Community service intelligent management system based on big data

Abstract

The invention discloses an intelligent community service management system based on big data, relates to the technical field of community service management, and realizes breakthrough improvement through multi-dimensional dynamic perception-self-evolution decision chain design: a cold start unit triggers the highest-level response at the moment of message release based on extreme event marks and geofences, and sends the highest-level response to a server; fatal delay of waiting for heat accumulation in a traditional scheme is thoroughly avoided; the space-time sensitivity quantification module adaptively adjusts the space-time contribution weight through a dynamic trade-off coefficient, and ensures that the night low-activity period does not report mistakenly and the information surge period does not miss; in the disaster evolution graph, a space-time attenuation model of node urgency degree coupling replaces an artificial rule, and a disaster diffusion path is automatically identified; and the multi-modal verification module calls an official disaster situation API and image recognition to block a false information propagation chain.

Description

Technical Field

[0001] The present invention relates to the technical field of community service management, and in particular to a community service intelligent management system based on big data. Background Art

[0002] Currently, the intelligent management system for community services is widely used in the public welfare field. The system integrates natural language processing (NLP) and basic recommendation algorithms to classify and prioritize massive amounts of UGC content. The current mainstream solutions mostly use a dissemination heat model to identify key help-seeking information by calculating comprehensive scores based on dimensions such as message forwarding volume, keyword matching, and publisher influence. Some solutions introduce graph neural networks to mine users' social relationship chains to predict information diffusion paths.

[0003] In actual applications, the above solution relies on popularity indicators such as the number of forwardings, resulting in newly released emergency requests needing to accumulate sufficient dissemination before reaching the administrator, missing the golden response period; at the same time, keyword matching is easily affected by dialects and abbreviated expressions; graph neural networks require a preset social graph and cannot handle temporarily formed emergency communities.

[0004] To address the above issues, some solutions have added synonym libraries to cover dialect words, but this cannot solve the problem of information cold start; some solutions have combined image / positioning data to assist in judgment, but network interruptions during disasters lead to high multimedia transmission failure rates. It can be seen that such solutions are still limited by static rule frameworks and cannot dynamically adapt to the complex information environment of evolving disaster situations. In essence, they have not broken through the logical shackles of post-event response. Summary of the Invention

[0005] In view of the above existing problems, the present invention is proposed.

[0006] The present invention provides a community service intelligent management system based on big data to solve the problems of existing systems relying on static rules and dissemination popularity, delayed response to key information in disasters and high misjudgment rate, and no adaptation solutions for temporary communities.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] The embodiment of the present invention provides a community service intelligent management system based on big data, which includes:

[0009] Multi-source data collection module, real-time acquisition of text messages, user location coordinates, message publishing timestamps, and attachment type identifiers within the community;

[0010] a spatiotemporal sensitivity quantification module, connected to the multi-source data acquisition module, generating an urgency score based on the geographic entity recognition results, temporal keyword density, and attachment type in the message;

[0011] The disaster evolution map construction module dynamically aggregates message nodes with adjacent geographical locations and urgency scores reaching a threshold, generating a regional disaster topology map with evolution direction markers.

[0012] The resource dynamic matching engine is connected to the disaster evolution map construction module, analyzes the material demand characteristics in the node and outputs the supplier matching plan.

[0013] As a preferred solution of the big data-based community service intelligent management system described in the present invention, the spatiotemporal sensitivity quantification module performs the following operations:

[0014] The geographical entity names in the message are extracted through named entity recognition technology, and associated with the preset administrative level-disaster type two-dimensional weight table;

[0015] Count the frequency of occurrence of keywords in the preset time-sensitive vocabulary in the message;

[0016] A weighted function is used to calculate the urgency score.

[0017] As a preferred solution of the big data-based community service intelligent management system described in the present invention, the step of calculating the urgency score using a weighted function in the spatiotemporal sensitivity quantification module includes:

[0018] After entity extraction and time word statistics, the collected single message enters the following calculation chain in sequence, and finally obtains the urgency score S j ,include:

[0019] Calculate geographic entity weight coefficients,

[0020]

[0021] Among them, i represents the sequence number of the geographic entity in the message, n j represents the total number of geographic entities contained in the jth message, α i,j If the i-th entity appears, it takes 1, otherwise it takes 0. Represents entity g i The corresponding administrative level weight, Represents entity g i The corresponding disaster type weight, g i represents the i-th geographic entity identifier, w g represents the geographical weight coefficient after aggregation;

[0022] Calculate the time keyword frequency factor,

[0023]

[0024] Where β represents the time weight scaling coefficient, c tIndicates the number of times the time-sensitive word appears in the message, l j Indicates the total number of words in the message, γ represents the time weight decay exponent, 0<γ≤1, f t Indicates the time keyword frequency factor;

[0025] Calculate basic urgency compensation,

[0026] E0=λ a +λ m ,

[0027] Among them, λ a represents the risk weight of the attachment type, λ m represents the message mode risk weight, E0 represents the basic urgency compensation amount,

[0028] Calculate the dynamic trade-off coefficient,

[0029]

[0030] Where κ is the density slope constant, d j Indicates the message density of the same group in the past ten minutes, μ d represents the median of historical benchmark message density, τ represents the spatiotemporal coupling dynamic trade-off coefficient;

[0031] Compute the urgency score map,

[0032] S j =τtanh(w g f t )+(1-τ)E0,

[0033] Among them, S j represents the final urgency score of the j-th message.

[0034] As a preferred solution of the big data-based community service intelligent management system described in the present invention, the disaster evolution graph construction module includes:

[0035] The geographic clustering unit aggregates messages of the same geographic entity into event nodes and labels the nodes with initial urgency scores;

[0036] Cross-node association unit,calculates the spatiotemporal association strength between adjacent nodes.

[0037] As a preferred solution of the big data-based community service intelligent management system described in the present invention, in the disaster evolution graph construction module, the calculation of the spatiotemporal correlation strength of adjacent nodes includes:

[0038] When the geographic clustering unit generates an event node, the cross-node association unit determines the association strength A between adjacent nodes p and q according to the following closed-loop calculation chain: pq;

[0039] Calculate the time decay factor using the formula:

[0040]

[0041] Where, Δt pq Indicates the difference in timestamps between the first message of node p and node q. represents the time difference attenuation coefficient, θ pq represents the attenuation factor at the time level;

[0042] Where:

[0043]

[0044] Among them, θ0 represents the time decay reference constant, θ1 represents the urgency coupling amplification constant, S p represents the urgency score of node p, S q represents the urgency score of node q, represents the time difference attenuation coefficient;

[0045] The spatial attenuation factor is defined as:

[0046]

[0047] Among them, d pq represents the spherical distance between node p and node q, represents the geographic distance attenuation coefficient, ψ pq represents the attenuation factor at the spatial level;

[0048] Where:

[0049]

[0050] Among them, ρ0 represents the spatial attenuation reference constant, ρ1 represents the distance normalized gain constant, and d pq represents the spherical distance between nodes, R represents the average radius of the earth, represents the geographic distance attenuation coefficient;

[0051] Calculate the spherical distance using the formula:

[0052]

[0053] Among them, Δφ represents the latitude difference, Δλ represents the longitude difference, and φ p represents the latitude of node p, φ q represents the latitude of node q, R represents the average radius of the earth, d pq represents the spherical distance;

[0054] Comprehensive association strength mapping:

[0055]

[0056] Among them, S p represents the urgency score of node p, S q represents the urgency score of node q, θ pq represents the time decay factor, ψ pq represents the spatial attenuation factor, A pq Represents the comprehensive spatiotemporal correlation strength of adjacent nodes.

[0057] As a preferred solution of the big data-based community service intelligent management system described in the present invention, the resource dynamic matching engine executes:

[0058] Parse the material demand phrases in the event nodes through the entity relationship extraction model and output structured demand tuples;

[0059] Connect to the real-time inventory database of the material dispatching platform to filter out suppliers whose transportation radius is smaller than the set value;

[0060] Generate a priority-sorted matching list and push it to the community management terminal.

[0061] As a preferred solution of the big data-based community service intelligent management system described in the present invention, the resource dynamic matching engine further includes a conflict resolution unit. When multiple event nodes require similar materials:

[0062] Prioritize nodes with higher urgency scores;

[0063] Initiate cross-regional inventory transfer agreement verification process for remaining demand.

[0064] As a preferred solution of the big data-based community service intelligent management system described in the present invention, the system further includes a multimodal verification module, the input end of which is connected to the disaster evolution graph construction module, and is used to:

[0065] Call the disaster notification API interface to verify the disaster status of the geographic entity in the event node;

[0066] When a node contains an image attachment, a pre-trained disaster feature detection model is used to identify damaged buildings, medical supplies, or abnormal natural features in the image.

[0067] As a preferred solution of the big data-based community service intelligent management system described in the present invention, the system further includes a hierarchical response module, the input end of which is connected to the spatiotemporal sensitivity quantification module, and executes:

[0068] The urgency score is mapped to a four-level response label, where the first-level label corresponds to the highest score interval;

[0069] Configure a mandatory display policy for the first-level tag, including pinning it to the top of the community chat interface and sending pop-up reminders to administrators every 5 minutes.

[0070] As a preferred solution of the big data-based community service intelligent management system described in the present invention, the hierarchical response module is further configured with a cold start determination unit, which directly assigns a first-level response label to a new message when it meets any of the following conditions:

[0071] The text content includes the earthquake magnitude value and the flood water level exceeding warning mark symbol;

[0072] The sender's location is within the red alert polygon geofence issued by the meteorological department.

[0073] The beneficial effects of the present invention are as follows: the present invention achieves a breakthrough improvement through the design of a multi-dimensional dynamic perception-self-evolution decision chain: the cold start unit triggers the highest level of response at the moment of message release based on extreme event markers and geographic fences, completely avoiding the fatal delay of traditional solutions waiting for popularity accumulation; the spatiotemporal sensitivity quantification module adaptively adjusts the spatiotemporal contribution weights through dynamic trade-off coefficients to ensure no false alarms during low activity periods at night and no omissions during periods of information surge; the disaster evolution map replaces manual rules with a spatiotemporal attenuation model coupled with node urgency to automatically identify disaster diffusion paths; the multimodal verification module calls the official disaster information API and image recognition to block the chain of false information dissemination;

[0074] In addition, a resource matching engine is used to parse material demands into structured tuples, and a conflict resolution unit dynamically allocates inventory based on node urgency scores to resolve the deadlock of multiple locations scrambling for similar materials. Cross-regional inventory adjustment agreements verify and activate backup supply chains to address the risk of local resource depletion.

[0075] In summary, the present invention establishes an adaptive disaster response logic in the three links of information capture, analysis, and response, which fundamentally eliminates the systemic defects of post-response in the background technology, and is particularly suitable for community management scenarios of sudden disasters such as earthquakes and floods. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0077] Figure 1 This is a schematic diagram of the framework of the big data-based community service intelligent management system in Example 1. DETAILED DESCRIPTION

[0078] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0079] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0080] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0081] Example 1, with reference to Figure 1 This embodiment provides a community service intelligent management system based on big data, including:

[0082] Multi-source data collection module, real-time acquisition of text messages, user location coordinates, message publishing timestamps, and attachment type identifiers within the community;

[0083] The spatiotemporal sensitivity quantification module connects to the multi-source data collection module and generates an urgency score based on the geographic entity recognition results, temporal keyword density, and attachment type in the message;

[0084] The spatiotemporal sensitivity quantization module performs the following operations:

[0085] The geographical entity names in the message are extracted through named entity recognition technology, and associated with the preset administrative level-disaster type two-dimensional weight table;

[0086] Count the frequency of occurrence of keywords in the preset time-sensitive vocabulary in the message;

[0087] A weighted function is used to calculate the urgency score;

[0088] In the spatiotemporal sensitivity quantification module, the steps of calculating the urgency score using a weighted function include:

[0089] After entity extraction and time word statistics, the collected single message enters the following calculation chain in sequence, and finally obtains the urgency score S j ,include:

[0090] Calculate geographic entity weight coefficients,

[0091]

[0092] Among them, i represents the sequence number of the geographic entity in the message, n j represents the total number of geographic entities contained in the jth message, α i,j If the i-th entity appears, it takes 1, otherwise it takes 0. Represents entity g i The corresponding administrative level weight, Represents entity g i The corresponding disaster type weight, g i represents the i-th geographic entity identifier, w g represents the geographical weight coefficient after aggregation;

[0093] Calculate the time keyword frequency factor,

[0094]

[0095] Where β represents the time weight scaling coefficient, c t Indicates the number of times the time-sensitive word appears in the message, l j Indicates the total number of words in the message, γ represents the time weight decay exponent, 0<γ≤1, f t Indicates the time keyword frequency factor;

[0096] Calculate basic urgency compensation,

[0097] E0=λ a +λ m ,

[0098] Among them, λ a represents the risk weight of the attachment type, λ m represents the message mode risk weight, E0 represents the basic urgency compensation amount,

[0099] Calculate the dynamic trade-off coefficient,

[0100]

[0101] Where κ is the density slope constant, d j Indicates the message density of the same group in the past ten minutes, μ d represents the median of historical benchmark message density, τ represents the spatiotemporal coupling dynamic trade-off coefficient;

[0102] Compute the urgency score map,

[0103] S j =τtanh(w g f t )+(1-τ)E0,

[0104] Among them, S j represents the final urgency score of the j-th message;

[0105] Specifically, the scoring chain first aggregates the levels of geographical entities and disaster associations to form w g , highlighting the spatial focus, and then mapping the time word frequency with the power exponential model to obtain f t , amplifying the outbreak trend of events, the attachment and message modes inject multimodal risk gain through E0, and the Sigmoid trade-off τ driven by the density of messages in the same group adaptively adjusts the spatiotemporal contribution, so that the system does not falsely alarm during the low activity period at night and quickly raises the score during the information surge period;

[0106] The disaster evolution map construction module dynamically aggregates message nodes with adjacent geographical locations and urgency scores reaching a threshold, generating a regional disaster topology map with evolution direction markers.

[0107] The disaster evolution map construction modules include:

[0108] The geographic clustering unit aggregates messages of the same geographic entity into event nodes and labels the nodes with initial urgency scores;

[0109] Cross-node association unit, calculates the spatiotemporal association strength between adjacent nodes;

[0110] In the disaster evolution graph construction module, the calculation of the spatiotemporal correlation strength of adjacent nodes includes:

[0111] When the geographic clustering unit generates an event node, the cross-node association unit determines the association strength A between adjacent nodes p and q according to the following closed-loop calculation chain: pq ;

[0112] Calculate the time decay factor using the formula:

[0113]

[0114] Where, Δt pq Indicates the difference in timestamps between the first message of node p and node q. represents the time difference attenuation coefficient, θ pq represents the attenuation factor at the time level;

[0115] Where:

[0116]

[0117] Among them, θ0 represents the time decay reference constant, θ1 represents the urgency coupling amplification constant, S p represents the urgency score of node p, S q represents the urgency score of node q, represents the time difference attenuation coefficient;

[0118] The spatial attenuation factor is defined as:

[0119]

[0120] Among them, d pq represents the spherical distance between node p and node q, represents the geographic distance attenuation coefficient, ψ pq represents the attenuation factor at the spatial level;

[0121] Where:

[0122]

[0123] Among them, ρ0 represents the spatial attenuation reference constant, ρ1 represents the distance normalized gain constant, and d pq represents the spherical distance between nodes, R represents the average radius of the earth, represents the geographic distance attenuation coefficient;

[0124] Calculate the spherical distance using the formula:

[0125]

[0126] Among them, Δφ represents the latitude difference, Δλ represents the longitude difference, and φ p represents the latitude of node p, φ q represents the latitude of node q, R represents the average radius of the earth, d pq represents the spherical distance;

[0127] Comprehensive association strength mapping:

[0128]

[0129] Among them, S p represents the urgency score of node p, S q represents the urgency score of node q, θ pq represents the time decay factor, ψ pq represents the spatial attenuation factor, A pq Represents the comprehensive spatiotemporal correlation strength of adjacent nodes;

[0130] Specifically, the formula chain takes the node urgency coupling as the starting point, maps the time difference and spatial distance through exponential decay, takes into account the propagation speed of emergencies and the geographical diffusion resistance, and the time decay coefficient The urgency weighted structure is adopted to make high-risk nodes more sensitive to time synchronization, and the spatial attenuation coefficient It is linearly amplified with distance to avoid distant nodes being misconnected due to weight dilution. The overall correlation strength A pq Quickly filter out low-coherence edges;

[0131] The resource dynamic matching engine connects to the disaster evolution graph construction module, analyzes the material demand characteristics in the node and outputs the supplier matching solution;

[0132] Resource dynamic matching engine execution:

[0133] Parse the material demand phrases in the event nodes through the entity relationship extraction model and output structured demand tuples (material type, quantity, and timeliness requirements);

[0134] Connect to the real-time inventory database of the material dispatching platform to filter out suppliers whose transportation radius is smaller than the set value;

[0135] Generate a priority-sorted matching list and push it to the community management terminal;

[0136] The resource dynamic matching engine also includes a conflict resolution unit. When multiple event nodes require the same type of resources:

[0137] Prioritize nodes with higher urgency scores;

[0138] Initiate cross-regional inventory transfer agreement verification process for remaining demand;

[0139] The system also includes a multimodal verification module, whose input is connected to the disaster evolution map construction module, which is used to:

[0140] Call the disaster notification API interface to verify the disaster status of the geographic entity in the event node;

[0141] When a node contains an image attachment, it uses a pre-trained disaster feature detection model to identify damaged buildings, medical supplies, or abnormal natural features in the image.

[0142] The system also includes a hierarchical response module, whose input is connected to the spatiotemporal sensitivity quantification module, which performs:

[0143] The urgency score is mapped to a four-level response label, where the first-level label corresponds to the highest score interval;

[0144] Configure mandatory display policies for first-level tags, including pinning them to the top of the community chat interface and sending pop-up reminders to administrators every 5 minutes.

[0145] The hierarchical response module is also equipped with a cold start determination unit, which directly assigns a first-level response tag to a new message if it meets any of the following conditions:

[0146] The text content includes the earthquake magnitude value and the flood water level exceeding warning mark symbol;

[0147] The sender's location is within the red alert polygon geofence issued by the meteorological department.

[0148] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A community service intelligent management system based on big data, characterized by: include, Multi-source data collection module, real-time acquisition of text messages, user location coordinates, message publishing timestamps, and attachment type identifiers within the community; a spatiotemporal sensitivity quantification module, connected to the multi-source data acquisition module, generating an urgency score based on the geographic entity recognition results, temporal keyword density, and attachment type in the message; The disaster evolution map construction module dynamically aggregates message nodes with adjacent geographical locations and urgency scores reaching a threshold, generating a regional disaster topology map with evolution direction markers. The resource dynamic matching engine is connected to the disaster evolution map construction module, analyzes the material demand characteristics in the node and outputs the supplier matching plan.

2. The big data-based community service intelligent management system according to claim 1, characterized in that: The spatiotemporal sensitivity quantification module performs the following operations: The geographical entity names in the message are extracted through named entity recognition technology, and associated with the preset administrative level-disaster type two-dimensional weight table; Count the frequency of occurrence of keywords in the preset time-sensitive vocabulary in the message; A weighted function is used to calculate the urgency score.

3. The big data-based community service intelligent management system according to claim 2, characterized in that: In the spatiotemporal sensitivity quantification module, the step of calculating the urgency score using a weighted function includes: After entity extraction and time word statistics, the collected single message enters the following calculation chain in sequence, and finally obtains the urgency score S j ,include: Calculate geographic entity weight coefficients, Among them, i represents the sequence number of the geographic entity in the message, n j represents the total number of geographic entities contained in the jth message, α i,j If the i-th entity appears, it takes 1, otherwise it takes 0. Represents entity g i The corresponding administrative level weight, Represents entity g i The corresponding disaster type weight, g i represents the i-th geographic entity identifier, w g represents the geographical weight coefficient after aggregation; Calculate the time keyword frequency factor, Where β represents the time weight scaling coefficient, c t Indicates the number of times the time-sensitive word appears in the message, l j Indicates the total number of words in the message, γ represents the time weight decay exponent, 0<γ≤1, f t Indicates the time keyword frequency factor; Calculate basic urgency compensation, E0=λ a +λ m , Among them, λ a represents the risk weight of the attachment type, λ m represents the message mode risk weight, E0 represents the basic urgency compensation amount, Calculate the dynamic trade-off coefficient, Where κ is the density slope constant, d j Indicates the message density of the same group in the past ten minutes, μ d represents the median of historical benchmark message density, τ represents the spatiotemporal coupling dynamic trade-off coefficient; Compute the urgency score map, S j =τtanh(w g f t )+(1-τ)E0, Among them, S j represents the final urgency score of the j-th message.

4. The big data-based community service intelligent management system according to claim 1, characterized in that: The disaster evolution graph construction module includes: The geographic clustering unit aggregates messages of the same geographic entity into event nodes and labels the nodes with initial urgency scores; Cross-node association unit,calculates the spatiotemporal association strength between adjacent nodes.

5. The big data-based community service intelligent management system according to claim 4, characterized in that: In the disaster evolution graph construction module, the calculation of the spatiotemporal correlation strength of adjacent nodes includes: When the geographic clustering unit generates an event node, the cross-node association unit determines the association strength A between adjacent nodes p and q according to the following closed-loop calculation chain: pq ; Calculate the time decay factor using the formula: Where Δt pq Indicates the difference in timestamps between the first message of node p and node q. represents the time difference attenuation coefficient, θ pq represents the attenuation factor at the time level; Where: in, represents the time decay reference constant, represents the urgency coupling amplification constant, S p represents the urgency score of node p, S q represents the urgency score of node q, represents the time difference attenuation coefficient; The spatial attenuation factor is defined as: Among them, d pq represents the spherical distance between node p and node q, represents the geographic distance attenuation coefficient, ψ pq represents the attenuation factor at the spatial level; Where: Among them, ρ0 represents the spatial attenuation reference constant, ρ1 represents the distance normalized gain constant, and d pq represents the spherical distance between nodes, R represents the average radius of the earth, represents the geographic distance attenuation coefficient; Calculate the spherical distance using the formula: Among them, Δφ represents the latitude difference, Δλ represents the longitude difference, and φ p represents the latitude of node p, φ q represents the latitude of node q, R represents the average radius of the earth, d pq represents the spherical distance; Comprehensive association strength mapping: Among them, S p represents the urgency score of node p, S q represents the urgency score of node q, θ pq represents the time decay factor, ψ pq represents the spatial attenuation factor, A pq Represents the comprehensive spatiotemporal correlation strength of adjacent nodes.

6. The big data-based community service intelligent management system according to claim 1, characterized in that: The resource dynamic matching engine performs: Parse the material demand phrases in the event nodes through the entity relationship extraction model and output structured demand tuples; Connect to the real-time inventory database of the material dispatching platform to filter out suppliers whose transportation radius is smaller than the set value; Generate a priority-sorted matching list and push it to the community management terminal.

7. The big data-based community service intelligent management system according to claim 6, characterized in that: The resource dynamic matching engine also includes a conflict resolution unit. When multiple event nodes require the same type of materials: Prioritize nodes with higher urgency scores; Initiate cross-regional inventory transfer agreement verification process for remaining demand.

8. The big data-based community service intelligent management system according to claim 7, characterized in that: The system also includes a multimodal verification module, the input end of which is connected to the disaster evolution graph construction module, and is used to: Call the disaster notification API interface to verify the disaster status of the geographic entity in the event node; When a node contains an image attachment, a pre-trained disaster feature detection model is used to identify damaged buildings, medical supplies, or abnormal natural features in the image.

9. The big data-based community service intelligent management system according to claim 8, characterized in that: The system further includes a hierarchical response module, the input of which is connected to the spatiotemporal sensitivity quantification module, and executes: The urgency score is mapped to a four-level response label, where the first-level label corresponds to the highest score interval; Configure a mandatory display policy for the first-level tag, including pinning it to the top of the community chat interface and sending pop-up reminders to administrators every 5 minutes.

10. The big data-based community service intelligent management system according to claim 9, characterized in that: The hierarchical response module is further configured with a cold start determination unit, which directly assigns a first-level response tag to a new message when it meets any of the following conditions: The text content includes earthquake magnitude values ​​and flood water level exceeding warning mark symbols; The sender's location is within the red alert polygon geofence issued by the meteorological department.

Images