Disaster response method, system and equipment based on social media data and medium

Abstract

The invention discloses a disaster response method, system and device based on social media data and a medium, and the method comprises the following steps: carrying out the preprocessing of real-time social media data, and obtaining the preprocessing data; obtaining a plurality of real-time themes related to the disaster situation by utilizing the preprocessed data; performing priority ranking on the plurality of real-time topics to generate a real-time topic sequence; and according to the real-time theme sequence, performing resource demand evaluation on each real-time theme in sequence, and generating a real-time action suggestion. According to the method, real-time action suggestions can be continuously adjusted according to real-time changes of real-time social media data or situation updating by analyzing the transmitted real-time social media data in real time and updating real-time themes and corresponding resource requirements so as to adapt to continuously changing scenes.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular, to a disaster response method, system, device and medium based on social media data. Background Art

[0002] Social media platforms have a wide coverage and real-time nature, and have become key tools for disseminating information, coordinating rescue efforts, and understanding public needs during natural disasters. These platforms generate a large amount of unstructured text data, such as tweets and posts, which usually contain key insights into the on-the-ground reality, such as medical emergencies, shelter needs, transportation disruptions, and environmental risks.

[0003] In the prior art, existing disaster response systems often have difficulty effectively utilizing social media data for the following reasons: the unstructured nature of the data makes real-time analysis complex; the ability to understand the emotional context and urgency behind social media data is limited; determining priorities and allocating resources according to real-time needs has a delay. Therefore, existing disaster management systems cannot meet the ever-changing resource needs during disasters and conduct real-time disaster response. Summary of the Invention

[0004] The object of the present invention is to provide a disaster response method, system, device and medium based on social media data, so as to solve the problem that real-time disaster response cannot be carried out due to the ever-changing resource needs during disasters.

[0005] To solve the above technical problems, a technical solution adopted by the present invention is to provide a disaster response method based on social media data. The method includes the following steps: preprocessing real-time social media data to obtain preprocessed data; using the preprocessed data to obtain multiple real-time themes related to the disaster situation; sorting the multiple real-time themes by priority to generate a real-time theme sequence; and sequentially evaluating the resource needs of each real-time theme according to the real-time theme sequence and generating real-time action suggestions.

[0006] In some embodiments, the preprocessed data includes restored data. The step of preprocessing real-time social media data to obtain preprocessed data includes: deleting special data from the real-time social media data to obtain simplified data; the real-time social media data includes multiple initial sub-data; restoring the simplified data to obtain the restored data; the restored data includes multiple restored sub-data, and the multiple restored sub-data correspond one-to-one to the multiple initial sub-data.

[0007] In some embodiments, the preprocessing of the real-time social media data to obtain preprocessed data further includes the steps of: extracting the real-time social media data to obtain metadata related to the real-time social media data, where the preprocessed data includes the metadata; the metadata includes multiple sub-metadata, and the multiple sub-metadata correspond one-to-one with the multiple initial sub-data.

[0008] In some embodiments, the obtaining of multiple real-time topics related to the disaster situation by using the preprocessed data includes the steps of: respectively performing semantic embedding on each restored sub-data to generate corresponding semantic embedding sub-data; calculating the cosine similarity between each semantic embedding sub-data and the reference data; the reference data includes multiple reference sub-data, and the multiple reference sub-data correspond one-to-one with the multiple real-time topics; grouping each restored sub-data according to the cosine similarity to obtain the multiple real-time topics.

[0009] In some embodiments, the grouping of each restored sub-data according to the cosine similarity to obtain the multiple real-time topics includes: grouping the i-th restored sub-data into the j-th real-time topic, which can be expressed as:

[0010]

[0011] where c j represents the j-th real-time topic, j ≤ n; t' i represents the i-th restored sub-data; e i represents the i-th semantic embedding sub-data; e j represents the j-th reference data related to the j-th real-time topic; Sim(e i , e j ) represents the cosine similarity between the i-th semantic embedding sub-data e i and the j-th reference data e j ; θ represents the similarity threshold; T” represents the restored data.

[0012] In some embodiments, the prioritizing the multiple real-time topics to generate a real-time topic sequence includes: respectively performing emotion extraction on the restored sub-data corresponding to each real-time topic to obtain the user emotions contained in the corresponding initial sub-data; calculating the urgency scores corresponding to each real-time topic according to the user emotions; sorting each real-time topic according to the urgency scores to obtain the real-time topic sequence.

[0013] In some embodiments, according to the real-time theme sequence, resource requirement assessments are sequentially performed on each real-time theme, and real-time action suggestions are generated, including the steps of: respectively performing requirement mapping on each real-time theme to obtain the resource requirements corresponding to each real-time theme; generating the executable real-time action suggestions according to the resource requirements.

[0014] The present invention also provides a disaster response system based on social media data. The system includes: a preprocessing unit for preprocessing real-time social media data to obtain preprocessed data; a theme acquisition unit for using the preprocessed data to obtain a plurality of real-time themes related to the disaster situation; a theme sorting unit for sorting the priorities of the plurality of real-time themes to generate a real-time theme sequence; a theme analysis unit for sequentially performing resource requirement assessments on each real-time theme according to the real-time theme sequence and generating real-time action suggestions.

[0015] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above method are implemented.

[0016] The present invention also provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0017] The beneficial effects of the present invention are as follows: The present invention discloses a disaster response method, system, device, and medium based on social media data. The method includes the following steps: preprocessing real-time social media data to obtain preprocessed data; using the preprocessed data to obtain a plurality of real-time themes related to the disaster situation; sorting the priorities of the plurality of real-time themes to generate a real-time theme sequence; sequentially performing resource requirement assessments on each real-time theme according to the real-time theme sequence and generating real-time action suggestions. By real-time analyzing the incoming real-time social media data and updating the real-time themes and corresponding resource requirements, this method can continuously adjust the real-time action suggestions according to the real-time changes of the real-time social media data or the situation to adapt to the changing scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a schematic diagram of a disaster response method based on social media data of the present invention;

[0019] Figure 2 is a schematic diagram of a disaster response method based on social media data of the present invention;

[0020] Figure 3 is a schematic diagram of a disaster response method based on social media data of the present invention;

[0021] Figure 4 It is a schematic diagram of a disaster response method based on social media data according to the present invention;

[0022] Figure 5 It is a schematic diagram of a disaster response method based on social media data according to the present invention;

[0023] Figure 6 It is a block diagram of the composition of a disaster response system based on social media data according to the present invention;

[0024] Figure 7 It is a schematic diagram of the architecture of an embodiment of an electronic device according to the present invention;

[0025] Figure 8 It is a schematic block diagram of an embodiment of a computer-readable storage medium according to the present invention. Detailed implementation manners

[0026] For the convenience of understanding the present invention, the present invention will be described in more detail below with reference to the accompanying drawings and specific embodiments. Preferred embodiments of the present invention are shown in the accompanying drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described in this specification. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present invention more thorough and comprehensive.

[0027] It should be noted that unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used in this specification includes any and all combinations of one or more of the related listed items.

[0028] As Figure 1 shown, a disaster response method based on social media data according to the present invention is shown, including the following steps:

[0029] Step S1: Preprocess the real-time social media data to obtain preprocessed data.

[0030] It should be noted that real-time social media data refers to various types of real-time data generated on social media platforms, including real-time user information, content, interactions, relationships, etc. The sources of real-time social media data include social media platforms such as Weibo, WeChat, Facebook, Twitter, etc. The data on these platforms mainly includes various forms such as text, pictures, and videos. Its dissemination speed is fast, timeliness is strong, and the user coverage is wide, reflecting various information such as user emotions, needs, and wishes. The preprocessing includes operations such as special data deletion, data restoration, and metadata extraction on real-time social media data to obtain preprocessed data with clearer meaning, which is convenient for effectively analyzing real-time social media data. The preprocessed data includes restored data and metadata.

[0031] Specifically, as Figure 2 shown, step S1 includes the following sub-steps:

[0032] Step S11: Perform special data deletion on real-time social media data to obtain simplified data.

[0033] In this embodiment, special data deletion refers to eliminating special characters and removing stop words in real-time social media data. Special characters include non-alphabetic characters such as punctuation marks, numbers, HTML tags, etc. Stop words are common words that are ignored or deleted during text processing. These words are usually frequently occurring function words or words with no practical meaning, such as prepositions, conjunctions, articles, pronouns, etc.

[0034] Among them, in text processing, the existence of special characters may interfere with text analysis and model training. Therefore, by eliminating special characters, the cleanliness of the text and the robustness of the model can be improved.

[0035] In addition, stop words usually do not contribute much to the analysis of the meaning of the text and will occupy a large amount of storage space and computing resources. Therefore, in text processing tasks (such as text classification, information retrieval, etc.), a set of stop words is often predefined and removed from the text during the processing. When selecting stop words, factors such as the characteristics of the language, the field and goal of the task need to be comprehensively considered. The specific content of stop words can be determined according to the specific needs of the task and usually includes common words such as "a", "an", "the", "is", "are", "and", "of", etc. In text analysis, these words may interfere with the training effect of the model, so they need to be removed from the text.

[0036] In addition to English stop words, Chinese also has many commonly used stop words. The characteristics of Chinese stop words are that there are more of them, and because the Chinese vocabulary structure is more complex than English, it is more difficult to judge Chinese stop words. In Chinese texts, some common stop words include: "的", "了", "和", "是", "在", "有", "不", "我", "他", "你", etc. These words appear frequently, but usually have no meaning in text analysis.

[0037] Furthermore, the real-time social media data includes multiple initial sub-data, each of which includes text and / or special characters, and the text may include stop words. Therefore, the real-time social media data can be represented as T = {t1, t2, ...t n}, after deleting special characters and stop words, the simplified data can be expressed as T'=T-{sc,st}; where t n represents the nth initial sub-data; sc represents special characters; st represents stop words.

[0038] In this embodiment, by removing special characters and stop words from real-time social media data, the text is simplified for further analysis and the interference of non-informative content is eliminated to ensure the clarity of the text.

[0039] Step S12: restoring the simplified data to obtain restored data.

[0040] It should be noted that data restoration uses lemmatization, which means converting words into their original form, which is called a lemma or base form. The processing steps of lemmatization include: using word segmentation technology to convert text into a word list; using part-of-speech tagging technology to mark the part of speech for each word; and restoring the words to their original form based on the part of speech and context of each word.

[0041] In natural language processing, there are two common lemmatization algorithms: rule-based and statistical-based. Rule-based algorithms use predefined rules to restore words to their original forms, while statistical-based algorithms infer the original forms of words based on the frequency of occurrence of words in the corpus and the context.

[0042] In this embodiment, the simplified data includes a plurality of simplified sub-data, and the restored data includes a plurality of restored sub-data.

[0043] Therefore, the process of lemmatization of simplified data can be expressed as:

[0044]

[0045] Among them, ti represents the ith simplified sub-data, where i = 1, 2, ..., n; Lemma() represents word form restoration processing; t' i Indicates the i-th restored sub-data.

[0046] In some embodiments, lemma restoration is mainly performed on English, removing the affixes of the words and extracting the main part of the words. Usually, the extracted words will be words in the dictionary. Unlike stem extraction, the extracted words may not appear in the words. For example, the word "running" is lemma "run", the word "cars" is lemma "car", and the word "ate" is lemma "eat".

[0047] Furthermore, by restoring words to their root form to standardize the text, and word form restoration to ensure that semantically similar words are grouped, it is helpful for the subsequent extraction and analysis of real-time topics related to the disaster. At the same time, word form restoration is combined with special character elimination, stop word removal and other technologies in step S1 to improve the effect of data preprocessing.

[0048] In some other embodiments, word form restoration can also be performed for other languages. For example, in Chinese, word form changes are less than English words, but the combination of words is more flexible, and the same word may have multiple different forms. Chinese word form restoration is slightly different from English. It usually refers to converting different forms of words into their original forms. For example, "吃了" and "吃饭" are both restored to "吃饭". And the implementation of Chinese word form restoration usually requires the help of Chinese word segmentation technology and part-of-speech tagging technology. Chinese word segmentation is the process of decomposing a text into words, and part-of-speech tagging is the process of marking the part of speech for each word. Through these two steps, the original form of each word can be identified and restored.

[0049] Step S13: extract metadata from the real-time social media data to obtain metadata.

[0050] It should be noted that metadata is data that describes data, including information such as data table structure, field definition, data partition, data dependency, etc. By extracting metadata, a metadata storage and management center for the data system can be established to help better understand, master and manage data.

[0051] In this application, metadata refers to data related to real-time social media data, including data such as user location, time, and tags.

[0052] In this embodiment, the metadata includes multiple sub-metadata. And the metadata can be expressed as:

[0053] M={m1,m2,...,mn};

[0054] wherein, m n represents the nth sub - metadata corresponding to the nth initial sub - data t n .

[0055] In this embodiment, through metadata extraction, the context of the initial sub - data can be further understood, and it helps to contextualize the restored data geographically and temporally, which is crucial for subsequent location - based prioritization and action recommendations.

[0056] Step S2: Using the pre - processed data, obtain multiple real - time topics related to the disaster situation.

[0057] It should be noted that in this application, an LLM is adopted to divide the restored data in the pre - processed data into multiple meaningful and actionable real - time topics related to the disaster scenario. A real - time topic refers to the resource requirements related to the disaster scenario, such as resource requirements related to the disaster situation like medical assistance, shelter, transportation needs, etc.

[0058] Among them, the LLM receives the first - type prompt, and the first - type prompt refers to the prompt for the LLM to analyze the restored data obtained from the pre - processed real - time social media data and classify it into actionable real - time topics.

[0059] As an example, the prompt for the LLM is: "You are a disaster response assistant responsible for analyzing the restored data related to natural disasters. At the same time, divide the restored data into multiple meaningful real - time topics and concisely summarize each real - time topic with key words and sentences."

[0060] In this embodiment, the LLM first performs semantic embedding on the restored data in the pre - processed data to generate corresponding semantic - embedded data; the semantic - embedded data includes multiple semantic - embedded sub - data, and the restored sub - data corresponds one - to - one with the semantic - embedded sub - data. Then, the multiple semantic - embedded sub - data are respectively grouped into corresponding real - time topics.

[0061] Specifically, as Figure 3 shown, step S2 includes the following sub - steps:

[0062] Step S21: Perform semantic embedding on each restored sub - data respectively to generate corresponding semantic - embedded sub - data.

[0063] In this embodiment, performing semantic embedding on each restored sub - data respectively to generate corresponding semantic - embedded sub - data, this process can be expressed as:

[0064] e i = f(t' i );

[0065] wherein, ei represents the i-th semantic embedding sub-data; f() represents the semantic embedding operation; t' i represents the i-th restored data.

[0066] Step S22: Calculate the cosine similarity between each semantic embedding sub-data and the reference data.

[0067] In this embodiment, the reference data includes multiple reference sub-data, and the multiple reference sub-data correspond to multiple real-time topics one by one. Each reference sub-data can be used to determine whether each semantic embedding sub-data (each restored data) is assigned to the corresponding real-time topic.

[0068] Furthermore, calculate the cosine similarity between each semantic embedding sub-data and each reference sub-data. This process can be expressed as:

[0069]

[0070] where, e i represents the i-th semantic embedding sub-data; e j represents the j-th reference sub-data.

[0071] Step S23: Group each restored data according to the cosine similarity to obtain multiple real-time topics.

[0072] In this embodiment, group all the restored data according to the magnitude of the cosine similarity, so as to obtain multiple real-time topics. If the cosine similarity between a certain semantic embedding sub-data and a reference sub-data is greater than the similarity threshold, the restored data corresponding to this semantic embedding sub-data can be grouped into the real-time topic corresponding to this reference sub-data. The similarity threshold is the minimum value of the cosine similarity required to group the restored data corresponding to each semantic embedding sub-data into the corresponding real-time topic.

[0073] Furthermore, grouping the i-th restored data into the j-th real-time topic can be expressed as:

[0074]

[0075] where, c j represents the j-th real-time topic, j ≤ n; t' i represents the i-th restored data; e i represents the i-th semantic embedding sub-data; e j represents the j-th reference data related to the j-th real-time topic; Sim(e i , e j ) represents the cosine similarity between the i-th semantic embedding sub-data ei and the j-th reference data e j ; θ represents the similarity threshold; T” represents the restored data.

[0076] Step S3: Prioritize multiple real-time topics to generate a real-time topic sequence.

[0077] It should be noted that this step also utilizes the LLM to evaluate the urgency, user sentiment, and criticality of the real-time social media data corresponding to each extracted real-time topic, so as to effectively sort each real-time topic.

[0078] Among them, the LLM receives the second type of prompt. The second type of prompt refers to prompting the LLM to evaluate the user sentiment (such as fear, anger) contained in the initial sub-data corresponding to each real-time topic and assign an urgency level (high, medium, low). Then, according to the user sentiment and urgency level, each real-time topic is sorted.

[0079] As an example, the LLM will be prompted: "Your task is to classify the urgency level of real-time topics according to the user sentiment and frequency contained in the initial sub-data corresponding to the real-time topics. At the same time, sort the real-time topics and explain the reasons."

[0080] Specifically, as Figure 4 shown, step S3 includes the following sub-steps:

[0081] Step S31: Perform sentiment extraction on the restored sub-data corresponding to each real-time topic to obtain the user sentiment contained in the corresponding initial sub-data.

[0082] In this embodiment, the LLM performs sentiment extraction based on one or more restored sub-data corresponding to each real-time topic to obtain the user sentiment contained in one or more initial sub-data corresponding to each real-time topic. This process can be expressed as:

[0083]

[0084] where s j represents the jth user sentiment extracted from one or more restored sub-data corresponding to the jth real-time topic c j ; represents the number of restored sub-data corresponding to the jth real-time topic c j ; t' a represents the ath restored sub-data corresponding to the jth real-time topic c j ; and sentiment() represents the sentiment extraction operation.

[0085] Step S32: Calculate the urgency score corresponding to each real-time topic according to the user sentiment.

[0086] It should be noted that the level of the urgency score indicates the urgency of the disaster area involved in the corresponding initial sub-data that requires a disaster response. The higher the urgency score, the more urgent the situation of the corresponding disaster area that requires a disaster response.

[0087] In this embodiment, for the j-th real-time theme c j the corresponding urgency score is:

[0088] U j = w1f j + w2s j + w3c j ;

[0089] where f j represents the first weight corresponding to the j-th real-time theme c j ; w1 represents the first weight coefficient corresponding to the first weight f j ; w2 represents the second weight coefficient corresponding to the j-th user emotion s j ; w3 represents the third weight coefficient corresponding to the j-th real-time theme c j .

[0090] Step S33: Sort each real-time theme according to the urgency score to obtain a real-time theme sequence.

[0091] It should be noted that according to the metadata corresponding to each initial sub-data in the social media data, a plurality of real-time themes have an initial theme sequence. This step mainly re-sorts each real-time theme in the initial theme sequence from largest to smallest according to the magnitude of the urgency score corresponding to each real-time theme, so as to obtain the corresponding real-time theme sequence.

[0092] In this embodiment, the obtained real-time theme sequence is:

[0093] C' = Sort(C, by U j );

[0094] where C represents the initial theme sequence, C = {c1, c2, c3,...}; Sort(C, byU j ) represents sorting all the real-time themes in the initial theme sequence C from largest to smallest according to the corresponding urgency score U j .

[0095] In this embodiment, while obtaining the real-time theme sequence, the urgency level of each real-time theme is also classified. The urgency level corresponding to the first one-third of the real-time themes in the real-time theme sequence is "high", the urgency level corresponding to the last one-third of the real-time themes is "low", and the urgency level corresponding to the other real-time themes is "medium".

[0096] Among them, the LLM will correspondingly output the specific ranking and ranking reason of each real-time theme in the real-time theme sequence, as well as the corresponding urgency level. For example, "Real-time Theme 1: Medical Needs" ranks 1 in the real-time theme sequence, and the reason for its ranking 1 in the real-time theme sequence is: "Due to the widespread shortage of medical resources and emergency calls in multiple regions, the situation is extremely urgent"; and the corresponding urgency level is "high".

[0097] In this application, by extracting actionable real-time themes and analyzing urgency and user sentiment, real-time insights into disaster scenarios can be provided; at the same time, using the LLM to process unstructured data can provide highly accurate results for subsequent disaster responses.

[0098] Step S4: According to the real-time theme sequence, evaluate the resource requirements of each real-time theme in turn and generate real-time action suggestions.

[0099] It should be noted that the LLM, according to the real-time theme sequence, uses the urgency score corresponding to each real-time theme for demand mapping in turn to determine the resource requirements of each real-time theme (for example, ambulance, medicine, and food supply requirements, etc.), and generates detailed real-time action suggestions according to the resource requirements (for example, logistics and action implementation strategies, etc.).

[0100] At this time, the LLM receives the third type of prompt, and the third type of prompt refers to prompting the LLM to determine the specific resources required for each real-time theme (for example, ambulances, food supplies, etc.) and provide practical real-time action suggestions.

[0101] As an example, the prompt for the LLM is: "According to the priority of the real-time theme, determine the specific resources required for each real-time theme and generate practical real-time action suggestions."

[0102] Specifically, as Figure 5 shown, step S3 includes the following sub-steps:

[0103] Step S41: Map the demand of each real-time theme respectively to obtain the resource requirements corresponding to each real-time theme.

[0104] It should be noted that the resource requirements include the specific type, quantity, and location of the required resources.

[0105] In this embodiment, according to the real-time theme sequence and using the demand mapping function, the urgency scores of each real-time theme are mapped for demand in turn to obtain the corresponding resource requirements. The resource requirements corresponding to the jth real-time theme c j can be expressed as:

[0106] D j = g(U j );

[0107] Among them, D j represents the resource requirement corresponding to the j-th real-time theme c j ; g() represents the demand mapping function between the real-time theme and the resource requirement.

[0108] Step S42: Generate executable real-time action suggestions according to the resource requirements corresponding to each real-time theme.

[0109] In this embodiment, combining the resource requirements corresponding to each real-time theme, a real-time action suggestion can be generated. The real-time action suggestion includes multiple suggested operations, such as resource allocation, setting up temporary clinics, and long-term measures for continuous recovery efforts, etc.

[0110] Therefore, a real-time action suggestion obtained can be expressed as:

[0111] A = {a1, a2,..., a k};

[0112] Among them, a k represents the k-th suggested operation in the real-time action suggestion A.

[0113] In this embodiment, generating executable real-time action suggestions according to the resource requirements corresponding to each real-time theme, enabling disaster response personnel to adopt the corresponding real-time action suggestions, can simplify and accelerate the decision-making process and reduce the delay of disaster relief operations.

[0114] In some embodiments, it is also necessary to optimize the resource allocation corresponding to each real-time theme. Among them, the optimization formula is used:

[0115]

[0116] Among them, minC represents the optimized resource allocation corresponding to the j-th real-time theme c j ; R s represents the s-th type of resource, s = 1, 2,... m.

[0117] In this application, metadata is also combined, and the priority of resource requirements is dynamically determined according to context analysis, so as to make faster and more informed resource allocation decisions; at the same time, the resource requirements corresponding to real-time themes with an emergency level of "high" are systematically prioritized to ensure fair and balanced resource allocation, minimize waste, and effectively solve urgent resource requirements.

[0118] It should be noted that the LLM integrates resource requirements and real-time action suggestions into a structured format for the use of the disaster response team.

[0119] For example, the LLM finally outputs "Resource requirements: 20 doctors and nurses, 10 ambulances with drivers, and 1000 medical supplies (bandages, medicines); Real-time action suggestions: First, immediately deploy emergency medical teams to high-demand areas and set up temporary clinics in nearby safe areas to address the overabundance of treatment needs; Then,..."

[0120] Further, repeat the above steps and update the database storing real-time social media data.

[0121] Among them, by re-analyzing the incoming real-time social media data and updating the priority and resource plans to adapt to changing scenarios, this ensures a dynamic and flexible response to changing disaster situations, and can continuously adjust the real-time action suggestions according to the real-time changes of real-time social media data or situational updates.

[0122] In this application, by combining real-time action suggestions with insights from real-time social media data, resource allocation can be optimized to minimize overabundance or shortage of material supplies; at the same time, it can also adapt to various types of natural disasters, regions, and the scale of social media activities.

[0123] Based on the same inventive concept, as Figure 6 shown, the present invention also provides a disaster response system based on social media data, which includes:

[0124] A preprocessing unit 101 for preprocessing real-time social media data to obtain preprocessed data.

[0125] A theme acquisition unit 102 for using the preprocessed data to obtain multiple real-time themes related to the disaster situation.

[0126] A theme ranking unit 103 for ranking the multiple real-time themes by priority to generate a real-time theme sequence.

[0127] A theme analysis unit 104 for sequentially evaluating the resource requirements of each real-time theme according to the real-time theme sequence and generating real-time action suggestions.

[0128] In this application, other technical features in the above disaster response system based on social media data are the same as those disclosed in the above method embodiment, and will not be elaborated here.

[0129] Based on the same inventive concept, this application also provides an electronic device, which includes a processor, a memory, and a communication circuit. The processor is respectively connected to the memory and the communication circuit; wherein, the communication circuit is used for communication connection, the memory is used for storing a computer program, and the processor is used for executing the computer program to implement the above method.

[0130] Please refer toFigure 7 In the embodiments of the present application, the described electronic device may specifically include a processor 210 and a memory 220. The memory 220 is coupled to the processor 210.

[0131] The processor 210 is used to control the operation of the electronic device. The processor 210 may also be referred to as a CPU (Central Processing Unit). The processor 210 may be an integrated circuit chip with signal processing capabilities. The processor 210 may also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general-purpose processor may be a microprocessor, or the processor 210 may also be any conventional processor, etc.

[0132] The memory 220 is used to store computer programs, which may be RAM, ROM, or other types of storage terminals. Specifically, the memory 220 may include one or more computer-readable storage media, which may be non-transitory or transitory. The memory 220 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage terminals and flash storage terminals. In some embodiments, the non-transitory computer-readable storage medium in the memory 220 is used to store at least one program code.

[0133] The processor 210 is used to execute the computer programs stored in the memory 220 to implement the methods described in the method embodiments of the present application.

[0134] In some embodiments, the electronic device may further include: a peripheral terminal interface 230 and at least one peripheral terminal. The processor 210, the memory 220, and the peripheral terminal interface 230 may be connected through a bus or signal lines. Each peripheral terminal may be connected to the peripheral terminal interface 230 through a bus, signal lines, or a circuit board. Specifically, the peripheral terminal includes at least one of a radio frequency circuit 240, a display screen 250, an audio circuit 260, and a power supply 270.

[0135] The peripheral terminal interface 230 may be used to connect at least one I / O (Input / Output) related peripheral terminal to the processor 210 and the memory 220. In some embodiments, the processor 210, the memory 220, and the peripheral terminal interface 230 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 210, the memory 220, and the peripheral terminal interface 230 may be implemented on a separate chip or circuit board, and this embodiment does not limit this.

[0136] The radio frequency circuit 240 is used to receive and transmit RF (Radio Frequency) signals, also known as electromagnetic signals. The radio frequency circuit 240 communicates with the communication network and other Internet of Things devices through electromagnetic signals, and the radio frequency circuit 240 is the communication circuit of the electronic device. The radio frequency circuit 240 converts electrical signals into electromagnetic signals for transmission, or converts the received electromagnetic signals into electrical signals. Optionally, the radio frequency circuit 240 includes: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a subscriber identity module card, and so on. The radio frequency circuit 240 can communicate with other terminals through at least one wireless communication protocol. The wireless communication protocol includes but is not limited to: the World Wide Web, metropolitan area network, intranet, generations of mobile communication networks (2G, 3G, 4G, and 5G), wireless local area network, and / or WiFi (Wireless Fidelity) network. In some embodiments, the radio frequency circuit 240 may further include a circuit related to NFC (Near Field Communication), which is not limited in this application.

[0137] The display screen 250 is used to display the UI (User Interface). The UI may include graphics, text, icons, videos, and any combination thereof. When the display screen 250 is a touch display screen, the display screen 250 also has the ability to collect touch signals on or above the surface of the display screen 250. The touch signal can be input to the processor 210 for processing as a control signal. At this time, the display screen 250 can also be used to provide virtual buttons and / or virtual keyboards, also known as soft buttons and / or soft keyboards. In some embodiments, there may be one display screen 250, which is set on the front panel of the electronic device; in other embodiments, there may be at least two display screens 250, which are respectively set on different surfaces of the electronic device or are in a folding design; in other embodiments, the display screen 250 may be a flexible display screen, which is set on the curved surface or folding surface of the electronic device. Even, the display screen 250 can be set into an irregular non-rectangular shape, that is, a special-shaped screen. The display screen 250 can be prepared from materials such as LCD (Liquid Crystal Display) and OLED (Organic Light-Emitting Diode).

[0138] The audio circuit 260 may include a microphone and a speaker. The microphone is used to collect sound waves of the operator and the environment, and convert the sound waves into electrical signals for input to the processor 210 for processing, or input to the radio frequency circuit 240 to implement voice communication. For the purpose of stereo collection or noise reduction, there may be multiple microphones, which are respectively arranged at different parts of the electronic device. The microphone may also be an array microphone or an omnidirectional collection microphone. The speaker is used to convert the electrical signal from the processor 210 or the radio frequency circuit 240 into sound waves. The speaker may be a traditional thin film speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can not only convert the electrical signal into audible sound waves for humans, but also convert the electrical signal into inaudible sound waves for humans for uses such as ranging. In some embodiments, the audio circuit 260 may further include a headphone jack.

[0139] The power supply 270 is used to supply power to each component in the electronic device. The power supply 270 may be alternating current, direct current, a primary battery or a rechargeable battery. When the power supply 270 includes a rechargeable battery, the rechargeable battery may be a wired rechargeable battery or a wireless rechargeable battery. A wired rechargeable battery is a battery charged through a wired line, and a wireless rechargeable battery is a battery charged through a wireless coil. The rechargeable battery may also be used to support fast charging technology.

[0140] For a detailed description of the functions and execution processes of each functional module or component in the electronic device embodiments of the present application, reference may be made to the descriptions in the above method embodiments of the present application, and details are not repeated here.

[0141] In several embodiments provided in the present application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the electronic device embodiments described above are merely illustrative. For example, the division of modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other may be through some interfaces, and the indirect couplings or communication connections of devices or units may be in electrical, mechanical or other forms.

[0142] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0143] In addition, in each embodiment of the present application, each functional unit can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0144] Based on the same inventive concept, the present application also provides a computer-readable storage medium storing a computer program that can be executed by a processor to implement the above method.

[0145] Please refer to Figure 8 , when the above integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in the computer-readable storage medium 300. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions / computer programs for causing an Internet of Things device (which can be a personal computer, a server, or a network terminal, etc.) or a processor to execute all or part of the steps of the methods in various embodiments of the present application. The foregoing storage medium includes: various media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk, or an optical disc, as well as electronic terminals such as a computer, a mobile phone, a notebook computer, a tablet computer, and a camera having the above storage medium.

[0146] The description of the execution process of the program data in the computer-readable storage medium can refer to the description in the above method embodiments of the present application and will not be repeated here.

[0147] As can be seen, the present invention discloses a disaster response method, system, device, and medium based on social media data. The method includes the following steps: preprocessing real-time social media data to obtain preprocessed data; using the preprocessed data to obtain multiple real-time topics related to the disaster situation; sorting the multiple real-time topics by priority to generate a real-time topic sequence; and according to the real-time topic sequence, sequentially evaluating the resource requirements for each real-time topic and generating real-time action suggestions. By analyzing the incoming real-time social media data in real time and updating the real-time topics and corresponding resource requirements, this method can continuously adjust the real-time action suggestions according to the real-time changes of the real-time social media data or the situation to adapt to the changing scenarios.

[0148] The above are only embodiments of the present invention, and thus do not limit the protection scope of the present invention. Any equivalent structural transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be included in the protection scope of the present invention.

Claims

1. A disaster response method based on social media data, characterized in that: The method comprises the following steps: Preprocessing the real-time social media data to obtain preprocessed data; Using the preprocessed data, a plurality of real-time topics related to the disaster situation are obtained; Prioritize the multiple real-time topics to generate a real-time topic sequence; According to the real-time topic sequence, resource demand assessment is performed on each real-time topic in turn, and real-time action suggestions are generated.

2. The disaster response method according to claim 1, characterized in that: The preprocessing data includes restoring data; the preprocessing of the real-time social media data to obtain the preprocessed data includes the steps of: Performing special data deletion on the real-time social media data to obtain simplified data; the real-time social media data includes a plurality of initial sub-data; The simplified data is restored to obtain the restored data; the restored data includes a plurality of restored sub-data, and the plurality of restored sub-data correspond to the plurality of initial sub-data in a one-to-one manner.

3. The disaster response method according to claim 2, characterized in that: The preprocessing of the real-time social media data to obtain preprocessed data also includes the steps of: The real-time social media data is extracted to obtain metadata related to the real-time social media data, wherein the preprocessed data includes the metadata; the metadata includes a plurality of sub-metadata, and the plurality of sub-metadata correspond one-to-one to the plurality of initial sub-data.

4. The disaster response method according to claim 2, characterized in that: The method of using the pre-processed data to obtain multiple real-time topics related to the disaster situation includes the following steps: Semantically embed each restored sub-data to generate corresponding semantically embedded sub-data; Calculating the cosine similarity between each semantic embedding sub-data and reference data; the reference data includes a plurality of reference sub-data, and the plurality of reference sub-data correspond one-to-one to a plurality of real-time topics; The restored sub-data are grouped according to the cosine similarity to obtain the multiple real-time topics.

5. The disaster response method according to claim 4, characterized in that: The step of grouping the restored sub-data according to the cosine similarity to obtain the multiple real-time topics includes: Grouping the i-th restored sub-data into the j-th real-time topic can be expressed as: Among them, c j represents the jth real-time topic, j≤n; t' i represents the i-th restored sub-data; e i represents the i-th semantic embedded sub-data; e j represents the jth reference data related to the jth real-time topic; Sim(e i ,e j ) represents the i-th semantic embedding sub-data e i and the jth reference data e j ; θ represents the similarity threshold; T" represents the restored data.

6. The disaster response method according to claim 4, characterized in that: The step of prioritizing the plurality of real-time topics to generate a real-time topic sequence comprises: Perform emotion extraction on the restored sub-data corresponding to each real-time topic to obtain the user emotions contained in the corresponding initial sub-data; Calculating the urgency score corresponding to each real-time topic according to the user emotion; The real-time topics are sorted according to the urgency scores to obtain the real-time topic sequence.

7. The disaster response method according to claim 1, characterized in that: According to the real-time topic sequence, resource demand assessment is performed on each real-time topic in turn, and real-time action suggestions are generated, including the steps of: Mapping the requirements of each real-time topic to obtain the resource requirements corresponding to each real-time topic; Generate the executable real-time action suggestion according to the resource demand.

8. A disaster response system based on social media data, characterized in that: The system comprises: A preprocessing unit, used for preprocessing the real-time social media data to obtain preprocessed data; A topic acquisition unit, used to obtain multiple real-time topics related to the disaster situation using the pre-processed data; A topic sorting unit, used to sort the multiple real-time topics by priority and generate a real-time topic sequence; The topic analysis unit is used to evaluate the resource requirements of each real-time topic in turn according to the real-time topic sequence and generate real-time action suggestions.

9. An electronic device, characterized in that: The method comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images