Sea squirt survey data analysis system and method based on multi-dimensional data acquisition

Abstract

The invention relates to the technical field of data analysis, and particularly discloses a sea squirt survey data analysis system and method based on multi-dimensional data acquisition, which uses an MINE algorithm to perform correlation analysis on various invasive organism data and various environmental factors so as to construct an MIC adjacency matrix between the invasive organism data and the environmental factor data, and carrying out feature extraction and enhancement processing on the adjacent matrix by adopting a data processing technology based on deep learning so as to identify and emphasize key connection between the environmental factors and invasive organisms, thereby intelligently identifying key environmental factors in the environmental factors. In this way, the recognition precision of the key participants in the environmental factors can be effectively improved, and therefore powerful decision support is provided for marine ecology management.

Description

Technical Field

[0001] The present application relates to the field of data analysis technology, and more specifically, to a system and method for analyzing ascidian survey data based on multi-dimensional data collection. Background Art

[0002] Tunicates, also known as sea urchins or sea urchins, are a class of invertebrates widely distributed in the world's oceans. Tunicates play an important role in the ecosystem. Not only do they support the survival of other marine organisms as part of the food chain, but they can also affect the nutrient cycle and bottom structure of the water body. However, some species of tunicates are also considered invasive species, such as the papillary tunicate and the saarbrook tunicate, which can rapidly reproduce and occupy ecological niches in non-native environments, leading to the decline or even extinction of local species, causing serious damage to the marine ecosystem. Therefore, in order to effectively manage and control the invasion of tunicates, in-depth research is needed on the key environmental factors that affect their distribution and spread.

[0003] At present, through the study of the genetic adaptation mechanism of the mastoid tunicate to different marine environments, it has been confirmed that the maximum salinity of seawater is one of the key driving forces for the local adaptation of different groups. The local adaptation driven by environmental factors can promote the expansion of marine invasive organisms and invade different marine environments. However, the key processes and key environmental factors that affect the colonization of the invasive organisms mastoid tunicate and sasborne tunicate are still unclear, which leads to unclear priority targets for compensatory restoration, making it difficult to quantify the ecological damage and economic impact caused by invasive organisms to nearshore ecosystems.

[0004] In recent years, new technologies have provided new perspectives for the analysis of large biological data sets (Speed, 2011). The maximum information-based nonparametric exploration (MINE algorithm) provides a statistical method for discovering potential associations in large data sets. The MINE algorithm is based on the maximum information coefficient (MIC), giving the same score to different types of the same "noise" (Reshef et al., 2011), providing a data mining solution for exploring various types of biological relationships.

[0005] Currently, the MINE algorithm has been used to explore the relationship between plankton groups (Genitsaris et al., 2019), macrobenthic groups (Gu et al., 2021), and environmental parameters, but has not been applied to research related to invasive organisms. In addition, although the MINE algorithm can effectively identify the association between variables, it mainly relies on the statistical correlation between variables when constructing the association network between variables. It has weak explanatory power for causal relationships and complex nonlinear relationships between variables, and often faces dimensional disasters when processing high-dimensional data, resulting in information loss and reduced model generalization ability. Therefore, in practical applications, it is often necessary to combine other methods for further analysis to determine the key environmental factors that affect the colonization of the mastoidea and sasieri populations.

[0006] Therefore, an optimized data analysis method for ascidian surveys based on multidimensional data collection is desired. Summary of the invention

[0007] In order to solve the above-mentioned technical problems, the present application is proposed. The embodiment of the present application provides a system and method for analyzing sea squirt survey data based on multidimensional data collection, which uses the MINE algorithm to perform correlation analysis on various invasive biological data and various environmental factors, to construct the MIC adjacency matrix between invasive biological data and environmental factor data, and uses the data processing technology based on deep learning to perform feature extraction and enhancement processing on the adjacency matrix, to identify and emphasize the key connection between environmental factors and invasive organisms, so as to intelligently identify the key environmental factors in various environmental factors. In this way, the recognition accuracy of key participants in environmental factors can be effectively improved, thereby providing powerful decision-making support for the management of marine ecology.

[0008] Accordingly, according to one aspect of the present application, a method for analyzing sea squirt survey data based on multidimensional data collection is provided, comprising: Collect data on invasive organisms and environmental factors; Based on the MINE algorithm, the adjacency matrix of invasive organisms and environmental factors was constructed; Extracting the adjacency topological features of the invasive organism-environmental factor adjacency matrix to obtain an invasive organism-environmental factor adjacency topological feature graph; Performing feature enhancement on the invasive organism-environmental factor adjacency topological feature map to obtain an invasive organism-environmental factor adjacency enhanced topological feature map; Based on the invasive organism-environmental factor adjacency enhanced topological feature map, key environmental factors are identified.

[0009] In the above-mentioned tunicate survey data analysis method based on multidimensional data collection, the feature enhancement of the invasive organism-environmental factor adjacency topology feature map is performed, including: calculating the node semantic association of the invasive organism-environmental factor adjacency topology feature map to obtain the invasive organism-environmental factor adjacency topology node semantic association topology matrix; based on the node spatial position information of the invasive organism-environmental factor adjacency topology feature map, the invasive organism-environmental factor adjacency topology node semantic association topology matrix is ​​spatially attended attenuated modulated to obtain the invasive organism-environmental factor adjacency topology node association space soft constraint topology feature matrix; based on the invasive organism-environmental factor adjacency topology node association space soft constraint topology feature matrix, the invasive organism-environmental factor adjacency topology feature map is node granularity spatial enhancement processing is performed to obtain the invasive organism-environmental factor adjacency enhanced topology feature map.

[0010] In the above-mentioned ascidian survey data analysis method based on multidimensional data collection, an invasive organism-environmental factor adjacency matrix is ​​constructed based on the MINE algorithm, including: using the MINE algorithm to calculate the MIC value between each pair of the invasive organisms and the environmental factors to obtain the invasive organism-environmental factor adjacency matrix.

[0011] In the above-mentioned tunicate survey data analysis method based on multidimensional data collection, the adjacency topological features of the invasive organism-environmental factor adjacency matrix are extracted to obtain an invasive organism-environmental factor adjacency topological feature map, including: inputting the invasive organism-environmental factor adjacency matrix into an adjacency topological feature extractor based on a void convolutional layer to obtain the invasive organism-environmental factor adjacency topological feature map.

[0012] In the above-mentioned tunicate survey data analysis method based on multidimensional data collection, the node semantic associations of the invasive organism-environmental factor adjacency topology feature graph are calculated to obtain an invasive organism-environmental factor adjacency topology node semantic association topology matrix, including: feature dispersion of the invasive organism-environmental factor adjacency topology feature graph along the channel dimension to obtain a set of invasive organism-environmental factor adjacency topology node feature vectors; calculating the semantic association score between any two invasive organism-environmental factor adjacency topology node feature vectors in the set of invasive organism-environmental factor adjacency topology node feature vectors to obtain the invasive organism-environmental factor adjacency topology node semantic association topology matrix.

[0013] In the above-mentioned tunicate survey data analysis method based on multidimensional data collection, the semantic association score between any two invasive organism-environmental factor adjacent topology node feature vectors in the set of invasive organism-environmental factor adjacent topology node feature vectors is calculated to obtain the invasive organism-environmental factor adjacent topology node semantic association topology matrix, including: performing nonlinear mapping on each invasive organism-environmental factor adjacent topology node feature vector in the set of invasive organism-environmental factor adjacent topology node feature vectors to obtain a set of mapped invasive organism-environmental factor adjacent topology node feature vectors; calculating the Poincare distance between any two mapped invasive organism-environmental factor adjacent topology node feature vectors in the set of mapped invasive organism-environmental factor adjacent topology node feature vectors as the semantic association score to obtain the invasive organism-environmental factor adjacent topology node semantic association topology matrix.

[0014] In the above-mentioned tunicate survey data analysis method based on multidimensional data collection, based on the node spatial position information of the invasive organism-environmental factor adjacency topology feature graph, the invasive organism-environmental factor adjacency topology node semantic association topology matrix is ​​spatially attended attenuated modulated to obtain the invasive organism-environmental factor adjacency topology node association space soft constraint topology feature matrix, including: based on the spatial distance between any two invasive organism-environmental factor adjacency topology node feature vectors in the set of invasive organism-environmental factor adjacency topology node feature vectors, the invasive organism-environmental factor adjacency topology node semantic association topology matrix is ​​spatially attended attenuated modulated to obtain the invasive organism-environmental factor adjacency topology node association space soft constraint topology matrix; the invasive organism-environmental factor adjacency topology node association space soft constraint topology matrix is ​​void convolutionally encoded to obtain the invasive organism-environmental factor adjacency topology node association space soft constraint topology feature matrix.

[0015] In the above-mentioned tunicate survey data analysis method based on multidimensional data collection, the invasive organism-environmental factor adjacent topology node semantic association topology matrix is ​​spatially focused attenuated modulated based on the spatial distance between any two invasive organism-environmental factor adjacent topology node feature vectors in the set of invasive organism-environmental factor adjacent topology node feature vectors to obtain an invasive organism-environmental factor adjacent topology node associated spatial soft constraint topology matrix, including: using the natural exponential function value of the Euclidean distance between any two invasive organism-environmental factor adjacent topology node feature vectors in the set of invasive organism-environmental factor adjacent topology node feature vectors as the spatial distance factor to obtain the invasive organism-environmental factor adjacent topology node spatial distance topology matrix; using a preset attenuation modulation constant to multiply the invasive organism-environmental factor adjacent topology node spatial distance topology matrix to obtain a spatial attention attenuation modulation matrix, and calculating the location point division of the invasive organism-environmental factor adjacent topology node semantic association topology matrix and the spatial attention attenuation modulation matrix to obtain the invasive organism-environmental factor adjacent topology node associated spatial soft constraint topology matrix.

[0016] In the above-mentioned tunicate survey data analysis method based on multidimensional data collection, based on the soft constraint topological feature matrix of the invasive organism-environmental factor adjacency topological node association space, the invasive organism-environmental factor adjacency topological feature map is subjected to node granularity space enhancement processing to obtain the invasive organism-environmental factor adjacency enhanced topological feature map, including: inputting the set of invasive organism-environmental factor adjacency topological node feature vectors and the invasive organism-environmental factor adjacency topological node association space soft constraint topological feature matrix into a graph convolutional coding module to obtain a set of context-related enhanced invasive organism-environmental factor adjacency topological node feature vectors; and performing feature shape reshaping on the set of context-related enhanced invasive organism-environmental factor adjacency topological node feature vectors to obtain the invasive organism-environmental factor adjacency enhanced topological feature map.

[0017] In the above-mentioned tunicate survey data analysis method based on multi-dimensional data collection, key environmental factors are identified based on the invasive organism-environmental factor adjacency enhanced topological feature map, including: inputting the invasive organism-environmental factor adjacency enhanced topological feature map into a classifier-based key participant identifier to obtain an identification result, and the identification result is used to represent the label of the key environmental factor in the environmental factor.

[0018] According to another aspect of the present application, a sea squirt survey data analysis system based on multi-dimensional data collection is provided, which comprises: Multidimensional data collection module, used to collect data on invasive organisms and environmental factors; The adjacency matrix construction module is used to construct the adjacency matrix of invasive organisms and environmental factors based on the MINE algorithm; An adjacency topology feature extraction module is used to extract the adjacency topology features of the invasive organism-environmental factor adjacency matrix to obtain an invasive organism-environmental factor adjacency topology feature graph; A feature enhancement module, used for performing feature enhancement on the invasive organism-environmental factor adjacency topological feature map to obtain an invasive organism-environmental factor adjacency enhanced topological feature map; The key participant identification module is used to identify key environmental factors based on the invasive organism-environmental factor adjacency enhanced topological feature map.

[0019] Compared with the prior art, the sea squirt survey data analysis system and method based on multidimensional data collection provided by the present application uses the MINE algorithm to perform correlation analysis on various invasive biological data and various environmental factors to construct the MIC adjacency matrix between invasive biological data and environmental factor data, and uses data processing technology based on deep learning to perform feature extraction and enhancement processing on the adjacency matrix to identify and emphasize the key connection between environmental factors and invasive organisms, thereby intelligently identifying the key environmental factors in various environmental factors. In this way, the recognition accuracy of key participants in environmental factors can be effectively improved, thereby providing strong decision-making support for the management of marine ecology. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0021] Figure 1 It is a flow chart of a method for analyzing ascidian survey data based on multi-dimensional data collection according to an embodiment of the present application.

[0022] Figure 2 It is a schematic diagram of the architecture of the ascidian survey data analysis method based on multi-dimensional data collection according to an embodiment of the present application.

[0023] Figure 3 This is a flowchart of step S140 in the ascidian survey data analysis method based on multi-dimensional data collection according to an embodiment of the present application.

[0024] Figure 4 It is a block diagram of a tunicate survey data analysis system based on multi-dimensional data collection according to an embodiment of the present application. DETAILED DESCRIPTION

[0025] Below, the embodiments of the present application will be described in more detail in conjunction with the accompanying drawings, and the above and other purposes, features and advantages of the present application will become more apparent. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the example embodiments described herein.

[0026] As mentioned in the above background technology, although the MINE algorithm can effectively identify the correlation between variables, it mainly relies on the statistical correlation between variables when constructing the association network between variables, and has a weaker ability to explain the causal relationship and complex nonlinear relationship between variables. In practical applications, it is often necessary to combine other methods for further analysis to determine the key environmental factors that affect the colonization of the mastoid tunicate and sassi populations. In response to the above technical problems, the technical concept of the present application is to use the MINE algorithm to perform correlation analysis on various invasive biological data and various environmental factors, to construct the MIC adjacency matrix between invasive biological data and environmental factor data, and to use the data processing technology based on deep learning to perform feature extraction and enhancement processing on the adjacency matrix, to identify and emphasize the key connection between environmental factors and invasive organisms, so as to intelligently identify the key environmental factors in various environmental factors. In this way, the recognition accuracy of key participants in environmental factors can be effectively improved, thereby providing powerful decision-making support for the management of marine ecology.

[0027] Figure 1 It is a flow chart of a method for analyzing ascidian survey data based on multi-dimensional data collection according to an embodiment of the present application. Figure 2 FIG. 1 is a schematic diagram of the architecture of a method for analyzing sea squirt survey data based on multi-dimensional data collection according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to the embodiment of the present application, the method for analyzing tunicate survey data based on multidimensional data collection includes the following steps: S110, collecting data on invasive organisms and data on environmental factors; S120, constructing an invasive organism-environmental factor adjacency matrix based on the MINE algorithm; S130, extracting the adjacency topological features of the invasive organism-environmental factor adjacency matrix to obtain an invasive organism-environmental factor adjacency topological feature map; S140, performing feature enhancement on the invasive organism-environmental factor adjacency topological feature map to obtain an invasive organism-environmental factor adjacency enhanced topological feature map; S150, identifying key environmental factors based on the invasive organism-environmental factor adjacency enhanced topological feature map.

[0028] In the above-mentioned method for analyzing sea squirt survey data based on multidimensional data collection, the step S110 collects data on invasive organisms and data on environmental factors. In the technical solution of the present application, the data on invasive organisms are various observation and record information related to sea squirts, including but not limited to information such as the type, quantity, distribution, growth rate, and reproduction method of invasive organisms, and the data on environmental factors are important factors that affect the survival, reproduction, and spread of invasive organisms, including but not limited to environmental parameters such as temperature, salinity, pH value, nutrient concentration, and bottom type.

[0029] In the above-mentioned method for analyzing sea squirt survey data based on multidimensional data collection, the step S120, based on the MINE algorithm, constructs an invasive organism-environmental factor adjacency matrix. In a specific example of the present application, the step S120 includes: using the MINE algorithm to calculate the MIC value between each pair of the invasive organisms and the environmental factors to obtain the invasive organism-environmental factor adjacency matrix. It should be understood that, due to the complex mutual correlation between invasive organisms and environmental factors in the marine ecosystem. Therefore, in order to effectively establish a correlation network between various environmental factors and invasive organism data, the present application uses the MINE algorithm to calculate the MIC value between each pair of invasive organisms and the environmental factors to construct an invasive organism-environmental factor adjacency matrix. In the invasive organism-environmental factor adjacency matrix, each element represents the strength of association between the corresponding invasive organism and the environmental factor, thereby intuitively showing the potential association pattern between different environmental factors and invasive organisms. Specifically, the MINE algorithm is based on the maximum information coefficient (MIC), and by giving the same score to different types of the same "noise", it can reveal the correlation topology between invasive organisms and environmental factors. Unlike other regression models based on specific functions, the MINE algorithm can detect complex relationships driven by multiple factors without knowing the types of these associations in advance, thus providing an effective technical means for quantitative analysis of the interactions between invasive organisms and environmental factors.

[0030] In the above-mentioned method for analyzing the sea squirt survey data based on multidimensional data collection, the step S130 extracts the adjacency topology features of the invasive organism-environmental factor adjacency matrix to obtain an invasive organism-environmental factor adjacency topology feature map. In a specific example of the present application, the step S130 includes: inputting the invasive organism-environmental factor adjacency matrix into an adjacency topology feature extractor based on a hole convolution layer to obtain the invasive organism-environmental factor adjacency topology feature map. That is, since the MINE algorithm mainly analyzes the correlation between two independent variables, in order to further consider the global association action mode between the data of the invasive organism and the data of the environmental factor, the present application uses an adjacency topology feature extractor based on a hole convolution layer to extract features from the invasive organism-environmental factor adjacency matrix, so as to learn the associated topological structure between the environmental factor and the invasive organism, and generate an invasive organism-environmental factor adjacency topology feature map. Among them, the atrous convolution layer can effectively expand the receptive field of the convolution kernel without increasing the amount of computation by introducing an adjustable atrous rate, thereby capturing a wider range of spatial structural information and enhancing the ability to understand the complex correlation structure between invasive organisms and environmental factors.

[0031] In the above-mentioned method for analyzing sea squirt survey data based on multidimensional data collection, the step S140 performs feature enhancement on the invasive organism-environmental factor adjacency topology feature map to obtain an invasive organism-environmental factor adjacency enhanced topology feature map. That is, in order to further identify and enhance the important correlation features in the invasive organism-environmental factor adjacency topology feature map to emphasize the key interaction mode between invasive organisms and environmental factors, the present application further performs feature enhancement processing on the invasive organism-environmental factor adjacency topology feature map to utilize the global context information of the invasive organism-environmental factor adjacency topology feature map to improve the distinguishability of the features and strengthen the expressive power of the features.

[0032] Figure 3 FIG. 1 is a flow chart of step S140 in the method for analyzing sea squirt survey data based on multi-dimensional data collection according to an embodiment of the present application. Figure 3As shown, the step S140 includes: S141, calculating the node semantic association of the invasive organism-environmental factor adjacency topology feature graph to obtain the invasive organism-environmental factor adjacency topology node semantic association topology matrix; S142, based on the node spatial position information of the invasive organism-environmental factor adjacency topology feature graph, performing spatial attention attenuation modulation on the invasive organism-environmental factor adjacency topology node semantic association topology matrix to obtain the invasive organism-environmental factor adjacency topology node association space soft constraint topology feature matrix; S143, based on the invasive organism-environmental factor adjacency topology node association space soft constraint topology feature matrix, performing node granularity spatial enhancement processing on the invasive organism-environmental factor adjacency topology feature graph to obtain the invasive organism-environmental factor adjacency enhanced topology feature graph.

[0033] Specifically, the step S141 includes: performing feature dispersion on the invasive organism-environmental factor adjacency topology feature graph along the channel dimension to obtain a set of invasive organism-environmental factor adjacency topology node feature vectors; calculating the semantic association score between any two invasive organism-environmental factor adjacency topology node feature vectors in the set of invasive organism-environmental factor adjacency topology node feature vectors to obtain the invasive organism-environmental factor adjacency topology node semantic association topology matrix. More specifically, the semantic association score between any two invasive organisms-environmental factor adjacent topology node feature vectors in the set of invasive organisms-environmental factor adjacent topology node feature vectors is calculated, including: performing nonlinear mapping on each invasive organisms-environmental factor adjacent topology node feature vector in the set of invasive organisms-environmental factor adjacent topology node feature vectors to obtain a set of mapped invasive organisms-environmental factor adjacent topology node feature vectors; calculating the Poincare distance between any two mapped invasive organisms-environmental factor adjacent topology node feature vectors in the set of mapped invasive organisms-environmental factor adjacent topology node feature vectors as the semantic association score to obtain the invasive organisms-environmental factor adjacent topology node semantic association topology matrix.

[0034] Specifically, the step S142 includes: performing spatial attention attenuation modulation on the semantic association topology matrix of the invasive organism-environmental factor adjacent topology node based on the spatial distance between any two invasive organism-environmental factor adjacent topology node feature vectors in the set of invasive organism-environmental factor adjacent topology node feature vectors to obtain a soft constraint topology matrix of the association space of the invasive organism-environmental factor adjacent topology node; performing atrous convolution coding on the soft constraint topology matrix of the association space of the invasive organism-environmental factor adjacent topology node to obtain a soft constraint topology feature matrix of the association space of the invasive organism-environmental factor adjacent topology node. More specifically, the spatial attention attenuation modulation is performed on the semantic association topology matrix of the invasive organism-environmental factor adjacent topology nodes, including: using the natural exponential function value of the Euclidean distance between any two invasive organism-environmental factor adjacent topology node feature vectors in the set of invasive organism-environmental factor adjacent topology node feature vectors as the spatial distance factor to obtain the invasive organism-environmental factor adjacent topology node spatial distance topology matrix; using a preset attenuation modulation constant to multiply the invasive organism-environmental factor adjacent topology node spatial distance topology matrix to obtain the spatial attention attenuation modulation matrix, and calculating the position point-by-position division of the invasive organism-environmental factor adjacent topology node semantic association topology matrix and the spatial attention attenuation modulation matrix to obtain the invasive organism-environmental factor adjacent topology node association space soft constraint topology matrix.

[0035] Specifically, the step S143 includes: inputting the set of feature vectors of the invasive organism-environmental factor adjacent topological nodes and the soft constraint topological feature matrix of the associated space of the invasive organism-environmental factor adjacent topological nodes into a graph convolutional coding module to obtain a set of context-related enhanced invasive organism-environmental factor adjacent topological node feature vectors; and reshaping the set of context-related enhanced invasive organism-environmental factor adjacent topological node feature vectors to obtain the invasive organism-environmental factor adjacent enhanced topological feature graph.

[0036] Accordingly, the step S140 includes: processing the invasive organism-environmental factor adjacency topology feature map with the following feature enhancement formula to obtain the invasive organism-environmental factor adjacency enhanced topology feature map, wherein the feature enhancement formula is:

[0037]

[0038]

[0039]

[0040]

[0041]

[0042]

[0043] in, is the adjacency topological characteristic graph of the invasive organisms and environmental factors, , and They respectively represent the height, width and number of channels of the adjacent topological characteristic graph of the invasive organisms and environmental factors, represents the feature shape reshaping, represents the set of feature vectors of the adjacency topological nodes of invasive organisms and environmental factors, and Respectively represent the first and The eigenvectors of the topological nodes of the invasive organisms-environmental factors adjacency, The value of is the number of characteristic vectors of the adjacent topological nodes of the invasive organism-environmental factor, and , The value of is the length of the characteristic vector of the adjacent topological node of the invasive organism-environmental factor, is the weight transformation matrix, is the weight transformation vector, and Respectively represent the first and The feature vector of the adjacent topological nodes of invasive organisms and environmental factors after mapping, represents the inverse hyperbolic cosine function in hyperbolic space, represents the norm of the eigenvector, Indicates the The feature vector of the adjacent topological nodes of the invasive organism-environmental factor after mapping and the The Poincare distance between the feature vectors of the adjacent topological nodes of the invasive organism-environmental factor after mapping, In the semantic association topology matrix of the adjacent topological nodes of the invasive organisms and environmental factors, The semantic relevance score of the location, is the preset attenuation modulation constant, Indicates the The invasive organism-environmental factor adjacency topological node feature vector and the The Euclidean distance between the feature vectors of the adjacent topological nodes of the invasive organism-environmental factor, is the soft constraint topological matrix of the invasive organism-environmental factor adjacency topological node association space, The soft constraint topological matrix of the adjacent topological node association space of the invasive organisms and environmental factors is The eigenvalues ​​of the positions, represents the dilated convolutional coding, is the soft constraint topological feature matrix of the invasive organism-environmental factor adjacency topological node association space, represents graph convolution processing, Represents the set of feature vectors of the adjacent topological nodes of the context-reinforced invasive organism-environmental factor, Represents the invasive organism-environmental factor adjacency enhanced topological feature map.

[0044] That is, first, the invasive organism-environmental factor adjacency topology feature graph is feature scattered along the channel dimension to split it into a feature vector set of node granularity. Among them, each node feature corresponds to the association between a specific invasive organism data and an environmental factor. Then, by calculating the semantic association score between any two node features in the set, the invasive organism-environmental factor adjacency topology node semantic association topology matrix is ​​constructed to reveal the semantic association between different environmental factor data and invasive organism data. In particular, in this process, considering that compared with Euclidean space, hyperbolic space can more effectively capture the nonlinear relationship between data, therefore, the present application first maps each node feature from Euclidean space to Poincare space, so that it adapts to the geometric characteristics of hyperbolic space while maintaining the original semantic information, and calculates the Poincare distance between any two mapped node features as the semantic association score between the two, so as to utilize the geometric characteristics of Poincare space to enhance the nonlinear association between features, improve the model's ability to distinguish features of different categories, and more accurately simulate the inherent semantic association structure between environmental factor data and invasive organism data.

[0045] Next, considering that in the semantic feature space, similar node features often have stronger correlations than distant node features, the present application further performs spatial attention attenuation modulation on the above-mentioned invasive organism-environmental factor adjacency topology node semantic association topology matrix based on the spatial distance between any two node features to simulate the association attenuation characteristics between features, so that the model pays more attention to the feature associations in adjacent areas. Furthermore, the hole convolution technique is used to perform hole convolution encoding on the score matrix adjusted by spatial soft constraints to capture the semantic association topology features between each node feature, so as to obtain the invasive organism-environmental factor adjacency topology node association space soft constraint topology feature matrix. Secondly, the set of node features and the invasive organism-environmental factor adjacency topology node association space soft constraint topology feature matrix are processed by graph convolution coding technology, and the graph convolution characteristics are used to update the feature representation of each node based on the semantic association information between nodes, and the global semantic context association expression of node features is enhanced by learning the intrinsic semantic association structure between environmental factor data and invasive organism data. Finally, by reshaping the characteristic shapes of each node feature after contextual association enhancement and restoring its original characteristic structure, an invasive organism-environmental factor adjacency enhanced topological characteristic map is obtained, thereby realizing the recognition and refined enhancement of important association patterns between environmental factor data and invasive organism data.

[0046] In the above-mentioned sea squirt survey data analysis method based on multidimensional data collection, the step S150, based on the invasive organism-environmental factor adjacency enhanced topological feature map, identifies key environmental factors. In a specific example of the present application, the step S150 includes: the invasive organism-environmental factor adjacency enhanced topological feature map is input into a key participant identifier based on a classifier to obtain an identification result, and the identification result is used to represent the label of the key environmental factor in the environmental factor. That is, a classification algorithm is used to classify and judge the invasive organism-environmental factor adjacency enhanced topological feature map to determine which environmental factors have a decisive influence on the distribution and activity of invasive organisms. In the technical solution of the present application, the classifier model can effectively capture the important association patterns contained in the invasive organism-environmental factor adjacency enhanced topological feature map through training, identify the key participants in the environmental factor, thereby accurately mapping the invasive organism-environmental factor adjacency enhanced topological feature map to the corresponding environmental factor label to determine the key environmental factors affecting the distribution of invasive organisms.

[0047] In the technical solution of the present application, the characteristic values ​​of each position in the invasive organism-environmental factor adjacency topological feature map are used to represent the MIC value between each pair of the invasive organism and the environmental factor. After the invasive organism-environmental factor adjacency matrix is ​​convolutionally encoded and feature enhanced based on spatial soft constraints, the obtained invasive organism-environmental factor adjacency enhanced topological feature map will also have feature distribution diversity due to the difference in attention weights of spatial soft constraints, affecting the consistency of classification mapping iteration and reducing the accuracy of classification results.

[0048] Preferably, inputting the invasive organism-environmental factor adjacency enhanced topological feature graph into a classifier-based key participant identifier to obtain an identification result comprises: Expanding the invasive organism-environmental factor adjacency enhanced topological feature graph into an invasive organism-environmental factor adjacency enhanced topological feature vector; Determine the characteristic mean and characteristic variance of the invasive organism-environmental factor adjacency enhancement topology characteristic vector, and divide the characteristic mean by the characteristic variance to obtain the invasive organism-environmental factor adjacency enhancement topology probability statistics value:

[0049] in, and They respectively represent the mean and variance of the feature set composed of the eigenvalues ​​of the adjacent enhanced topological feature vector of the invasive organism-environmental factor, It represents the probability statistics of the topological enhancement of the adjacency between invasive organisms and environmental factors; The invasive organism-environmental factor adjacency enhanced topology feature vector is point-multiplied with the inverse of the maximum eigenvalue in the invasive organism-environmental factor adjacency enhanced topology feature vector to obtain the invasive organism-environmental factor adjacency enhanced topology probability constraint vector:

[0050] in, represents the inverse of the maximum eigenvalue of the adjacency enhanced topological eigenvector of the invasive organism-environmental factor, represents the invasive organism-environmental factor adjacency enhanced topological feature vector, It means point multiplication by position. represents the probability constraint vector of the adjacent enhanced topology of invasive organisms and environmental factors; After performing point addition of the invasive organism-environmental factor adjacency enhanced topology probability constraint vector and the invasive organism-environmental factor adjacency enhanced topology probability statistic value, the logarithmic value with base 2 is calculated to obtain the invasive organism-environmental factor adjacency enhanced topology information interaction vector:

[0051] in, represents the invasive organism-environmental factor adjacency enhanced topological probability constraint vector, represents the probability statistics of the topological enhancement of the adjacency between invasive organisms and environmental factors. Indicates adding by position point. represents the interaction vector of the adjacency-enhanced topological information between invasive organisms and environmental factors; After performing point subtraction on the invasive organism-environmental factor adjacency enhanced topology probability constraint vector using the invasive organism-environmental factor adjacency enhanced topology probability statistics, the inverse of each eigenvalue is calculated to obtain the invasive organism-environmental factor adjacency enhanced topology sequence constraint vector:

[0052] in, represents the invasive organism-environmental factor adjacency enhanced topological probability constraint vector, represents the probability statistics of the topological enhancement of the adjacency between invasive organisms and environmental factors. It means to subtract by position point. represents the invasive organism-environmental factor adjacency enhanced topological sequence constraint vector; The invasive organism-environmental factor adjacency enhanced topology information interaction vector is interpolated with the invasive organism-environmental factor adjacency enhanced topology sequence constraint vector to obtain an optimized invasive organism-environmental factor adjacency enhanced topology feature vector; The optimized invasive organism-environmental factor adjacency enhanced topological feature vector is passed through a classifier-based key participant identifier to obtain an identification result.

[0053] In this case, the probability statistical characteristics of the invasive organism-environmental factor adjacency enhanced topological feature vector obtained by expanding the invasive organism-environmental factor adjacency enhanced topological feature graph are used to simulate the mesoscale interaction structure under probability constraints between the eigenvalue scale and the feature graph scale of the invasive organism-environmental factor adjacency enhanced topological feature vector, so as to construct a bidirectional latent variable motif based on the mesoscale short sequence relative to the probability statistical value to perform mesoscale bidirectional migration, and to perform posterior recovery based on the sequence constraints on the interactive information, thereby improving the convergence effect in the probability density domain and improving the accuracy of the recognition result obtained by the classifier-based key participant identifier of the invasive organism-environmental factor adjacency enhanced topological feature graph.

[0054] In summary, the method for analyzing sea squirt survey data based on multidimensional data collection according to the embodiment of the present application is explained, which uses the MINE algorithm to perform correlation analysis on various invasive biological data and various environmental factors to construct the MIC adjacency matrix between invasive biological data and environmental factor data, and uses deep learning-based data processing technology to perform feature extraction and enhancement processing on the adjacency matrix to identify and emphasize the key connection between environmental factors and invasive organisms, thereby intelligently identifying the key environmental factors in various environmental factors. In this way, the accuracy of identifying key participants in environmental factors can be effectively improved, thereby providing strong decision-making support for the management of marine ecology.

[0055] Figure 4 FIG. 1 is a block diagram of a sea squirt survey data analysis system based on multi-dimensional data collection according to an embodiment of the present application. Figure 4 As shown, according to the embodiment of the present application, the ascidian survey data analysis system 100 based on multidimensional data collection includes: a multidimensional data collection module 110, which is used to collect data on invasive organisms and environmental factors; an adjacency matrix construction module 120, which is used to construct an invasive organism-environmental factor adjacency matrix based on the MINE algorithm; an adjacency topology feature extraction module 130, which is used to extract the adjacency topology features of the invasive organism-environmental factor adjacency matrix to obtain an invasive organism-environmental factor adjacency topology feature map; a feature enhancement module 140, which is used to perform feature enhancement on the invasive organism-environmental factor adjacency topology feature map to obtain an invasive organism-environmental factor adjacency enhanced topology feature map; a key participant identification module 150, which is used to identify key environmental factors based on the invasive organism-environmental factor adjacency enhanced topology feature map.

[0056] Here, those skilled in the art can understand that the specific operations of each module in the above-mentioned sea squirt survey data analysis system based on multi-dimensional data collection have been referred to above. Figures 1 to 3 The description of the ascidian survey data analysis method based on multidimensional data acquisition has been introduced in detail, and therefore, its repeated description will be omitted.

[0057] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details of the above embodiments are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

[0058] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed system and method can be implemented in other ways. For example, the system embodiment described above is only schematic. For example, the module division is only a logical function division, and there may be other division methods in actual implementation. The modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0059] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference to a figure in a claim should not be considered as limiting the claim to which it relates.

[0060] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. The multiple units stated in the system claims can also be implemented by one unit through software or hardware.

[0061] Finally, it should be noted that the above description has been given for the purpose of illustration and description. In addition, the above embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for analyzing sea squirt survey data based on multidimensional data collection, characterized in that: include: Collect data on invasive organisms and environmental factors; Based on the MINE algorithm, the adjacency matrix of invasive organisms and environmental factors was constructed; Extracting the adjacency topological features of the invasive organism-environmental factor adjacency matrix to obtain an invasive organism-environmental factor adjacency topological feature graph; Performing feature enhancement on the invasive organism-environmental factor adjacency topological feature map to obtain an invasive organism-environmental factor adjacency enhanced topological feature map; Based on the invasive organism-environmental factor adjacency enhanced topological feature map, identifying key environmental factors; Among them, the feature enhancement of the invasive organism-environmental factor adjacency topology feature map includes: calculating the node semantic association of the invasive organism-environmental factor adjacency topology feature map to obtain the invasive organism-environmental factor adjacency topology node semantic association topology matrix; based on the node spatial position information of the invasive organism-environmental factor adjacency topology feature map, performing spatial attention attenuation modulation on the invasive organism-environmental factor adjacency topology node semantic association topology matrix to obtain the invasive organism-environmental factor adjacency topology node association space soft constraint topology feature matrix; based on the invasive organism-environmental factor adjacency topology node association space soft constraint topology feature matrix, performing node granularity spatial enhancement processing on the invasive organism-environmental factor adjacency topology feature map to obtain the invasive organism-environmental factor adjacency enhanced topology feature map.

2. The method for analyzing sea squirt survey data based on multidimensional data collection according to claim 1, characterized in that: Based on the MINE algorithm, the adjacency matrix of invasive organisms and environmental factors is constructed, including: The MINE algorithm was used to calculate the MIC value between each pair of the invasive organisms and the environmental factors to obtain the invasive organism-environmental factor adjacency matrix.

3. The method for analyzing sea squirt survey data based on multidimensional data collection according to claim 2, characterized in that: Extracting the adjacency topological features of the invasive organism-environmental factor adjacency matrix to obtain an invasive organism-environmental factor adjacency topological feature graph includes: The invasive organism-environmental factor adjacency matrix is ​​input into an adjacency topology feature extractor based on a dilated convolutional layer to obtain the invasive organism-environmental factor adjacency topology feature graph.

4. The method for analyzing sea squirt survey data based on multidimensional data collection according to claim 3, characterized in that: Calculating the node semantic association of the invasive organism-environmental factor adjacency topology feature graph to obtain an invasive organism-environmental factor adjacency topology node semantic association topology matrix includes: The invasive organism-environmental factor adjacency topology feature graph is feature scattered along the channel dimension to obtain a set of invasive organism-environmental factor adjacency topology node feature vectors; The semantic association score between any two invasive organisms-environmental factor adjacent topology node feature vectors in the set of the invasive organisms-environmental factor adjacent topology node feature vectors is calculated to obtain the invasive organisms-environmental factor adjacent topology node semantic association topology matrix.

5. The method for analyzing sea squirt survey data based on multidimensional data collection according to claim 4, characterized in that: Calculating the semantic association score between any two invasive organisms-environmental factor adjacent topology node feature vectors in the set of invasive organisms-environmental factor adjacent topology node feature vectors to obtain the invasive organisms-environmental factor adjacent topology node semantic association topology matrix, including: Nonlinearly mapping each invasive organism-environmental factor adjacent topology node feature vector in the set of invasive organism-environmental factor adjacent topology node feature vectors to obtain a set of mapped invasive organism-environmental factor adjacent topology node feature vectors; The Poincare distance between any two mapped invasive organism-environmental factor adjacent topology node feature vectors in the set of mapped invasive organism-environmental factor adjacent topology node feature vectors is calculated as the semantic association score to obtain the invasive organism-environmental factor adjacent topology node semantic association topology matrix.

6. The method for analyzing sea squirt survey data based on multidimensional data collection according to claim 5, characterized in that: Based on the node spatial position information of the invasive organism-environmental factor adjacency topology feature graph, the invasive organism-environmental factor adjacency topology node semantic association topology matrix is ​​subjected to spatial attention attenuation modulation to obtain the invasive organism-environmental factor adjacency topology node association space soft constraint topology feature matrix, including: Based on the spatial distance between any two invasive organism-environmental factor adjacent topological node feature vectors in the set of invasive organism-environmental factor adjacent topological node feature vectors, the invasive organism-environmental factor adjacent topological node semantic association topological matrix is ​​subjected to spatial attention attenuation modulation to obtain an invasive organism-environmental factor adjacent topological node association space soft constraint topological matrix; The soft constraint topological matrix of the adjacent topological node association space of the invasive organisms and environmental factors is subjected to dilated convolution coding to obtain the soft constraint topological feature matrix of the adjacent topological node association space of the invasive organisms and environmental factors.

7. The method for analyzing sea squirt survey data based on multidimensional data collection according to claim 6, characterized in that: Based on the spatial distance between any two invasive organism-environmental factor adjacent topological node feature vectors in the set of invasive organism-environmental factor adjacent topological node feature vectors, the invasive organism-environmental factor adjacent topological node semantic association topological matrix is ​​subjected to spatial attention attenuation modulation to obtain an invasive organism-environmental factor adjacent topological node association space soft constraint topological matrix, including: Using the natural exponential function value of the Euclidean distance between any two invasive organisms-environmental factor adjacent topological node feature vectors in the set of invasive organisms-environmental factor adjacent topological node feature vectors as a spatial distance factor to obtain an invasive organisms-environmental factor adjacent topological node spatial distance topological matrix; The preset attenuation modulation constant is used to multiply the spatial distance topological matrix of the invasive organism-environmental factor adjacent topological nodes to obtain the spatial attention attenuation modulation matrix, and the semantic association topological matrix of the invasive organism-environmental factor adjacent topological nodes and the spatial attention attenuation modulation matrix are calculated and divided by the position points to obtain the invasive organism-environmental factor adjacent topological node association space soft constraint topological matrix.

8. The method for analyzing sea squirt survey data based on multidimensional data collection according to claim 7, characterized in that: Based on the invasive organism-environmental factor adjacency topology node association space soft constraint topology feature matrix, the invasive organism-environmental factor adjacency topology feature graph is subjected to node granularity space enhancement processing to obtain the invasive organism-environmental factor adjacency enhanced topology feature graph, including: Inputting the set of invasive organism-environmental factor neighbor topology node feature vectors and the invasive organism-environmental factor neighbor topology node association space soft constraint topology feature matrix into a graph convolutional coding module to obtain a set of context-related enhanced invasive organism-environmental factor neighbor topology node feature vectors; The set of the context-related enhanced invasive organism-environmental factor adjacency topological node feature vectors is reshaped to obtain the invasive organism-environmental factor adjacency enhanced topological feature graph.

9. The method for analyzing sea squirt survey data based on multidimensional data collection according to claim 8, characterized in that: Based on the invasive organism-environmental factor adjacency enhanced topological feature map, key environmental factors are identified, including: The invasive organism-environmental factor adjacency enhanced topological feature graph is input into a classifier-based key participant identifier to obtain an identification result, and the identification result is used to represent the label of the key environmental factor in the environmental factor.

10. A sea squirt survey data analysis system based on multi-dimensional data collection, characterized in that: include: Multidimensional data collection module, used to collect data on invasive organisms and environmental factors; The adjacency matrix construction module is used to construct the adjacency matrix of invasive organisms and environmental factors based on the MINE algorithm; An adjacency topology feature extraction module is used to extract the adjacency topology features of the invasive organism-environmental factor adjacency matrix to obtain an invasive organism-environmental factor adjacency topology feature graph; A feature enhancement module, used for performing feature enhancement on the invasive organism-environmental factor adjacency topological feature map to obtain an invasive organism-environmental factor adjacency enhanced topological feature map; The key participant identification module is used to identify key environmental factors based on the invasive organism-environmental factor adjacency enhanced topological feature map.

Images