Disaster Countermeasure Information Provision System
Patent Information
- Application Number
- JP2025057981
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2025-03-31
- Publication Date
- 2026-10-14
AI Technical Summary
【0007】 本発明によれば、必要とする災害対策情報をAI処理によって取得する際に、あらかじめ災害対策データベースから類似文章及び類似データの少なくともいずれかを検索して参照情報を取得し、この参照情報を用いて修正質問文を作成し、この修正質問文をAI処理に用いることで、より正確で的確な回答文を得ることができる。特に、災害現場で災害対策情報を得なければならない状況では、緊急性、専門性、特殊性、固有性が高いことからも質問文に曖昧さが含まれてしまう可能性が高いが、参照情報を用いた修正質問文を作成することで曖昧さを無くし、災害防御活動の指揮に必要な正確で的確な災害対策情報を得ることができる。
Smart Images

Figure 2026165546000001_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a disaster countermeasure information providing system that provides disaster countermeasure information necessary for commanding disaster prevention activities at a disaster site.
Background Art
[0002] Patent Document 1 discloses a disaster support system that can safely and swiftly proceed with activities when a disaster occurs, and can promptly perform rescue work for disaster victims. Patent Document 2 discloses a portable alarm system that allows a commanding captain to grasp the status of team members at a remote location, and can immediately issue an alarm when an abnormality has occurred or may occur to a team member. Patent Document 3 discloses emergency communication command equipment that relates to the guidance of emergency vehicles at the time of fire and disaster occurrence, and in particular notifies the arrangement of each emergency vehicle with reference to the weather information at the fire and disaster occurrence point.
Prior Art Documents
Patent Documents
[0003]
Patent Document 1
Patent Document 2
Patent Document 3
Summary of the Invention
Problems to be Solved by the Invention
[0004] For example, at a disaster site, a commanding captain must quickly grasp information related to firefighting team members and people requiring rescue, and issue accurate commands to firefighting vehicles and firefighting team members. To achieve this, it is necessary to quickly and accurately grasp the disaster countermeasure information required for commanding disaster prevention activities. However, a disaster countermeasure information providing system that provides the required disaster countermeasure information has not yet been proposed.
[0005] The present invention aims to provide a disaster response information provision system that can obtain accurate and appropriate disaster response information necessary for directing disaster prevention activities. [Means for solving the problem]
[0006] The disaster response information provision system of the present invention as described in claim 1 is a disaster response information provision system that provides disaster response information necessary for directing disaster prevention activities at a disaster site, characterized in that when a computer obtains a question sentence requesting the disaster response information as text data, it executes a classification item determination command sentence creation step which creates a classification item determination command sentence using classification item-specific disaster prevention information, which pre-classifies the disaster prevention activities into classification items, and the question sentence; a reference information acquisition step which searches for at least one of similar sentences and similar data from disaster response databases 4 and 5 based on the classification items obtained by AI processing based on the classification item determination command sentence, and obtains as reference information those containing the similar sentences or similar data with high similarity; a question sentence modification command sentence creation step which creates a question sentence modification command sentence using the reference information obtained in the reference information acquisition step and the question sentence; an answer creation command sentence creation step which creates an answer creation command sentence for the disaster response information using a modified question sentence obtained by AI processing based on the question sentence modification command sentence; and an answer sentence output step which outputs an answer sentence obtained by AI processing based on the answer creation command sentence. The present invention as described in claim 2 is a disaster countermeasure information provision system as described in claim 1, wherein when the computer obtains the revised question, it performs a reclassification item determination command creation step, which creates a reclassification item determination command using the classification item-specific disaster prevention information and the revised question, and a rereference information acquisition step, which searches the disaster countermeasure databases 4 and 5 for at least one of the most similar documents and similar data based on the reclassification items obtained by the AI processing based on the reclassification item determination command, and obtains the most similar documents or similar data with high similarity as rereference information, and in the response creation command creation step, it creates the response creation command including the rereference information. The present invention as described in claim 3 is characterized in that, in the disaster response information provision system described in claim 1, if the question is disaster response information based on fire safety regulations or fire safety manuals, the reference information acquisition step uses the literature reference database 4 as the disaster response databases 4 and 5, and if the question is disaster response information based on the situation at the disaster site, the reference information acquisition step uses the site information database 5 as the disaster response databases 4 and 5. The present invention as described in claim 4 is characterized in that, in the disaster response information provision system described in claim 1, the question is in voice and the text data is converted from the voice to text. The present invention as described in claim 5 is characterized in that, in the disaster response information provision system described in claim 3, the on-site information database 5 includes any of the following information: personnel information including personnel personal information, personnel location information, or personnel vital information, and personnel information including personnel personal information, personnel location information, or triage information, together with time information based on the time of occurrence or input time. [Effects of the Invention]
[0007] According to the present invention, when acquiring necessary disaster response information through AI processing, reference information is obtained by first searching a disaster response database for at least one of similar sentences and similar data, creating a revised question using this reference information, and then using this revised question in AI processing, thereby obtaining a more accurate and precise answer. In particular, in situations where disaster response information must be obtained at a disaster site, the question is likely to contain ambiguity due to its urgency, specialization, specificity, and uniqueness. However, by creating a revised question using reference information, ambiguity can be eliminated, and accurate and precise disaster response information necessary for directing disaster prevention activities can be obtained. [Brief explanation of the drawing]
[0008] [Figure 1] Conceptual diagram of a disaster site utilizing a disaster response information provision system according to one embodiment of the present invention. [Figure 2] A flowchart illustrating the processing flow when the question in the disaster response information provision system according to this embodiment is a reference to literature. [Figure 3] A flowchart illustrating the processing flow when the question in the disaster response information provision system according to this embodiment is on-site information. [Figure 4] Image of reference information obtained from S8 [Figure 5] Image of the question correction command created in the question correction command step. [Figure 6] Image of the command statement for determining reclassification items created in the command statement creation step for determining reclassification items. [Figure 7] Image of rereference information obtained in S14 [Figure 8] Image of the command statement for determining reclassification items created in the command statement creation step for determining reclassification items. [Figure 9] Image of rereference information obtained by S34 (when the command is "What are the characteristics of the person to be rescued?") [Figure 10] Image of rereference information obtained by S34 (when the command is "What is the status of the team member?") [Figure 11]Image of the response text obtained in S36 [Modes for carrying out the invention]
[0009] The disaster response information provision system according to the first embodiment of the present invention, when a computer obtains a question sentence requesting disaster response information as text data, executes the following steps: a classification item determination command creation step, which creates a classification item determination command sentence using classification item-specific disaster prevention information, which pre-classifies disaster prevention activities into classification items, and the question sentence; a reference information acquisition step, which searches the disaster response database for at least one of similar sentences and similar data based on the classification items obtained by AI processing based on the classification item determination command sentence, and obtains similar sentences or similar data with high similarity as reference information; a question sentence modification command sentence creation step, which creates a question sentence modification command sentence using the reference information obtained in the reference information acquisition step and the question sentence; an answer creation command sentence creation step, which creates an answer creation command sentence for disaster response information using a modified question sentence obtained by AI processing based on the question sentence modification command sentence; and an answer sentence output step, which outputs an answer sentence obtained by AI processing based on the answer creation command sentence. According to this embodiment, when acquiring necessary disaster response information through AI processing, reference information is obtained by first searching the disaster response database for at least one of similar sentences and similar data, creating a revised question using this reference information, and then using this revised question in AI processing, thereby obtaining a more accurate and precise answer. In particular, in situations where disaster response information must be obtained at a disaster site, the question is likely to contain ambiguity due to its urgency, specialization, specificity, and uniqueness. However, by creating a revised question using reference information, ambiguity can be eliminated, and accurate and precise disaster response information necessary for directing disaster prevention activities can be obtained.
[0010] In a disaster countermeasure information providing system according to the first embodiment, the second embodiment of the present invention executes: a reclassification item determining instruction sentence creation step in which, when a computer acquires a modified question sentence, the computer creates a reclassification item determining instruction sentence using disaster prevention information by classification item and the modified question sentence; and a re-reference information acquisition step of retrieving at least one of re-similar sentences and re-similar data from a disaster countermeasure database based on a reclassification item obtained by AI processing based on the reclassification item determining instruction sentence, and acquiring information including the re-similar sentences or re-similar data with high similarity as re-reference information, and in the response creation instruction sentence creation step, a response creation instruction sentence is created including the re-reference information. According to the present embodiment, when required disaster countermeasure information is acquired by AI processing, re-reference information is acquired in advance from the disaster countermeasure database using a modified question sentence, and the re-reference information is used in the AI processing together with the modified question sentence, whereby a more accurate and precise response sentence can be obtained.
[0011] In the disaster countermeasure information providing system according to the first embodiment, the third embodiment of the present invention uses, in the reference information acquisition step, a document reference database as the disaster countermeasure database if the question sentence is disaster countermeasure information based on fire fighting regulations or a fire fighting manual, and uses, in the reference information acquisition step, a site information database as the disaster countermeasure database if the question sentence is disaster countermeasure information based on the situation of a disaster site. According to the present embodiment, by using the document reference database when the question sentence is disaster countermeasure information based on fire fighting regulations or a fire fighting manual, and using the site information database when the question sentence is disaster countermeasure information based on the situation of a disaster site, an accurate and precise modified question sentence free of ambiguity can be created.
[0012] In the disaster countermeasure information providing system according to the first embodiment, the fourth embodiment of the present invention is configured such that the question sentence is based on voice, and the text data is obtained by converting the voice into text. According to the present embodiment, since voice input is possible, response sentences can be obtained quickly at disaster sites where urgency is required, and by using the corrected question sentences for AI processing, it is possible to eliminate the influence of ambiguity caused by voice input and noise generated by the surrounding environment.
[0013] In the disaster countermeasure information providing system according to the third embodiment, the fifth embodiment of the present invention includes, in the site information database, any one of member information including member personal information, member position information, or member vital information, and rescuee information including rescuee personal information, rescuee position information, or triage information, together with time information based on the occurrence time or input time. According to the present embodiment, the situations of members and rescuees at the disaster site can be quickly grasped. [Examples]
[0014] Hereinafter, a disaster countermeasure information providing system according to an embodiment of the present invention will be described. Figure 1 is a conceptual diagram of a disaster site using the disaster countermeasure information providing system according to the present embodiment. The disaster countermeasure information providing system according to the present embodiment is a system that provides disaster countermeasure information necessary to support the command that the command chief 1 issues to the members 2 and the mobile devices 3, and the disaster countermeasure information can also be used by the members 2 and the mobile devices 3. The mobile devices 3 include fire vehicles, as well as drones and fire-fighting robots. The disaster countermeasure information providing system is connected to the document reference database 4 and the site information database 5 via communication means. In the document reference database 4, manuals or document data concerning fire defense activities are stored as fire defense activity data for each fire defense activity category. The field information database 5 stores field activity data, including structural data of the damaged object, floor plan data, map data of a designated area including the damaged object, team member location data, mobile vehicle location data, vital data, mobile vehicle status data, rescue victim location data, safe area data, dangerous area data, fire location data, conversation data, team member data, mobile vehicle data, rescue victim data, and triage data, along with time data.
[0015] Figures 2 and 3 are flowcharts showing the processing flow of the disaster response information provision system according to this embodiment. Figure 2 shows the processing when the question is a reference to literature, and Figure 3 shows the processing when the question is on-site information. When Commander 1 inputs a question via voice requesting disaster response information necessary for directing disaster relief activities at the disaster site (S1), the computer converts the input voice into text and creates text data (S2). The computer determines whether the entered question is a reference to literature (S3). This determination in S3 can be made based on keywords in the text data, or it may be made based on the input operation. If the computer determines in S3 that the question is a reference (Yes in S3), it retrieves the question as text data (S4). References are disaster response information based on fire safety regulations or fire safety manuals, and reference database 4 is used as the disaster response database. When the computer obtains a question requesting disaster response information as text data, it executes a step to create a classification item determination command statement using the question and the disaster prevention information categorized by classification item, which is a classification of disaster prevention activities in advance.
[0016] The classification items in disaster prevention information by classification category include, for example, 'buildings', 'training', 'emergency investigation', 'special disasters', 'command activities', 'rescue activities', 'other', 'communication command', 'firefighting activities', 'general firefighting activities', 'mass casualty incidents', 'fire extinguishing', 'water accidents', 'NBC disasters', 'fire prevention plan', 'subway fire prevention', 'on-site deployment', 'tunnels', 'general on-site activities', 'return to station / reporting', 'vehicle fires', 'ship fires', 'gas leaks', 'fire prevention plan', 'train accidents', 'vehicles / equipment', 'underground shopping mall fire prevention', 'information management', 'odor investigation', 'support deployment', 'traffic accidents', 'earthquakes', 'tunnel deployment', and 'emergency medical activities'. For example, under the classification category "Information Management," there is further information such as "Activity records and analysis of firefighting activities (firefighting, rescue, emergency medical services, command), good examples, bad examples, annual firefighting activity statistics, guidelines for recording activity records, and materials for considering incidents." Under the classification category "General Firefighting Activities," there is further information such as "Materials that cover the entire content of firefighting activities, from fire prevention plans, dispatch systems and plans, to materials concerning basic techniques in on-site operations, basic firefighting procedures, and basic techniques necessary for rescue operations." Under the classification category "Fire Prevention Plan (Fire Prevention): Materials concerning the basic plan for fire prevention in the jurisdictional fire area and its creation procedures, regulations, accompanying dispatch plans and systems, and plans for specific incidents that deviate from the basic plan." Under the classification category "Fire Prevention Plan (Preparation) Regarding the category "Prevention," there is further information that it contains "materials related to prevention work, on-site inspections, and inspections in the fire service area under its jurisdiction." Regarding the category "Communication and Command," there is further information that it contains "materials related to the communication and command center." Regarding the category "On-site Deployment," there is further information that it contains "materials related to on-site deployments established based on the systems formulated in the fire prevention plan and deployment plan." Regarding the category "General On-site Activities," there is further information that it contains "materials that summarize the basic knowledge that each team, including firefighting, rescue, and emergency medical services, needs to understand when conducting on-site activities, and are not specific to particular incidents." Regarding the category "Firefighting Activities," there is further information that it contains "materials related to the basic knowledge necessary for firefighting teams to operate, firefighting procedures, vehicle deployment locations, fire hydrants, relays, water discharge tactics, etc."
[0017] In the step of creating a command statement for determining classification items, for example, if the command statement is "Precautions when spraying water to extinguish an EV vehicle fire," then this command statement will be inserted, and the above classification items will be inserted as "the list below."
[0018] Based on the classification item determination command statement created in the classification item determination command statement creation step in S5, the classification items are obtained by AI processing (S6). In S6, for example, the classification items "vehicle fire" and "firefighting activities" are acquired. In the reference information acquisition step, based on the classification items obtained in S6 ("vehicle fire" and "firefighting activities"), similar documents are searched from the literature reference database 4 (S7), and those containing highly similar documents are acquired as reference information (S8). Vector store search is suitable for S7. Figure 4 shows an image of the reference information obtained by S8.
[0019] In the step for creating a command to modify the question in S9, a command to modify the question is created using the reference information obtained in the reference information acquisition step and the question itself. Figure 5 shows an example of a question correction command created in the question correction command step. Based on the question correction instruction created in the question correction instruction step in S9, the corrected question is obtained by AI processing (S10). The revised question obtained in S10 might be, for example, "Precautions when using water to extinguish a fire in a vehicle equipped with a lithium-ion battery." In this way, "EV vehicle" in the original question is replaced with "vehicle equipped with a lithium-ion battery" in the revised question.
[0020] When the computer obtains the revised query, it executes a step to create a reclassification item determination command statement using the classification item-specific disaster prevention information and the revised query statement (S11). Figure 6 shows an image of the command statement for determining reclassification items that is created in the command statement creation step for determining reclassification items. The reclassification item determination command created in the reclassification item determination command creation step is basically the same as the classification item determination command created in the classification item determination command creation step, except that the "question" is replaced with a "modified question." However, the number of classification items requested in the reclassification item determination command is less than the number of classification items requested in the classification item determination command. In this embodiment, the number of classification items requested in the classification item determination command is set to two, and the number of classification items requested in the reclassification item determination command is set to one.
[0021] Based on the reclassification item determination command statement created in the reclassification item determination command statement creation step in S11, the reclassification items are obtained by AI processing (S12). In S12, for example, "vehicle fire" is acquired as a classification item. In the re-reference information acquisition step, based on the reclassification item "vehicle fire" acquired in S12, the most similar documents are searched for in the literature reference database 4 (S13), and those containing highly similar documents are acquired as reference information (S14). Vector store search is suitable for S13. Figure 7 is an image of the rereference information obtained by S14.
[0022] In the step of creating the response command in S15, the AI process uses the revised question and re-referenced information obtained from the question revision command to create the response command for disaster countermeasures information. Based on the response creation command created in the response creation step in S15, the response is obtained by AI processing (S16). In the response output step in S17, the response text obtained through AI processing is output based on the response creation instruction text.
[0023] If the computer determines in S3 that the question is not a reference (resulting in "No" in S3), it proceeds to process "A" in Figure 2. Figure 3 shows the process from [A] onwards. If the computer determines in S23 that the question is on-site information (Yes in S23), it retrieves the question as text data (S24). Field information includes information on disaster response activities at disaster sites, such as information on team members and information on people who need rescuing. Field information database 5 is used as the disaster response database. Team member information includes not only personal information but also location and vital signs, along with time information. Information on those requiring care includes not only personal information but also location and triage information, along with time information. Time information is recorded by the time of occurrence or input time. When the computer obtains a question requesting disaster response information as text data, it executes a step to create a classification item determination command statement using the question and the disaster prevention information categorized by classification item, which is a classification of disaster prevention activities in advance (S25).
[0024] Examples of classification items in disaster prevention information by classification category include 'members', 'incidents', 'rescurees', 'fire_trucks', 'units', 'triages', 'structure_info', and 'triage_vitals'. Furthermore, for example, the classification item "incidents" contains information such as "a table that manages incidents, especially including information on the location, time of occurrence, and end time of the incident," the classification item "units" contains information such as "a table that manages a list of members of the dispatched unit, including a list of the commander and members," the classification item "members" contains information such as "a table that manages real-time information on dispatched members," the classification item "fire_trucks" contains information such as "a table that manages dispatched vehicles, including vehicle type and pump information," the classification item "structure_info" contains information such as "a table that manages structural information, including drawing information and water pipe information," the classification item "rescurees" contains information such as "a table that manages people to be rescued, including a summary of the rescued person and rescue status information," and the classification item "triages" contains information such as "triage_vitals: information on triage vitals."
[0025] In the step of creating a command statement for determining classification items, for example, if the command statement is "What are the characteristics of the person in need of rescue?", then this command statement will be inserted, and the above classification items will be inserted as "the list below".
[0026] Based on the classification item determination command statement created in the classification item determination command statement creation step in S25, the classification items are obtained by AI processing (S26). In S26, for example, if the command is "What are the characteristics of the person to be rescued?", the classification items obtained will be "rescurees" and "triages". Also, if the command is "What is the condition of the team members?", the classification items obtained will be "members" and "units". In the reference information acquisition step, based on the classification items obtained in S26 ("rescurees" and "triages," or "members" and "units"), the system searches the field information database 5 for at least one of the similar documents and similar data (S27), and acquires the documents containing the highly similar documents or similar data as reference information (S28). Vector store search is suitable for S27.
[0027] In the step for issuing a command to modify the question in S29, a command to modify the question is created using the reference information obtained in the reference information acquisition step and the question itself. Based on the question text modification instruction created in the question text modification instruction step in S29, the modified question text is obtained by AI processing (S30). When the computer obtains the revised query, it executes a step to create a reclassification item determination command statement using the disaster prevention information by classification item and the revised query statement (S31). Figure 8 shows an example of a command statement for determining reclassification items, which is created in the command statement creation step. Figure 8(a) shows the case where the command statement is "What are the characteristics of the person to be rescued?", and Figure 8(b) shows the case where the command statement is "What is the condition of the team member?". The reclassification item determination command created in the reclassification item determination command creation step is basically the same as the classification item determination command created in the classification item determination command creation step, except that the "question" is replaced with a "modified question." However, the number of classification items requested in the reclassification item determination command is less than the number of classification items requested in the classification item determination command. In this embodiment, the number of classification items requested in the classification item determination command is set to two, and the number of classification items requested in the reclassification item determination command is set to one.
[0028] Based on the reclassification item determination command statement created in the reclassification item determination command statement creation step in S31, the reclassification items are obtained by AI processing (S32). In the re-reference information acquisition step, based on the reclassification items acquired in S32, at least one of the most similar documents and similar data is searched from the field information database 5 (S33), and the documents containing the most similar documents or similar data are acquired as reference information (S34). Vector store search is suitable for S33. Figures 9 and 10 are images of the rereference information obtained by S34. Figure 9 shows the case where the command is "What are the characteristics of the person to be rescued?", and Figure 10 shows the case where the command is "What is the condition of the team member?".
[0029] In the step of creating response commands in S35, response commands for disaster response information are created using the revised question text obtained by AI processing and the re-referenced information, based on the question text revision command text. Based on the response creation command created in the response creation step in S35, the response is obtained by AI processing (S36). Figure 11 shows an example of the response text that may be obtained. Figure 11(a) shows the case where the command is "What are the characteristics of the person to be rescued?", and Figure 11(b) shows the case where the command is "What is the condition of the team member?". In the response output step, the response obtained through AI processing is output based on the response creation instruction (S37).
[0030] If the computer determines in S23 that the question is not a field reference (resulting in No in S23), it will not process the question.
[0031] As described above, according to this embodiment, when acquiring necessary disaster response information through AI processing, reference information is obtained by first searching for at least one of similar sentences and similar data from disaster response databases 4 and 5, creating revised questions using this reference information, and using these revised questions in AI processing, thereby obtaining more accurate and precise answers. In particular, in situations where disaster response information must be obtained at a disaster site, the questions are likely to contain ambiguity due to their urgency, specialization, uniqueness, and individuality. However, by creating revised questions using reference information, ambiguity can be eliminated, and accurate and precise disaster response information necessary for directing disaster prevention activities can be obtained. Furthermore, according to this embodiment, when acquiring necessary disaster response information through AI processing, re-reference information is obtained in advance from disaster response databases 4 and 5 using modified question sentences, and the re-reference information is used in the AI processing along with the modified question sentences, thereby obtaining more accurate and precise answers. Furthermore, according to this embodiment, if the question is disaster countermeasure information based on fire safety regulations or fire safety manuals, the literature reference database 4 is used, and if the question is disaster countermeasure information based on the situation at the disaster site, the field information database 5 is used, thereby enabling the creation of accurate and precise revised questions without ambiguity. Furthermore, according to this embodiment, voice input enables rapid acquisition of answers in emergency situations such as disaster sites, and the use of revised question sentences in AI processing eliminates ambiguity caused by voice input and the influence of noise generated by the surrounding environment. Furthermore, according to this embodiment, by including personnel information, including personnel location information and personnel vital information, and information on persons to be rescued, including the location information and triage information of persons to be rescued, along with time information based on the time of occurrence or input time, the situation of personnel 2 and persons to be rescued at the disaster site can be quickly grasped. [Industrial applicability]
[0032] The disaster response information provision system according to the present invention can quickly and accurately provide disaster response information necessary for commanders to direct operations, especially in large-scale disasters and special disaster situations. [Explanation of symbols]
[0033] 1 Commander 2 members 3 Mobile Unit 4. Reference Database (Disaster Countermeasure Database) 5. Field Information Database (Disaster Countermeasure Database)
Claims
1. A disaster response information provision system that provides disaster response information necessary for directing disaster prevention activities at disaster sites, Computers When the question requesting the aforementioned disaster countermeasure information is obtained as text data, a classification item determination command creation step is performed to create a classification item determination command using the classification item-specific disaster prevention information, which is a classification of the aforementioned disaster prevention activities into classification items in advance, and the aforementioned question. A reference information acquisition step involves searching the disaster countermeasures database for at least one of similar documents and similar data based on the classification items obtained by AI processing based on the aforementioned instruction statement for determining classification items, and obtaining reference information containing the similar documents or similar data with high similarity. A question text modification instruction step, which creates a question text modification instruction using the reference information obtained in the reference information acquisition step and the question text, A step to create an answer creation command, which uses the modified question text obtained by the AI processing based on the aforementioned question text modification command to create an answer creation command for the disaster countermeasure information, A response output step that outputs a response obtained by the AI processing based on the aforementioned response creation command statement. Execute A disaster response information provision system characterized by the following features.
2. The aforementioned computer, Upon obtaining the aforementioned revised question, the process involves creating a command statement for determining reclassification items, which uses the disaster prevention information by classification item and the revised question. Based on the command statement for determining the reclassification items and the reclassification items obtained by the AI processing, the re-reference information acquisition step involves searching the disaster countermeasures database for at least one of the most similar sentences and most similar data, and obtaining the most similar sentences or data with high similarity as re-reference information. Execute, In the step of creating the command statement for generating the answer, the command statement for generating the answer is created including the re-reference information. The disaster response information provision system according to feature 1.
3. If the aforementioned question is based on the disaster response information according to the Fire Service Act or Fire Service Manual, In the aforementioned reference information acquisition step, a literature reference database is used as the disaster countermeasures database. If the aforementioned question is the disaster response information based on the situation at the disaster site, In the aforementioned reference information acquisition step, the field information database is used as the disaster response database. The disaster response information provision system according to feature 1.
4. The aforementioned question is in audio format, and the aforementioned text data is a text conversion from the audio. The disaster response information provision system according to feature 1.
5. The aforementioned field information database includes: Personnel information including personal information of personnel, personnel location information, or personnel vital information, And, Personal information of persons in need of rescue, location information of persons in need of rescue, or information of persons in need of rescue including triage information The information includes either of the following, along with time information based on the time of occurrence or input time. The disaster response information provision system according to feature 3.
Citation Information
Patent Citations
Friction false twisting processing method
JP1988028924A
Urgent communication command facility
JP2002203289A
Disaster support systems and programs
JP6994594B1