data through locality sensitive hashing (LSH). The specific steps are as follows: First, the fingerprint generated by SimHash is processed by the LSH algorithm to build a hash table. Then, based on the LSH algorithm, similar fingerprints are mapped to the same bucket with a higher probability, so as to achieve fast search and filtering. Furthermore, for each captured data, its SimHash fingerprint is calculated, and the hash table is searched by the LSH algorithm to determine whether there is similar data. If similar data is found in the hash table, the data is considered redundant and can be discarded or marked. If no similar data is found, the SimHash fingerprint of the data is inserted into the hash table. After LSH filtering, the non-redundant data is retained as global public health data, and the global public health data is stored in a database or file system for subsequent analysis and processing.

[0046] It can be understood that through the above steps, the working nodes can efficiently execute the crawling tasks in parallel and use SimHash and LSH algorithms to filter redundant data, and finally obtain high-quality global public health data.

[0047] Step S200, using a preset BERT model and a preset YOLO model to process the global public health data to extract key information and classify risk levels; Specifically, the step of using the preset BERT model and the preset YOLO model to process the global public health data to extract key information and classify risk levels also includes steps S210 to S230: Step S210, using a preset BERT model to identify and classify the risks of text data in the global public health data, and output predefined category labels; Step S220, using a preset YOLO model to perform target positioning and classification on the image data in the global public health data, and outputting a recognition result with a confidence level higher than a preset confidence threshold; Step S230: extract key information and classify risk levels based on the predefined category labels and the recognition results.

[0048] It should be noted that text analysis mainly relies on the BERT (Bidirectional EncoderRepresentations from Transformers) model. Through the bidirectional transformer structure, this model can deeply understand the semantics of text context. In the field of public health, BERT has been fine-tuned to identify specific text features such as medical terms and risk levels. The specific process includes: first, text vectorization (converting text into high-dimensional semantic space vectors), then, semantic understanding (identifying key entities, event types, and risk levels in the text), and further, classification output (dividing the text into predefined categories, such as routine reports, research progress, etc.) It should be noted that the image analysis uses the YOLO (You Only Look Once) target detection model, which is specifically used to quickly and accurately identify key information in medical images. In the field of public health, YOLO can identify: medical statistical charts, pathogen microscopic images, etc. The model completes target positioning and classification through a single forward propagation, greatly improving the efficiency of image analysis. In addition, the recognition process includes target box positioning, category prediction, and confidence assessment.

[0049] In addition, it should be noted that considering that the captured data is not of the same voice, it needs to be translated. Therefore, the emergency response device also integrates Meta's NLLB-200 model, supports translation between 200+ languages, and optimizes the uncommon terminology library (such as medical terms) through user feedback.

[0050] It should be noted that in the multilingual translation system, first of all, a multi-level user feedback channel is built. Furthermore, an intuitive feedback entry is set up in the translation result interface, including an accuracy scoring button and a term correction shortcut. For users who want to provide more detailed opinions, a professional feedback form is provided. This form not only collects the original term and the suggested correct translation, but also asks the user's professional background to evaluate the credibility of the feedback. The collected feedback is not simply entered into the system directly, but is strictly structured. Each piece of feedback is converted into a data object containing multiple dimensions, such as original term, source language, target language, incorrect translation, correct translation, feedback type, professional field, etc. The system will check whether the feedback of multiple users on the same term is consistent. Secondly, key or controversial terms will be submitted to the expert database for review. Through statistical significance tests, we can determine whether it is really necessary to update the terminology database.

[0051] In the model optimization phase, we use a small sample incremental learning method. The core of this method is to fine-tune the local parameters of terminology in a specific field while retaining the basic capabilities of the pre-trained model. This fine-tuning strategy not only avoids damaging the overall capabilities of the model, but also continuously improves the translation accuracy in specific fields.

[0052] Step S300: triggering a differentiated alarm based on the key information and the risk level.

[0053] It should be noted that before triggering a differentiated alert, a risk prediction model will be used for prediction: LSTM-based time series analysis is used to predict the spread trend of public health events, and differentiated alerts are triggered based on multiple factors.

[0054] In this embodiment, risks are divided into different levels according to the results of key information extraction and risk assessment. Common risk level classification methods include: low risk (Low Risk: low risk, conventional monitoring measures can be taken), medium risk (Medium Risk: medium risk, further attention and analysis are required), high risk (High Risk: high risk, emergency measures need to be taken), and very high risk (Very High Risk: extremely high risk, immediate action is required).

[0055] Furthermore, differentiated alarm triggering mechanisms are designed according to different risk levels: If it is a low-risk alert, the notification method is: sent to relevant monitoring personnel via email or system notification. The notification content provides risk warnings and recommended follow-up monitoring measures. The response time is not urgent and can be handled according to normal procedures.

[0056] If it is a medium-risk alert, the notification method is to send it to the relevant person in charge via SMS or instant messaging tools. The notification content provides risk details and recommended initial response measures. The response time is to recommend an initial response within the specified time.

[0057] If it is a high-risk alert, the notification method is to send it to key decision-makers via phone or emergency text message. The notification content is a detailed risk analysis and emergency response measures, and the response time requires immediate response and measures.

[0058] If it is an extremely high risk alert, the notification method is to send it to all relevant personnel through multiple channels (phone, text message, email). The content of the notification provides an explanation of the severity of the risk and an immediate action plan. The response time is to require a response and measures to be taken in the shortest possible time.

[0059] The present application proposes an emergency response method and related equipment, which relate to the field of data processing technology. Compared with the related technology, public health information cannot be quickly and accurately summarized and analyzed, thereby affecting the efficiency and effectiveness of public health emergency response. In the present application, firstly, a distributed crawler cluster based on the Scrapy-Redis framework is used to crawl global public health data in real time, and then, the global public health data is processed using a preset BERT model and a preset YOLO model to extract key information and divide risk levels. Finally, based on the key information and the risk levels, differentiated alarms are triggered.

[0060] It can be understood that this application uses a distributed crawler cluster based on the Scrapy-Redis framework to quickly capture multimodal data, and combines the preset BERT model and the preset YOLO model to quickly and accurately process the captured multimodal data, effectively improving the efficiency of public health emergency response.

[0061] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can refer to the above introduction, and will not be repeated later. Figure 2 Before step S200, the emergency response method further includes steps A10 to A20: Step A10, based on SimHash and local sensitive hashing, fast hash processing is performed on the global public health data to determine whether there is similar or repeated content; Step A20: If similar or repeated content exists, a deduplication operation is performed on the global public health data to obtain deduplicated global public health data.

[0062] Based on the first and second embodiments of the present application, in the third embodiment of the present application, the same or similar contents as those of the first and second embodiments can be referred to the above description, and will not be described in detail later. Figure 3 Before step S200, the emergency response method further includes steps B10 to B40: Step B10, obtaining sample data, wherein the processing result corresponding to the sample data is the first processing result; It should be noted that sample data is the basic data used to train and validate the model. This data may include: text data (such as news articles, research papers, etc.) and image data.

[0063] It should be noted that the first processing result refers to the annotation information of the sample data, that is, the manually annotated semantic classification and object recognition results. These annotation information are used as "standard answers" for subsequent model training and verification.

[0064] Step B20, using the current BERT model to perform semantic classification on the sample data, and using the current YOLO model to perform target recognition on the sample data, to obtain a second processing result; It should be noted that the second processing result refers to the output results of the current BERT model and the current YOLO model.

[0065] Step B30, determining whether the first processing result is consistent with the second processing result; Step B40, if they are inconsistent, adjust the parameters of the current BERT model and the current YOLO model, and based on the current BERT model and the current YOLO model after the adjustment of the parameters, return to the step of using the current BERT model to perform semantic classification on the sample data, and using the current YOLO model to perform target recognition on the sample data to obtain a second processing result, until the first processing result is consistent with the second processing result, and obtain the preset extraction model.

[0066] It should be noted that through consistency judgment (comparing the semantic classification results of the BERT model and the target recognition results of the YOLO model with the first processing results of manual annotation), if they are consistent, it means that the model performs well on the current sample; if they are inconsistent, it means that the model needs further optimization.

[0067] Furthermore, the performance of the BERT model can be optimized by fine-tuning. If the object recognition result of the YOLO model is inconsistent with the annotation result, adjust the parameters of the YOLO model. The performance of the YOLO model can be optimized by adjusting the learning rate, optimizer, regularization parameters, etc.

[0068] After adjusting the parameters, the sample data is reprocessed to obtain a new second processing result. Steps B20 and B30 are repeated until the first processing result is consistent with the second processing result. When the model can accurately process all sample data, these optimized models are considered to be "preset extraction models" and can be used for practical applications.

[0069] It can be understood that through the above steps, semantic classification and target recognition of global public health data can be achieved, and by continuously adjusting the model parameters until the output of the model is consistent with the results of manual annotation. The optimized BERT model and YOLO model can be used for actual public health data processing tasks.

[0070] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the emergency response method of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0071] This application also provides an emergency response device, please refer to Figure 4 , the emergency response device comprises: A crawling module 10, which is used for a distributed crawler cluster based on the Scrapy-Redis framework to crawl global public health data in real time; A data processing module 20, which is used to process the global public health data using a preset BERT model and a preset YOLO model to extract key information and classify risk levels; The trigger module 30 is used to trigger a differentiated alarm based on the key information and the risk level.

[0072] In one embodiment, the data processing module includes: A first data processing unit, configured to use a preset BERT model to identify and classify the risks of text data in the global public health data, and output a predefined category label; A second data processing unit is used to use a preset YOLO model to perform target positioning and classification on the image data in the global public health data, and output a recognition result with a confidence level higher than a preset confidence threshold; The second data processing unit is used to extract key information and classify risk levels based on the predefined category label and the recognition result.

[0073] In one embodiment, the capture module includes: A first calling unit, configured to call the master node to assign a crawling task to each working node in response to a crawling instruction; The crawling unit is used to call the working nodes to execute the crawling tasks in parallel, and to perform redundant data filtering operations based on the SimHash algorithm and local sensitive hashing to obtain global public health data.

[0074] In one embodiment, the capture module further includes: The first calculation unit is used to calculate the PageRank score of each data source in the global public health data source, and calculate the timeliness weight of each group of data in each data source; A second calculation unit, configured to calculate a URL priority based on the PageRank score and the timeliness weight; The second calling unit is used to call the main node to allocate the high priority URL to the working node based on the priority.

[0075] In one embodiment, the emergency response further includes a deduplication module, and the deduplication module includes: A first judgment unit, configured to perform fast hash processing on the global public health data based on SimHash and local sensitive hashing to determine whether there is similar or repeated content; The deduplication unit is used to perform a deduplication operation on the global public health data if similar or repeated content exists, so as to obtain the deduplicated global public health data.

[0076] In one embodiment, the emergency response further includes a training module, and the training module includes: An acquisition unit, used for acquiring sample data, wherein the processing result corresponding to the sample data is a first processing result; A third data processing unit is used to perform semantic classification on the sample data using the current BERT model, and to perform target recognition on the sample data using the current YOLO model to obtain a second processing result; A second judging unit, configured to judge whether the first processing result is consistent with the second processing result; The training unit is used to adjust the parameters of the current BERT model and the current YOLO model if they are inconsistent, and based on the current BERT model and the current YOLO model after adjusting the parameters, return to the step of using the current BERT model to semantically classify the sample data, and using the current YOLO model to perform target recognition on the sample data to obtain a second processing result, until the first processing result is consistent with the second processing result, and a preset extraction model is obtained. The emergency response device provided in the present application adopts the emergency response method in the above-mentioned embodiment to solve the technical problems of emergency response. Compared with the prior art, the beneficial effects of the emergency response device provided in the present application are the same as the beneficial effects of the emergency response method provided in the above-mentioned embodiment, and other technical features in the emergency response device are the same as the features disclosed in the above-mentioned embodiment method, which will not be repeated here.

[0077] The present application provides an emergency response device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the emergency response method in the above-mentioned embodiment one.

[0078] Reference below Figure 5 , which shows a schematic diagram of the structure of an emergency response device suitable for implementing the embodiment of the present application. The emergency response device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The emergency response equipment shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0079] like Figure 5As shown, the emergency response equipment may include a processing device 1001 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the emergency response equipment are also stored. The processing device 1001, ROM1002, and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the emergency response device to communicate with other devices wirelessly or by wire to exchange data. Although the emergency response device with various systems is shown in the figure, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0080] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0081] The emergency response device provided by the present application adopts the emergency response method in the above embodiment to solve the technical problems. Compared with the prior art, the beneficial effects of the emergency response device provided by the present application are the same as the beneficial effects of the emergency response method provided by the above embodiment, and the other technical features in the emergency response device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0082] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0083] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0084] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the emergency response method in the above-mentioned embodiment.

[0085] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM: Random Access Memory), a read-only memory (ROM: Read Only Memory), an erasable programmable read-only memory (EPROM: Erasable Programmable Read Only Memory or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM: CD-Read Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency: Radio Frequency), etc., or any suitable combination of the above.

[0086] The computer-readable storage medium may be included in the emergency response device; or may exist independently without being installed in the emergency response device.

[0087] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the emergency response device, the emergency response device: A distributed crawler cluster based on the Scrapy-Redis framework to capture global public health data in real time; Using a preset BERT model and a preset YOLO model to process the global public health data to extract key information and classify risk levels; Based on the key information and the risk level, a differentiated alarm is triggered.

[0088] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0089] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0090] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0091] The readable storage medium provided in this application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned emergency response method, and can solve the technical problems of emergency response. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the emergency response method provided in the above-mentioned embodiment, and will not be repeated here.

[0092] The present application also provides a computer program product, including a computer program, which implements the steps of the emergency response method as described above when executed by a processor.

[0093] The computer program product provided by this application can solve the technical problems of emergency response. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as the beneficial effects of the emergency response method provided by the above embodiment, which will not be repeated here.

[0094] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. An emergency response method, characterized in that: The emergency response method includes: A distributed crawler cluster based on the Scrapy-Redis framework to capture global public health data in real time; Using a preset BERT model and a preset YOLO model to process the global public health data to extract key information and classify risk levels; Based on the key information and the risk level, a differentiated alarm is triggered.

2. The emergency response method according to claim 1, characterized in that: The step of using a preset BERT model and a preset YOLO model to process the global public health data to extract key information and classify risk levels also includes: Use the preset BERT model to identify and classify the text data in the global public health data, and output predefined category labels; Use a preset YOLO model to locate and classify the image data in the global public health data, and output a recognition result with a confidence level higher than a preset confidence threshold; Based on the predefined category labels and the recognition results, key information is extracted and risk levels are divided.

3. The emergency response method according to claim 1, characterized in that: The distributed crawler cluster based on the Scrapy-Redis framework, the step of crawling global public health data in real time, also includes: In response to the crawling instruction, the master node is called to assign crawling tasks to each working node; The working nodes are called to execute the crawling tasks in parallel, and redundant data filtering operations are performed based on the SimHash algorithm and local sensitive hashing to obtain global public health data.

4. The emergency response method according to claim 3, characterized in that: The step of calling the master node to assign crawling tasks to each working node in response to the crawling instruction also includes: Calculate the PageRank score of each data source in the global public health data source, and calculate the timeliness weight of each set of data in each data source; Calculate URL priority based on the PageRank score and the timeliness weight; Based on the priority, the master node is called to assign high priority URLs to the worker nodes.

5. The emergency response method according to claim 1, characterized in that: Before the step of using the preset BERT model and the preset YOLO model to process the global public health data to extract key information and classify risk levels, the method further includes: Based on SimHash and local sensitive hashing, the global public health data is quickly hashed to determine whether there is similar or repeated content; If similar or repeated content exists, a deduplication operation is performed on the global public health data to obtain deduplicated global public health data.

6. The emergency response method according to claim 1, characterized in that: Before the step of using the preset BERT model and the preset YOLO model to process the global public health data to extract key information and classify risk levels, the method further includes: Acquire sample data, wherein the processing result corresponding to the sample data is a first processing result; Using the current BERT model to perform semantic classification on the sample data, and using the current YOLO model to perform target recognition on the sample data, to obtain a second processing result; Determining whether the first processing result is consistent with the second processing result; If they are inconsistent, adjust the parameters of the current BERT model and the current YOLO model, and based on the current BERT model and the current YOLO model after the adjustment of the parameters, return to the step of using the current BERT model to perform semantic classification on the sample data, and using the current YOLO model to perform target recognition on the sample data to obtain a second processing result, until the first processing result is consistent with the second processing result, and obtain the preset extraction model.

7. An emergency response device, characterized in that: The emergency response device comprises: A crawling module, which is used for a distributed crawler cluster based on the Scrapy-Redis framework to crawl global public health data in real time; A data processing module, the data processing module is used to process the global public health data using a preset BERT model and a preset YOLO model to extract key information and classify risk levels; A trigger module is used to trigger a differentiated alarm based on the key information and the risk level.

8. An emergency response device, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the emergency response method according to any one of claims 1 to 6.

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

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the emergency response method according to any one of claims 1 to 6 are implemented.

Images