Flood disaster condition analysis method and related device

Abstract

The application provides a flood disaster analysis method, comprising: determining a disaster area, obtaining a social media data set in a preset time range from an internet multi-source search platform by setting a theme keyword and disaster sub-area position information in the disaster area, and constructing a social media data position pointing network according to the social media data set; obtaining remote sensing image data of the disaster area, combining the social media data position pointing network and the remote sensing image data of the disaster area, and analyzing the disaster degree of the disaster sub-area in the disaster area. The application can construct a social media data position pointing network by obtaining social media data, can evaluate and analyze the disaster degree of the disaster sub-area in the disaster area in detail by combining the social media data position pointing network and the remote sensing image data of the disaster area, and can continuously monitor the disaster situation of the disaster sub-area in the disaster area.

Description

Technical Field

[0001] This invention relates to the field of flood disaster assessment and analysis technology, and in particular to a flood disaster analysis method and related equipment. Background Technology

[0002] In recent years, abnormal global climate change has significantly increased the frequency of torrential rains and floods, causing numerous casualties and property losses, and severely impacting the sustainable development of human society. With advancements in science and technology, Earth observation technologies, such as satellite remote sensing, have played a crucial role in flood disaster monitoring. Satellite remote sensing offers a wide monitoring range and can analyze disaster conditions by comparing imaging changes before and after a disaster, providing effective information for disaster assessment and disaster relief. However, this technology also has limitations. For example, the long revisit period of satellites makes continuous monitoring of affected areas difficult. Satellite remote sensing primarily provides a macroscopic description of the disaster situation, such as determining the extent of inundation, but it struggles to further understand the extent of damage within the inundated areas and to specifically assess which regions are most severely affected. With the development of the internet and the widespread use of smart mobile devices, crowdsourced data, represented by social media, has brought new opportunities for disaster mitigation. Compared to remote sensing data, social media provides detailed descriptions of the disaster from the perspective of the affected entities themselves, effectively supplementing remote sensing imagery data.

[0003] Currently, many researchers use social media data in disaster mitigation research, including combining social media data with remote sensing imagery for disaster assessment. However, related research is insufficient, particularly regarding the lack of spatial information in social media data. When floods occur, the extent of damage in inundated areas often limits the distribution of social media data, making it difficult to obtain information about the damage in these areas. Often, no social media data exists in these regions, hindering effective disaster assessment and analysis. Summary of the Invention

[0004] This invention provides a method and related equipment for analyzing flood disasters, which can at least partially solve the above-mentioned technical problems.

[0005] This invention provides a method for analyzing flood disaster situations, including:

[0006] The affected areas are identified, and social media datasets within a preset time range are obtained from a multi-source internet search platform by setting topic keywords and the location information of the affected sub-areas within the affected areas. A social media data location pointing network is constructed based on the social media datasets.

[0007] Remote sensing image data of the disaster-stricken area is acquired, and combined with the location-pointing network of social media data and the remote sensing image data of the disaster-stricken area, the degree of disaster in the sub-regions of the disaster-stricken area is analyzed.

[0008] Optionally, constructing a social media data location pointing network based on the social media dataset includes:

[0009] Based on a pre-trained deep learning model, the location information of the disaster-stricken sub-region corresponding to each social media data point is extracted from the social media dataset;

[0010] Based on the location information of the disaster-affected sub-region corresponding to each of the social media data, and the location information of the data acquisition region corresponding to each of the social media data, a location pointing relationship is obtained;

[0011] Based on the aforementioned location pointing relationship, a social media data location pointing network is constructed.

[0012] Optionally, obtaining the location pointing relationship based on the location information of the disaster-stricken sub-region corresponding to each of the social media data and the location information of the data acquisition area corresponding to each of the social media data includes:

[0013] Using the location information of the disaster-stricken sub-region as the first type of node division standard, a first node corresponding to each social media data is determined based on the location information of the disaster-stricken sub-region corresponding to each social media data, wherein the first type of node includes multiple first nodes;

[0014] Using the data acquisition area location information as the second type of node division standard, and based on the data acquisition area location information corresponding to each social media data, a second node corresponding to each social media data is determined, wherein the second type of node includes multiple second nodes;

[0015] Based on the first node corresponding to each piece of social media data and the second node corresponding to each piece of social media data, at least one second node that has a pointing association with the first node is determined, and a position pointing relationship between the first node and the second node is established, wherein the first node is the node being pointed to.

[0016] Optionally, modeling the location pointing relationship and constructing a social media data location pointing network includes:

[0017] Based on the positional relationship between the first node and the second node, determine the weight of the pointing edge between the first node and the second node.

[0018] Based on the location information of the disaster-affected sub-region corresponding to each piece of social media data, the number of social media data included in each second node is obtained, and the weight of each second node is determined.

[0019] The weight of the first node is determined based on at least one second node that has a pointing association with the first node, and the weight of the pointing edge between the first node and the second node.

[0020] A social media data location pointing network is constructed based on the weights of the pointing edges between the first node and the second node, the weight degree of the first node, and the weight degree of the second node.

[0021] Optionally, the step of acquiring remote sensing image data of the disaster-stricken area, combining the social media data location pointing network and the remote sensing image data of the disaster-stricken area, and analyzing the degree of disaster in the affected cities within the disaster-stricken area includes:

[0022] Remote sensing image data of the disaster-stricken area is acquired, and combined with the location pointing network of social media data and the remote sensing image data of the disaster-stricken area, the degree of disaster in the disaster-stricken sub-regions within the disaster-stricken area is analyzed by spatial overlay.

[0023] Optionally, the method further includes:

[0024] Based on the location information of the social media data pointing to the network and the location information of the affected sub-regions within the disaster area, a subset of social media data pointing to the affected sub-regions within the disaster area is determined;

[0025] By analyzing the subset of social media data pointing to the disaster-stricken sub-regions within the disaster-stricken area, the thematic keywords corresponding to the disaster-stricken sub-regions within the disaster-stricken area in multiple time ranges are obtained;

[0026] Based on the thematic keywords corresponding to the affected sub-regions within the disaster-stricken area at multiple time intervals, the disaster situation of the affected sub-regions within the disaster-stricken area at multiple time intervals is obtained.

[0027] The present invention also provides a flood disaster analysis device, comprising:

[0028] The module is used to determine the disaster-stricken area. By setting the theme keywords and the location information of the disaster-stricken sub-areas within the disaster-stricken area, it obtains social media datasets within a preset time range from the Internet multi-source search platform, and constructs a social media data location pointing network based on the social media datasets.

[0029] The analysis module is used to acquire remote sensing image data of the disaster-stricken area, and combine the location pointing network of social media data with the remote sensing image data of the disaster-stricken area to analyze the degree of disaster in the disaster-stricken sub-areas within the disaster-stricken area.

[0030] Optionally, the building module includes:

[0031] An extraction submodule is used to extract the location information of the disaster-stricken sub-region corresponding to each social media data from the social media dataset based on a pre-trained deep learning model;

[0032] The first construction submodule is used to obtain the location relationship based on the location information of the disaster-stricken sub-region corresponding to each of the social media data and the data acquisition area information corresponding to each of the social media data.

[0033] The second construction submodule is used to model based on the location pointing relationship and construct a social media data location pointing network.

[0034] 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, wherein the processor executes the program to implement the flood disaster analysis method described above.

[0035] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the flood disaster analysis method as described above.

[0036] The flood disaster analysis method and related equipment provided by this invention construct a social media data location pointing network by acquiring social media data. By combining the social media data location pointing network with remote sensing image data of the disaster-stricken area, it is possible to conduct a detailed assessment and analysis of the disaster severity of the disaster-stricken sub-regions within the disaster-stricken area, and to continuously monitor the disaster situation of the disaster-stricken sub-regions within the disaster-stricken area. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0038] Figure 1 This is one of the flowcharts illustrating the flood disaster analysis method provided by the present invention;

[0039] Figure 2This is a structural diagram of the social media data location pointing to the network provided by the present invention;

[0040] Figure 3 This is a schematic diagram of flood disaster analysis based on the location pointing to a network of social media data provided by the present invention;

[0041] Figure 4 This is a schematic diagram of continuous disaster monitoring in disaster-stricken areas, provided by the present invention, which integrates social media data;

[0042] Figure 5 This is the second flowchart illustrating the flood disaster analysis method provided by the present invention;

[0043] Figure 6 This is a schematic diagram of the flood disaster analysis device provided by the present invention;

[0044] Figure 7 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0046] Figure 1 This is one of the flowcharts illustrating the flood disaster analysis method provided by this invention, such as... Figure 1 As shown, the present invention provides a method for analyzing flood disaster situation, including steps 110 and 120:

[0047] Step 110: Determine the disaster-stricken area. By setting topic keywords and the location information of the disaster-stricken sub-areas within the disaster-stricken area, obtain social media datasets within a preset time range from the Internet multi-source search platform, and construct a social media data location pointing network based on the social media datasets.

[0048] In this embodiment of the invention, in the event of a flood disaster, the affected area is determined, and based on the affected area, the detailed location information of the affected sub-areas within the affected area can be determined.

[0049] It should be noted that, based on the disaster-stricken area, the location information of multiple disaster-stricken sub-areas can be determined. The location information of the disaster-stricken sub-areas can be set according to the actual location information of the disaster-stricken area, and this invention does not impose any limitations on it.

[0050] The location information of the affected sub-region can be either the domain name of the affected area or the name of the affected city.

[0051] By setting topic keywords and location information of affected sub-regions within the disaster-stricken area, and using a search method of "location information of affected sub-regions + flood-related topic keywords," social media data related to affected sub-regions within the disaster-stricken area can be accurately obtained from multi-source internet search platforms.

[0052] For example, the geographical names of the affected sub-regions within the disaster area are used as the location information of the affected cities. Combined with flood-related keywords such as "flood" and "rainstorm", searches are conducted on multiple internet search platforms, including Sina Weibo and WeChat public accounts, to obtain social media data including geographical names and keywords of the affected sub-regions.

[0053] By acquiring social media datasets within a preset time range from a multi-source internet search platform, the selected disaster-stricken area can be subjected to a period of severe impact from the disaster. This allows for continuous analysis of the disaster situation in the disaster-stricken area within the preset time range.

[0054] Based on deep learning or natural language processing methods, the acquired social media dataset is analyzed to obtain the location information of the disaster-affected sub-regions corresponding to each social media data in the dataset and the geographical location information that the social media data itself carries when it is acquired. Using the location information of the disaster-affected sub-regions corresponding to the social media data as the pointed nodes and the geographical location information that the social media data itself carries when it is acquired as the pointing nodes, a social media data location pointing network is constructed.

[0055] In the embodiments provided by this invention, by extracting the location information of disaster-affected sub-regions from social media data and the geographical location information inherent in the social media data, the shortcomings of the lack of geographical location information in social media data itself are effectively solved. Furthermore, by combining the construction method of complex networks, the pointing relationship between the location information of disaster-affected sub-regions in social media data and the geographical location information inherent in the social media data is modeled, thereby constructing a social media data location pointing network. From the perspective of public observation, the degree of attention received by different disaster-affected sub-regions within the disaster area is quantified, thereby reflecting the degree of disaster in different regions.

[0056] Step 120: Obtain remote sensing image data of the disaster-stricken area, and combine the location pointing network of the social media data with the remote sensing image data of the disaster-stricken area to analyze the degree of disaster in the sub-areas of the disaster-stricken area.

[0057] In this embodiment of the invention, remote sensing image data of the disaster-stricken area is obtained, and the distribution of flood-inundated areas within the disaster-stricken area is obtained based on the flood-inundated area extraction method.

[0058] Based on remote sensing image data of the disaster-stricken area, macro-level disaster information of the flooded areas within the disaster-stricken area is obtained. Combined with location data from social media networks, the local details of the disaster are reflected in detail from the public's perspective. The degree of disaster in the sub-regions within the disaster-stricken area is analyzed, and the disaster situation in the sub-regions within the disaster-stricken area is continuously monitored.

[0059] The flood disaster analysis method and related equipment provided by this invention can construct a social media data location pointing network by acquiring social media data, and combine the social media data location pointing network with remote sensing image data of the disaster-stricken area to conduct a detailed assessment and analysis of the disaster severity of the disaster-stricken sub-regions within the disaster-stricken area, and can also continuously monitor the disaster situation of the disaster-stricken sub-regions within the disaster-stricken area.

[0060] Optionally, constructing a social media data location pointing network based on the social media dataset includes:

[0061] Based on a pre-trained deep learning model, the location information of the disaster-stricken sub-region corresponding to each social media data point is extracted from the social media dataset;

[0062] Based on the location information of the disaster-affected sub-region corresponding to each of the social media data, and the location information of the data acquisition region corresponding to each of the social media data, a location pointing relationship is obtained;

[0063] Based on the aforementioned location pointing relationship, a social media data location pointing network is constructed.

[0064] In this embodiment of the invention, a pre-trained deep learning model is used to automatically label the location information of disaster-affected sub-regions included in social media data, thereby forming the corresponding geographic location nodes of the social media data and obtaining the location information of the disaster-affected sub-regions corresponding to each social media data.

[0065] Based on the geolocation information inherent in the social media data, the location information of the data acquisition area corresponding to each social media data is obtained, and the corresponding geolocation node is obtained.

[0066] The data acquisition area location information can be either the data acquisition region name or the data acquisition city name.

[0067] The location information of the data acquisition area corresponding to each social media data is used as the pointing geographical location node corresponding to the social media data, and the location information of the disaster-affected sub-area corresponding to each social media data is used as the pointing geographical location node corresponding to the social media data, so as to obtain the location pointing relationship between the nodes.

[0068] For example, the location corresponding to social media data is G. c Social media data corresponds to the geographic location G o G c and G o The geographic location node V corresponding to social media data c The corresponding geographic location node V for social media data o The actual geographical location attribute is used to establish connections between nodes through location pointers, and the pointer relationship is: V o →V c .

[0069] Modeling is performed based on the positional pointing relationships between nodes, obtaining the weights of nodes and the weights of the pointing edges between nodes, and constructing a social media data positional pointing network based on the positional pointing relationships between nodes, as well as the weights of nodes and the weights of the pointing edges between nodes.

[0070] Optionally, obtaining the location pointing relationship based on the location information of the disaster-stricken sub-region corresponding to each of the social media data and the location information of the data acquisition area corresponding to each of the social media data includes:

[0071] Using the location information of the disaster-stricken sub-region as the first type of node division standard, a first node corresponding to each social media data is determined based on the location information of the disaster-stricken sub-region corresponding to each social media data, wherein the first type of node includes multiple first nodes;

[0072] Using the data acquisition area location information as the second type of node division standard, and based on the data acquisition area location information corresponding to each social media data, a second node corresponding to each social media data is determined, wherein the second type of node includes multiple second nodes;

[0073] Based on the first node corresponding to each piece of social media data and the second node corresponding to each piece of social media data, at least one second node that has a pointing association with the first node is determined, and a position pointing relationship between the first node and the second node is established, wherein the first node is the node being pointed to.

[0074] In this embodiment of the invention, the location information of the disaster-affected sub-region is used as the standard for dividing the first type of nodes. Based on the location information of the disaster-affected sub-region corresponding to each social media data, the set of location information of the disaster-affected sub-region included in the social media dataset is determined. The location information of the disaster-affected sub-region in the set of location information of the disaster-affected cities is used as the first type of node. The first type of node includes multiple first nodes, and each first node corresponds to a location information of a disaster-affected sub-region. The first node corresponding to each social media data is determined.

[0075] For example, the disaster-affected sub-regions corresponding to social media data are used as the standard for dividing the first type of nodes. Disaster-affected place names are extracted from the social media data in the social media dataset. Each disaster-affected place name corresponds to multiple social media data. The disaster-affected place name is used as the first node. Based on the disaster-affected place name corresponding to each social media data, the first node corresponding to each social media data is determined.

[0076] The location information of the data acquisition area of ​​social media data is used as the second type of node division standard. Based on the location information of the data acquisition area corresponding to each social media data, the set of location information of the data acquisition area included in the social media dataset is determined. The location information of the data acquisition area in the set of location information of the data acquisition area is used as the second type of node. The second type of node includes multiple second nodes, and each second node corresponds to a location information of the data acquisition area. The second node corresponding to each social media data is determined.

[0077] For example, from the geolocation information inherent in social media data, the publishing region domain name of each social media data is used as the data acquisition area location information of the social media data. The publishing region domain name of each social media data is extracted from the social media data in the social media dataset. Each publishing region includes multiple social media data. The publishing region name of the social media is used as the second node. Based on the publishing region domain name corresponding to each social media data, the second node corresponding to each social media data is determined.

[0078] Obtain the first node and second node corresponding to each social media data in the social media dataset. With the first node as the pointed-to node and the second node as the pointing node, establish the pointing relationship between nodes based on the first node and second node corresponding to each social media data. Since the first node corresponds to multiple social media data and the second node includes multiple social media data, there must be at least one second node that has a pointing relationship with the first node.

[0079] It should be noted that each piece of social media data can determine the relationship between the first node and the second node; that is, each piece of social media data identifies the relationship between the first node and the second node corresponding to the social media data.

[0080] Optionally, modeling the location pointing relationship and constructing a social media data location pointing network includes:

[0081] Based on the positional relationship between the first node and the second node, determine the weight of the pointing edge between the first node and the second node.

[0082] Based on the location information of the disaster-affected sub-region corresponding to each piece of social media data, the number of social media data included in each second node is obtained, and the weight of each second node is determined.

[0083] The weight of the first node is determined based on at least one second node that has a pointing association with the first node, and the weight of the pointing edge between the first node and the second node.

[0084] A social media data location pointing network is constructed based on the weights of the pointing edges between the first node and the second node, the weight degree of the first node, and the weight degree of the second node.

[0085] In this embodiment of the invention, a quantifiable association relationship between nodes is constructed based on the positional pointing relationship between the first node and the second node, and the weight corresponding to the pointing edge between the first node and the second node is determined based on the number of social media data pointing from the second node to the first node.

[0086] For example, the weight of the edge between nodes is W. o_c Its formula is as follows:

[0087]

[0088] Where, N o_c For the second node V o Pointing to the first node V c The number of social media data, where N is the total number of social media data in the social media dataset.

[0089] Based on the location information of the disaster-affected sub-region corresponding to each social media data, a quantifiable node size representation index is constructed to obtain the second node corresponding to each social media data, and then the number of social media data included in each second node is obtained to determine the weight of each second node.

[0090] The weight of the first node is determined based on at least one second node that has a pointing association with the first node, and the weights of the pointing edges between the first and second nodes.

[0091] For example, the formula for calculating the weight is as follows:

[0092]

[0093]

[0094] Among them, D o For the second node V o The weighting, N o_o This indicates the geographic location G corresponding to social media data. o The amount of social media data obtained from D c For the first node V c The weight of the pointer is the value pointing to the first node V. c The sum of weights corresponding to all pointing edges, where i is the index corresponding to the location information of the data acquisition area in the data acquisition area location information set.

[0095] Based on the weights of the pointing edges between the first and second nodes, the weight degree of the first node, and the weight degree of the second node, a social media data location pointing network is constructed.

[0096] It should be noted that the weight of the pointing edge between the first node and the second node determines the thickness of the pointing edge line in the social media data location pointing network. The greater the weight of the pointing edge, the thicker the pointing edge line in the social media data location pointing network.

[0097] The weight of the first and second nodes determines the size of their node identifiers in the social media data location pointing network. The greater the weight of a node, the larger its identifier in the social media data location pointing network.

[0098] Figure 2 This is a structural diagram of the social media data location pointing to the network provided by the present invention, such as... Figure 2 As shown, the greater the weight of a node, the larger its corresponding node identifier. In the diagram, the circle represents the second node, and the square represents the first node. A greater weight for the second node indicates that it contains more social media data, suggesting higher public activity in the region. A greater weight for the first node indicates that it receives more public attention, indirectly indicating a more severe disaster in the area. The greater the weight of the pointing edges between nodes, the thicker the connections between them, indicating closer relationships between nodes and reflecting the spatial distribution and magnitude of attention received by the first node.

[0099] Optionally, the step of acquiring remote sensing image data of the disaster-stricken area, combining the social media data location pointing network and the remote sensing image data of the disaster-stricken area, and analyzing the degree of disaster in the affected cities within the disaster-stricken area includes:

[0100] Remote sensing image data of the disaster-stricken area is acquired, and combined with the location pointing network of social media data and the remote sensing image data of the disaster-stricken area, the degree of disaster in the disaster-stricken sub-regions within the disaster-stricken area is analyzed by spatial overlay.

[0101] In this embodiment of the invention, remote sensing image data of the disaster-stricken area is acquired, and based on the remote sensing image data of the disaster-stricken area and combined with social media data location pointing to the network, the degree of disaster in the sub-areas within the disaster-stricken area is analyzed.

[0102] By using spatial overlay technology of Geographic Information System (GIS), social media data location pointing to the network and remote sensing image data of the disaster area are spatially overlaid to analyze the degree of disaster in the sub-regions of the disaster area.

[0103] In this embodiment of the invention, by using GIS spatial overlay technology, remote sensing image data and social media data location pointing network of the disaster-stricken area are combined. This not only intuitively shows the disaster situation of the disaster-stricken sub-areas within the disaster-stricken area, but also reflects the local details of the disaster from the public's perspective, enabling a better assessment and analysis of the degree of disaster in the disaster-stricken area.

[0104] Figure 3 This is a schematic diagram of flood disaster analysis based on the location-based network of social media data provided by the present invention, as shown in the figure. Figure 3 As shown, combining social media data location data with information extracted from flooded areas in remote sensing imagery data reveals that the remote sensing imagery data reflects the macroscopic extent of the disaster in the affected areas. Figure 3 In the data analysis, the areas where social media data was collected are marked with circles. Due to the flooding, there is very little social media data near the inundated areas, highlighting the limitations of traditional methods that rely on the upload location of social media data to determine flood severity. The affected sub-areas indicated by the social media data are marked with squares; these sub-areas are the areas of greatest public concern. Figure 3 In the image, sub-regions 1 to 7, as shown by the remote sensing image data, all have significant flooded areas, and the public has paid close attention to these sub-regions, indicating that the disaster in these sub-regions is quite severe and has attracted widespread public attention.

[0105] Optionally, the method further includes:

[0106] Based on the location information of the social media data pointing to the network and the location information of the affected sub-regions within the disaster area, a subset of social media data pointing to the affected sub-regions within the disaster area is determined;

[0107] By analyzing the subset of social media data pointing to the disaster-stricken sub-regions within the disaster-stricken area, the thematic keywords corresponding to the disaster-stricken sub-regions within the disaster-stricken area in multiple time ranges are obtained;

[0108] Based on the thematic keywords corresponding to the affected sub-regions within the disaster-stricken area at multiple time intervals, the disaster situation of the affected sub-regions within the disaster-stricken area at multiple time intervals is obtained.

[0109] In this embodiment of the invention, a sub-region severely affected by floods is selected, and a subset of social media data pointing to the sub-region severely affected by floods is obtained based on the location pointing network of social media data.

[0110] The social media data subset is divided according to time range to obtain the social media data time segment subset. Data processing is performed on each social media data time segment subset to obtain the thematic keywords corresponding to the disaster-stricken sub-regions in the disaster area in multiple time ranges.

[0111] Thematic keywords corresponding to the affected sub-regions within the disaster-stricken area are analyzed at multiple time intervals to obtain the disaster situation of the affected sub-regions within the disaster-stricken area at multiple time intervals.

[0112] Figure 4 This is a schematic diagram of continuous disaster monitoring in disaster-stricken areas, provided by the present invention, which integrates social media data. Figure 4 As shown, on July 22, the disaster-stricken cities were severely affected, and relevant departments had launched rescue efforts. The keywords "excavator," "rescue," "dike breach," and "danger" described the specific situation and rescue efforts. From July 23 to July 25, the keywords "injured," "life," "flood control," and "coma" indicated the arduousness of the current disaster relief and rescue work, while the keywords "escape," "improvement," and "transfer" also indicated the effectiveness of the current rescue efforts. From July 26 to July 27, the keywords "sacrifice" and "search and rescue" indicated the specific casualties among rescue personnel during the disaster relief efforts, while the keywords "hero" and "heroic" reflected the public's satisfaction with the disaster relief work and their gratitude to the rescue personnel.

[0113] In this embodiment of the invention, by processing and analyzing social media data within different time ranges, thematic keywords corresponding to disaster-stricken sub-regions within the disaster-stricken area and disaster situations corresponding to disaster-stricken sub-regions within the disaster-stricken area within multiple time ranges are obtained. This allows for continuous analysis of the disaster situation in disaster-stricken sub-regions within the disaster-stricken area within a set time range.

[0114] Figure 5 This is the second flowchart of the flood disaster analysis method provided by the present invention, as shown below. Figure 5 As shown, in another embodiment of the present invention, based on the location information of the affected sub-region, the main keywords such as "flood" and "rainstorm" are combined as search keyword conditions to obtain social media datasets from a multi-source search platform on the Internet.

[0115] Based on deep learning models or natural language processing methods, the social media data in the acquired social media dataset is deduplicated and structured, and the posting time, posting location tags, and text content of the social media data are extracted. The posting time, location tags, and text content of the analyzed social media data are then stored.

[0116] Based on the location tags corresponding to each social media data point, the location information of the data acquisition area corresponding to each social media data point is obtained; based on the text content corresponding to each social media data point, deep learning is performed to obtain the location information of the disaster-affected sub-area corresponding to each social media data point.

[0117] Based on the data corresponding to each social media data point, regional location information and disaster-affected sub-region location information corresponding to each social media data point are obtained, and the geographical location relationship between nodes is constructed.

[0118] Based on the geographic location pointing relationships between nodes, the weight degree of each node is calculated, and the weight of the pointing edges between nodes is calculated. Based on the geographic location pointing relationships between nodes, the weight degree of each node, and the weight of the pointing edges between nodes, the social media data location pointing network is obtained.

[0119] Remote sensing image data of the disaster-stricken area is acquired. Based on the flooded area extraction method, flooded area extraction data is obtained. The flooded area extraction data of the remote sensing image data and the location pointing network of social media data are spatially overlaid.

[0120] Based on the spatial overlay results of data extracted from the flooded areas of remote sensing imagery and location-pointing networks from social media data, the degree of disaster in the affected areas is analyzed, and the analysis results of the degree of disaster in the affected areas are obtained.

[0121] Data processing of social media data across different timeframes yields thematic keywords corresponding to sub-regions within the disaster-stricken area across multiple timeframes, enabling continuous analysis of the disaster situation in the affected region.

[0122] Figure 6 This is a schematic diagram of the flood disaster analysis device provided by the present invention, as shown below. Figure 6 As shown, the flood disaster analysis device provided by the present invention includes:

[0123] Module 610 is used to determine the disaster-stricken area, and obtain social media datasets within a preset time range from a multi-source Internet search platform by setting topic keywords and the location information of the disaster-stricken sub-areas within the disaster-stricken area, and construct a social media data location pointing network based on the social media datasets;

[0124] Analysis module 620 is used to acquire remote sensing image data of the disaster-stricken area, and combine the social media data location pointing network and the remote sensing image data of the disaster-stricken area to analyze the degree of disaster in the disaster-stricken sub-areas within the disaster-stricken area.

[0125] Optionally, the building module includes:

[0126] An extraction submodule is used to extract the location information of the disaster-stricken sub-region corresponding to each social media data from the social media dataset based on a pre-trained deep learning model;

[0127] The first construction submodule is used to obtain the location relationship based on the location information of the disaster-stricken sub-region corresponding to each of the social media data and the data acquisition area information corresponding to each of the social media data.

[0128] The second construction submodule is used to model based on the location pointing relationship and construct a social media data location pointing network.

[0129] Optionally, the first construction submodule is specifically used for:

[0130] Using the location information of the disaster-stricken sub-region as the first type of node division standard, a first node corresponding to each social media data is determined based on the location information of the disaster-stricken sub-region corresponding to each social media data, wherein the first type of node includes multiple first nodes;

[0131] Using the data acquisition area location information as the second type of node division standard, and based on the data acquisition area location information corresponding to each social media data, a second node corresponding to each social media data is determined, wherein the second type of node includes multiple second nodes;

[0132] Based on the first node corresponding to each piece of social media data and the second node corresponding to each piece of social media data, at least one second node that has a pointing association with the first node is determined, and a position pointing relationship between the first node and the second node is established, wherein the first node is the node being pointed to.

[0133] Optionally, modeling the location pointing relationship and constructing a social media data location pointing network includes:

[0134] Based on the positional relationship between the first node and the second node, determine the weight of the pointing edge between the first node and the second node.

[0135] Based on the location information of the disaster-stricken city corresponding to each piece of social media data, the number of social media data included in each second node is obtained, and the weight of each second node is determined.

[0136] The weight of the first node is determined based on at least one second node that has a pointing association with the first node, and the weight of the pointing edge between the first node and the second node.

[0137] A social media data location pointing network is constructed based on the weights of the pointing edges between the first node and the second node, the weight degree of the first node, and the weight degree of the second node.

[0138] Optionally, the analysis module 620 is specifically used for:

[0139] Remote sensing image data of the disaster-stricken area is acquired, and combined with the location pointing network of social media data and the remote sensing image data of the disaster-stricken area, the degree of disaster in the disaster-stricken sub-regions within the disaster-stricken area is analyzed by spatial overlay.

[0140] Optionally, the device further includes a continuous analysis module for...

[0141] Based on the location information of the social media data pointing to the network and the location information of the affected sub-regions within the disaster area, a subset of social media data pointing to the affected sub-regions within the disaster area is determined;

[0142] By analyzing the subset of social media data pointing to the disaster-stricken sub-regions within the disaster-stricken area, the thematic keywords corresponding to the disaster-stricken sub-regions within the disaster-stricken area in multiple time ranges are obtained;

[0143] Based on the thematic keywords corresponding to the affected sub-regions within the disaster-stricken area at multiple time intervals, the disaster situation of the affected sub-regions within the disaster-stricken area at multiple time intervals is obtained.

[0144] It should be noted that the flood disaster situation analysis device provided in the embodiments of the present invention can realize all the method steps implemented in the above-mentioned flood disaster situation analysis method embodiments, and can achieve the same technical effect, which will not be repeated here.

[0145] The flood disaster analysis device provided by this invention constructs a social media data location pointing network by acquiring social media data. By combining the social media data location pointing network with remote sensing image data of the disaster-stricken area, it can conduct a detailed assessment and analysis of the disaster severity of the disaster-stricken sub-regions within the disaster-stricken area, and can continuously monitor the disaster situation of the disaster-stricken sub-regions within the disaster-stricken area.

[0146] Figure 7 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 7 As shown, the electronic device may include a processor 710, a communications interface 720, a memory 730, and a communication bus 740. The processor 710, communications interface 720, and memory 730 communicate with each other via the communication bus 740. The processor 710 can call logical instructions in the memory 730 to execute a flood disaster analysis method. This method includes: determining the disaster-stricken area; obtaining a social media dataset within a preset time range from an internet multi-source search platform by setting topic keywords and the location information of disaster-stricken sub-areas within the disaster-stricken area; constructing a social media data location pointing network based on the social media dataset; obtaining remote sensing image data of the disaster-stricken area; and analyzing the degree of disaster in the disaster-stricken sub-areas within the disaster-stricken area by combining the social media data location pointing network and the remote sensing image data of the disaster-stricken area.

[0147] Furthermore, the logical instructions in the aforementioned memory 730 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0148] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the flood disaster analysis method provided by the above methods. The method includes: determining the disaster-stricken area; obtaining a social media dataset within a preset time range from an Internet multi-source search platform by setting topic keywords and the location information of the disaster-stricken sub-areas within the disaster-stricken area; constructing a social media data location pointing network based on the social media dataset; obtaining remote sensing image data of the disaster-stricken area; and analyzing the degree of disaster in the disaster-stricken sub-areas within the disaster-stricken area by combining the social media data location pointing network and the remote sensing image data of the disaster-stricken area.

[0149] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the flood disaster analysis method provided by the above methods. The method includes: determining the disaster-stricken area; obtaining a social media dataset within a preset time range from an internet multi-source search platform by setting topic keywords and location information of disaster-stricken sub-areas within the disaster-stricken area; constructing a social media data location pointing network based on the social media dataset; obtaining remote sensing image data of the disaster-stricken area; and analyzing the degree of disaster in the disaster-stricken sub-areas within the disaster-stricken area by combining the social media data location pointing network and the remote sensing image data of the disaster-stricken area.

[0150] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0151] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0152] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for analyzing flood disaster situations, characterized in that, include: The disaster-stricken area is determined by setting topic keywords and the location information of the disaster-stricken sub-areas within the disaster-stricken area. Social media datasets within a preset time range are obtained from a multi-source search platform on the Internet. Based on a pre-trained deep learning model, the location information of the disaster-stricken sub-areas corresponding to each social media data is extracted from the social media dataset. Based on the location information of the disaster-stricken sub-areas corresponding to each social media data and the data corresponding to each social media data, the regional location information is obtained to obtain the location pointing relationship. Based on the location pointing relationship, a model is built to construct a social media data location pointing network. The step of obtaining a location pointing relationship based on the location information of the disaster-stricken sub-region corresponding to each social media data and the location information of the data acquisition region corresponding to each social media data includes: using the location information of the disaster-stricken sub-region as a first type of node division standard, determining a first node corresponding to each social media data according to the location information of the disaster-stricken sub-region corresponding to each social media data, wherein the first type of node includes multiple first nodes; using the location information of the data acquisition region as a second type of node division standard, determining a second node corresponding to each social media data according to the location information of the data acquisition region corresponding to each social media data, wherein the second type of node includes multiple second nodes; determining at least one second node that has a pointing association with the first node according to the first node and the second node corresponding to each social media data, and establishing a location pointing relationship between the first node and the second node, wherein the first node is the pointed-to node; Remote sensing image data of the disaster-stricken area is acquired, and combined with the location-pointing network of social media data and the remote sensing image data of the disaster-stricken area, the degree of disaster in the sub-regions of the disaster-stricken area is analyzed.

2. The flood disaster analysis method according to claim 1, characterized in that, The modeling based on the location pointing relationship to construct a social media data location pointing network includes: Based on the positional relationship between the first node and the second node, determine the weight of the pointing edge between the first node and the second node. Based on the location information of the disaster-affected sub-region corresponding to each piece of social media data, the number of social media data included in each second node is obtained, and the weight of each second node is determined. The weight of the first node is determined based on at least one second node that has a pointing association with the first node, and the weight of the pointing edge between the first node and the second node. A social media data location pointing network is constructed based on the weights of the pointing edges between the first node and the second node, the weight degree of the first node, and the weight degree of the second node.

3. The flood disaster analysis method according to claim 1, characterized in that, The process of acquiring remote sensing image data of the disaster-stricken area, combining the location-based social media data with the remote sensing image data of the disaster-stricken area, and analyzing the degree of disaster in the sub-regions within the disaster-stricken area includes: Remote sensing image data of the disaster-stricken area is acquired, and combined with the location pointing network of social media data and the remote sensing image data of the disaster-stricken area, the degree of disaster in the disaster-stricken sub-regions within the disaster-stricken area is analyzed by spatial overlay.

4. The flood disaster analysis method according to claim 1, characterized in that, The method further includes: Based on the location information of the social media data pointing to the network and the location information of the affected sub-regions within the disaster area, a subset of social media data pointing to the affected sub-regions within the disaster area is determined; By analyzing the subset of social media data pointing to the disaster-stricken sub-regions within the disaster-stricken area, the thematic keywords corresponding to the disaster-stricken sub-regions within the disaster-stricken area in multiple time ranges are obtained; Based on the thematic keywords corresponding to the affected sub-regions within the disaster-stricken area at multiple time intervals, the disaster situation of the affected sub-regions within the disaster-stricken area at multiple time intervals is obtained.

5. A flood disaster analysis device, characterized in that, include: A construction module is used to determine the disaster-stricken area. By setting topic keywords and the location information of disaster-stricken sub-regions within the disaster-stricken area, it obtains a social media dataset within a preset time range from a multi-source internet search platform. Based on the social media dataset, it constructs a social media data location pointing network. The construction module includes: an extraction sub-module, used to extract the location information of the disaster-stricken sub-region corresponding to each social media data from the social media dataset based on a pre-trained deep learning model; a first construction sub-module, used to obtain regional location information based on the location information of the disaster-stricken sub-region corresponding to each social media data and the data corresponding to each social media data, thus obtaining a location pointing relationship; and a second construction sub-module, used to model based on the location pointing relationship to construct a social media data location pointing network. The step of obtaining a location pointing relationship based on the location information of the disaster-stricken sub-region corresponding to each social media data and the location information of the data acquisition region corresponding to each social media data includes: using the location information of the disaster-stricken sub-region as a first type of node division standard, determining a first node corresponding to each social media data according to the location information of the disaster-stricken sub-region corresponding to each social media data, wherein the first type of node includes multiple first nodes; using the location information of the data acquisition region as a second type of node division standard, determining a second node corresponding to each social media data according to the location information of the data acquisition region corresponding to each social media data, wherein the second type of node includes multiple second nodes; determining at least one second node that has a pointing association with the first node according to the first node and the second node corresponding to each social media data, and establishing a location pointing relationship between the first node and the second node, wherein the first node is the pointed-to node; The analysis module is used to acquire remote sensing image data of the disaster-stricken area, and combine the location pointing network of social media data with the remote sensing image data of the disaster-stricken area to analyze the degree of disaster in the disaster-stricken sub-areas within the disaster-stricken area.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the flood disaster analysis method as described in any one of claims 1 to 4.

7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the flood disaster analysis method as described in any one of claims 1 to 4.

Images