Method and system for constructing climatic risk comprehensive evaluation index system with multi-level structure

Abstract

The invention provides a method and a system for constructing a climatic risk comprehensive assessment index system with a multi-level structure, and the method comprises the steps: obtaining multi-source data related to climatic risks, and carrying out the preprocessing of the data, and obtaining a climatic risk data set; constructing a climate risk secondary index according to the climate risk data set, and quantifying the climate risk secondary index to obtain a climate risk secondary index value; and summarizing the climate risk secondary indexes to corresponding macroscopic dimensions based on the climate risk secondary index values, and constructing a climate risk comprehensive assessment index system through the macroscopic dimensions and the climate risk secondary indexes. According to the method, a multi-source data fusion and automatic topic modeling method is used, climate change related text information is collected from multiple sources, category labels or artificial experience do not need to be predefined, a comprehensive, accurate and hierarchical climate risk comprehensive assessment index system is constructed, the assessment efficiency and accuracy are improved, and the assessment efficiency is improved. And the interpretability and practicability of the result are enhanced.

Description

Technical Field

[0001] The present invention relates to the field of climate assessment technology, and in particular to a method and system for constructing a multi-level climate risk comprehensive assessment indicator system. Background Art

[0002] Climate risk assessment is a complex issue covering multiple levels, including environment, policy, enterprise and society. Each level reflects the evolution of a specific type of climate risk. Identifying these changing trends is crucial for assessing the impact of climate risks, understanding their inherent characteristics and predicting future risks. However, traditional assessment methods rely on expert experience and a single data source, and usually require manual setting of risk parameters or follow fixed assessment criteria. They lack the flexibility and accuracy to handle large-scale multidimensional data. As climate change issues become increasingly complex, traditional assessment methods have become difficult to fully capture the complexity of climate risks, and it is also difficult to make accurate assessments of future climate risks. Summary of the invention

[0003] The purpose of the present invention is to provide a method and system for constructing a multi-level climate risk comprehensive assessment indicator system, aiming to solve the above-mentioned problems in the prior art.

[0004] The embodiment of the present invention provides a method for constructing a multi-level climate risk comprehensive assessment index system, including:

[0005] Acquire multi-source data related to climate risk, and pre-process the multi-source data to obtain a climate risk data set;

[0006] constructing a secondary climate risk indicator according to the climate risk data set, and quantifying the secondary climate risk indicator to obtain a secondary climate risk indicator value; and

[0007] Based on the values ​​of the climate risk secondary indicators, the climate risk secondary indicators are summarized into corresponding macro dimensions, and a climate risk comprehensive assessment indicator system is constructed through the macro dimensions and the climate risk secondary indicators.

[0008] The embodiment of the present invention provides a system for constructing a climate risk comprehensive assessment index system with a multi-level structure, including:

[0009] A data module, used to obtain multi-source data related to climate risks, and pre-process the multi-source data to obtain a climate risk data set;

[0010] a quantification module, configured to construct a secondary climate risk indicator according to the climate risk data set, and quantify the secondary climate risk indicator to obtain a secondary climate risk indicator value; and

[0011] A construction module is used to summarize the climate risk secondary indicators into corresponding macro dimensions based on the values ​​of the climate risk secondary indicators, and to construct a climate risk comprehensive assessment indicator system through the macro dimensions and the climate risk secondary indicators.

[0012] An embodiment of the present invention also provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the method for constructing the multi-level structured comprehensive climate risk assessment indicator system.

[0013] An embodiment of the present invention also provides a computer-readable storage medium, on which a program for implementing information transmission is stored. When the program is executed by a processor, the steps of the method for constructing the multi-level climate risk comprehensive assessment indicator system are implemented.

[0014] The use of the embodiments of the present invention can include the following beneficial effects: the embodiments of the present invention propose a method for constructing a multi-level structured comprehensive climate risk assessment index system, using Top2Vec to automatically identify topics in documents without manually specifying the number of topics, and generating high-quality and highly interpretable topic representations by combining word embedding and document embedding, so that non-professionals can easily understand the climate risk index system; at the same time, the HDBSCAN algorithm is used to identify clusters by constructing a hierarchical clustering tree and selecting the optimal segmentation point. There is no need to pre-specify the number of clusters, and clusters of any shape can be discovered, thereby revealing the complex structure of different levels of data, which is very suitable for identifying macro dimensions in climate risk data and helps to more deeply understand the multi-level structure of climate risk data. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate one or more embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0016] Figure 1 is a flow chart of a method for constructing a multi-level climate risk comprehensive assessment indicator system according to an embodiment of the present invention;

[0017] Figure 2 is a method flow chart of an embodiment of the present invention;

[0018] Figure 3 It is a flowchart of a specific implementation process of an embodiment of the present invention;

[0019] Figure 4is a flowchart of Top2vec topic modeling in an embodiment of the present invention;

[0020] Figure 5 is a flow chart of HDBSCAN hierarchical clustering according to an embodiment of the present invention;

[0021] Figure 6 It is a schematic diagram of a system for constructing a multi-level climate risk comprehensive assessment indicator system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the following will be combined with the drawings in one or more embodiments of this specification to clearly and completely describe the technical solutions in one or more embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this document.

[0023] Method Embodiment

[0024] According to an embodiment of the present invention, a method for constructing a climate risk comprehensive assessment indicator system with a multi-level structure is provided. Figure 1 is a flow chart of a method for constructing a multi-level climate risk comprehensive assessment indicator system according to an embodiment of the present invention. Figure 1 As shown, the method for constructing a multi-level climate risk comprehensive assessment indicator system according to an embodiment of the present invention specifically includes:

[0025] Step S101, obtaining multi-source data related to climate risk, preprocessing the multi-source data to obtain a climate risk data set, specifically including:

[0026] Preliminarily processing the multi-source data, and using natural language processing technology to perform word segmentation on the preliminarily processed multi-source data, and constructing a climate risk data set based on the word items after the word segmentation processing;

[0027] Wherein, the multi-source data is text data;

[0028] The preliminary processing includes removing stop words, punctuation marks and numbers, and standardization;

[0029] Step S102, constructing a secondary climate risk indicator according to the climate risk data set, and quantifying the secondary climate risk indicator to obtain a secondary climate risk indicator value, specifically including:

[0030] The climate risk dataset is trained using Doc2Vec word embedding technology to generate word vectors, which are mapped to vector space to obtain high-dimensional document vectors. The high-dimensional document vectors are reduced in dimension using the UMAP algorithm, and the reduced-dimensional document vectors are clustered using the Mean Shift algorithm. Keywords are extracted using the TF-IDF method based on the clustering results to obtain the theme of each cluster and its corresponding keywords, and the theme is used as a secondary indicator of climate risk.

[0031] Among them, the secondary climate risk indicators include climate risk indicators, climate physical risk indicators, climate policy uncertainty indicators, management climate concern indicators, climate risk impact indicators, climate change news indicators, climate risk concern indicators, climate risk awareness indicators and corporate biodiversity indicators;

[0032] The weight of each keyword is obtained using Formula 1 according to the Top2vec probability distribution and TF-IDF value of each keyword. The frequency of each keyword at each time point under each secondary indicator is weighted and summed according to the weight using Formula 2 to obtain the specific value of each secondary indicator at each time point. The specific value of each secondary indicator at each time point is subjected to Z-score normalization to obtain the final climate risk secondary indicator value;

[0033] Weight(w,M j )=α·p(w|M j )+(1-α)·TF-IDF(w,M j ) Formula 1;

[0034]

[0035] Among them, p(w|M j ) represents each keyword w∈K j Top2Vec probability distribution, TF-IDF(w,M j ) represents the TF-IDF value of each keyword w, M j is the jth secondary index, M j ={K j}, α is the adjustment factor, for each time point t∈T, T represents the set of time points, V(M j ,t) represents the jth secondary index M j The specific value at time point t, f(w,t,M j ) indicates the statistics of each secondary index M j Frequency of keywords;

[0036] Step S103, based on the values ​​of the climate risk secondary indicators, the climate risk secondary indicators are classified into corresponding macro dimensions, and a climate risk comprehensive assessment indicator system is constructed through the macro dimensions and the climate risk secondary indicators, specifically including:

[0037] The characteristic vector of each secondary indicator is obtained by keyword weight, and the HDBSCAN algorithm is used to cluster the characteristic vectors of each secondary indicator. Different macro dimensions are defined according to the optimal clustering results, and the climate risk secondary indicators are classified into corresponding macro dimensions;

[0038] The method further comprises:

[0039] Interactive visualization tools are used to display and analyze the climate risk comprehensive assessment indicator system to obtain the climate risk comprehensive assessment results.

[0040] The following is a detailed description of the method for constructing a multi-level climate risk comprehensive assessment index system according to an embodiment of the present invention. Figure 2 As shown, the above technical solution of the embodiment of the present invention is described in detail.

[0041] The embodiment of the present invention proposes a method for constructing a multi-level structured comprehensive assessment index system for climate risks, aiming to break through the limitations of traditional methods. The method uses Top2Vec automatic topic modeling technology to integrate multi-source data such as research reports, policy documents, news reports, social media information, and corporate reports related to climate risks, and automatically extracts valuable key information from massive texts, greatly expanding the coverage of the data, while discovering more potential risk factors through the correlation between data. Top2Vec technology can automatically identify topics in documents without presetting the number of topics, reducing human intervention, and improving the efficiency and accuracy of climate risk assessment; at the same time, the embodiment of the present invention uses the HDBSCAN algorithm to automatically cluster secondary indicators, revealing the hierarchical structure of climate risks, thereby gaining an in-depth understanding of the impact of climate risks from a macro perspective.

[0042] By integrating technologies such as natural language processing, machine learning, data mining, text quantification and risk assessment, the embodiment of the present invention can automatically extract and construct a comprehensive climate risk indicator system from multi-source data. The method supports the automated adjustment of the assessment system and has a real-time monitoring and early warning mechanism to ensure the timeliness of climate risk assessment; by scientifically setting climate risk scoring standards and calculating the comprehensive scores of various risks, the accuracy of the assessment is further guaranteed; at the same time, with the help of interactive visualization tools, users can easily browse and analyze climate risk situations, quickly obtain key information, improve assessment efficiency, and enhance the interpretability and practicality of the results. This data-driven decision-making climate risk assessment method, through real-time automated adjustment of the assessment system, brings a more scientific and comprehensive perspective to climate risk management, effectively fills the gap in existing technologies in the field of complex climate risk assessment, intuitively displays and analyzes climate risks while providing comprehensive risk assessments, and provides strong theoretical support for decision makers.

[0043] Specifically, the method for constructing a multi-level climate risk comprehensive assessment index system proposed in the embodiment of the present invention includes multi-source data integration, secondary index setting, climate risk quantification, macro-dimensional induction and climate risk assessment, such as Figure 3 As shown, the specific implementation steps are as follows:

[0044] S1. Multi-source data integration: collect text information related to climate change from different sources, pre-process and segment the text, and construct a climate risk dataset, denoted as D;

[0045] S2. Secondary indicator setting: Top2vec automatic topic modeling method is used to perform topic modeling on dataset D to identify topics closely related to climate risk. Based on the identified topics, a series of secondary climate risk indicators are established;

[0046] S3. Climate risk quantification: Extract representative keywords for each secondary indicator, and determine the weight of the keyword by combining the probability distribution of Top2vec with the TF-IDF value. Then, perform weighted summation of the keywords under each secondary indicator to obtain the specific value of each secondary indicator.

[0047] S4, macro-dimension summary: HDBSCAN algorithm is applied to cluster the secondary indicators into different macro-dimensions;

[0048] S5. Climate risk assessment: Utilize the constructed climate risk comprehensive indicator system in combination with interactive visualization tools to intuitively browse and analyze climate risks, conduct a comprehensive assessment of climate risks, and obtain climate risk assessment results.

[0049] The core steps of multi-source data integration described in step S1 include:

[0050] S11. Collect text information related to climate change from multiple data sources, including but not limited to research reports, policy documents, news reports, social media content, and corporate sustainability reports. The text collections of all data sources are merged into a total text collection A = {A1, A2, …, A3}, where A i Represents the text collection provided by the i-th data source;

[0051] S12. Preprocess the collected text data, including removing stop words, punctuation marks, numbers, etc., and converting all texts to lowercase to standardize the data format. Preprocess each text a∈A, and the preprocessed text set is recorded as A. ′ ={P(a)|a∈A}, where P is the preprocessing function.

[0052] S13, use natural language processing technology to segment the preprocessed text, split the text into independent terms, and the set of terms after segmentation is recorded as W(A ′ )={W(a ′ )|a ′ ∈A ′}, where W is the word segmentation function.

[0053] S14. Construct the segmented terms into a climate risk dataset D = {(d1,w1),(d2,w2),…,(d m ,w m )}, where d i represents the i-th document, w i Represents the word segmentation result of the document. The dataset contains all processed text data and provides basic data for subsequent topic modeling and indicator quantification.

[0054] Step S2 of the embodiment of the present invention uses Top2Vec automatic topic modeling technology to automatically construct a secondary indicator system for climate risk. The automatic topic modeling technology identifies topics closely related to climate risk, and constructs a series of detailed secondary indicators based on these topics, such as climate physical risk indicators, climate policy uncertainty indicators, etc. These indicators provide specific quantitative standards for climate risk assessment, making the assessment more detailed and comprehensive, helping non-professionals to easily understand complex climate risk patterns, and improving the accuracy and interpretability of the assessment. The specific steps of Top2vec topic modeling technology are as follows: Figure 4 As shown:

[0055] (1) Use Doc2Vec word embedding technology to train text data, generate word vectors, and capture the semantic relationship between words.

[0056] (2) The UMAP algorithm is used to reduce the dimensionality of high-dimensional document vectors to improve the effectiveness and efficiency of subsequent clustering algorithms.

[0057] (3) Use the Mean Shift algorithm to cluster the document vectors after dimensionality reduction and identify different topics related to climate risks.

[0058] (4) Based on the clustering results, the TF-IDF method is used to extract keywords, determine the theme of each cluster, and output the theme of each cluster and its corresponding keywords.

[0059] (5) Each theme corresponds to a secondary indicator, and a series of secondary indicators are constructed to provide a quantitative indicator system for climate risk assessment. The specific implementation process is as follows:

[0060] S21. Use Doc2Vec word embedding technology to train text data, generate word vectors, and map each word into a vector space of fixed length to capture the semantic relationship between words. The specific formula is as follows:

[0061]

[0062] Among them, v w and v d are vectors of fixed length, representing the representation of vocabulary w and document d in the vector space, respectively. That is the vector set of the document.

[0063] S22. Use the dimensionality reduction algorithm UMAP (Uniform Manifold Approximation and Projection) to reduce the dimensionality of the high-dimensional document vector generated by Doc2Vec to solve the dimensionality disaster problem and improve the effect and efficiency of the subsequent clustering algorithm. The specific formula is as follows:

[0064] V′ D =UMAP(V D );

[0065] in, and is the document vector after dimensionality reduction.

[0066] S23. Use the Mean Shift algorithm to cluster the document vectors after dimensionality reduction to identify different topics related to climate risks. The specific formula is as follows:

[0067]

[0068] {C1,C2,…,C K = MeanShift(V′ D );

[0069] Among them, K is the kernel function, h is the bandwidth parameter, and each point x is initialized i Center of i =x i , x i represents the i-th document vector in the document embedding space, x j represents the jth document vector in the document embedding space, when all centers c i When convergence, stop iteration and get clusters C1, C2, ..., C K .

[0070] S24. Based on the clustering results of Mean Shift, the TF-IDF method is applied to extract keywords from the documents of each cluster, so as to determine the theme of each cluster, and output the theme of each cluster and its corresponding keywords. The specific formula is as follows:

[0071]

[0072]

[0073] Among them, C j is the document set of the jth cluster, TF(w,d) is the frequency of word w in document d, DF(w) is the number of documents containing word w, m is the total number of documents, and K j It is cluster C j A collection of keywords.

[0074] S25. Each theme corresponds to a secondary indicator. Based on the identified themes, a series of secondary indicators are constructed, such as climate risk indicators, climate physical risk indicators, climate policy uncertainty indicators, management climate concern indicators, climate risk impact indicators, climate change news indicators, climate risk concern indicators, climate risk awareness indicators and corporate biodiversity indicators, etc., to provide a quantitative indicator system for climate risk assessment. The specific formula is as follows:

[0075] M j ={K j};

[0076] M={M1,M2,…,M k};

[0077] Among them, M j is the jth secondary indicator, M is the set of secondary indicators, and k represents the total number of indicators.

[0078] The climate risk quantification described in step S3 combines the probability distribution of Top2Vec and the TF-IDF value to determine the weight of each keyword. The frequency of each keyword at each time point under each secondary indicator is weighted and summed to obtain the specific value of each secondary indicator at each time point. The specific value of each secondary indicator calculated is dimensionless, i.e., Z-score normalization, to ensure comparability between different indicators.

[0079] The main steps include:

[0080] S31. Determine the weight of each keyword by combining the probability distribution of Top2vec and the TF-IDF value. j ), the specific formula is as follows:

[0081] Weight(w,M j )=α·p(w|M j )+(1-α)·TF-IDF(w,M j );

[0082] Among them, each keyword w∈K j The Top2Vec probability distribution is p(w|M j ), the TF-IDF value is TF-IDF(w,M j ), α is a regulation factor used to balance the importance of Top2Vec probability distribution and TF-IDF value (usually 0≤α≤1).

[0083] S32, weighted sum of the frequencies of each keyword at each time point under each secondary indicator, to obtain the specific value V (M) of each secondary indicator at each time point j ,t), the specific calculation formula is as follows:

[0084]

[0085] Among them, for each time point t∈T, T represents the set of time points, and each secondary index M is counted j The frequency of the keyword f(w,t,M j ).

[0086] S33. The specific values ​​of each secondary indicator calculated are dimensionless, i.e., Z-score normalization, to ensure the comparability between different indicators. The specific formula is as follows:

[0087]

[0088] Among them, V9M j ,t) represents the jth secondary index M j Specific value at time point t, μVj It is the secondary indicator M j The mean value at all time points, σV j It is the secondary indicator M j The standard deviation across all time points.

[0089] Step S4 of the embodiment of the present invention uses the HDBSCAN algorithm to automatically summarize the macro dimensions to which the secondary climate risk indicators belong, and classifies the secondary indicators into different macro dimensions, such as the natural dimension, policy dimension, company dimension, and social dimension, through advanced clustering algorithms, revealing the multi-level structure of climate risk data, making the assessment results more systematic and structured, helping decision makers understand the complexity of climate risks from a global perspective, providing more comprehensive and in-depth risk insights, and supporting more effective risk management, standard setting, and corporate strategic planning. The specific steps of the HDBSCAN algorithm are as follows: Figure 5 As shown:

[0090] (1) Each secondary indicator is represented as a high-dimensional feature vector, which consists of the weights of the keywords.

[0091] (2) Using the mutual reachability distance, HDBSCAN constructs a minimum spanning tree to connect the data points and form a preliminary clustering structure.

[0092] (3) Starting from the highest density level, gradually reduce the density threshold. At each density level, connect the points that meet the density conditions to form clusters. As the density threshold decreases, the clusters will gradually merge to form a hierarchical structure.

[0093] (4) Select clusters that persist at different density levels to evaluate the stability of clusters.

[0094] (5) Based on the optimal clustering results, define different macro dimensions, such as natural dimension, policy dimension, company dimension and social dimension, etc. Each cluster represents a macro dimension, and classify the secondary indicators into these dimensions. The specific implementation process of step S4 is as follows:

[0095] S41. Construct feature vector: Represent each secondary index as a high-dimensional feature vector v j , the specific calculation formula is as follows:

[0096] v j =[Weight(w1,M j ),Weight(w2,M j ),…,Weight(w n ,M j )];

[0097] Where n is the total number of keywords, M jis each secondary index, and its weight with keyword w is Weight(w,M j ).

[0098] S42. Based on density estimation, HDBSCAN introduces the concept of mutual reachability distance to measure the density-aware distance between two points. This distance metric takes into account the local density of the two points and the actual distance between them. Using the mutual reachability distance, HDBSCAN constructs a minimum spanning tree to connect the data points to form a preliminary structure of the cluster. The specific calculation formula is as follows:

[0099]

[0100] G = MST(X,mrd);

[0101] Where X = {v1,v2,…,v m} is a set of secondary indicators, v i is the eigenvector of the ith index, v j is the eigenvector of the jth indicator, core_dist k (v i ) refers to the kth nearest core distance, d is the distance metric, mrd(v i ,v j ) refers to the mutual reachability distance between the i-th and j-th indicators, G is a minimum spanning tree constructed using the mutual reachability distance, MST is the minimum spanning tree, Indicates v i The feature vector of the kth nearest neighbor, core_dist k (v j ) represents the kth nearest core distance of the jth index, d(v i ,v j ) indicates v i and v j The distance metric of .

[0102] S43, starting from the highest density level, gradually reduce the density threshold, at each density level, connect the points that meet the density conditions to form clusters. As the density threshold decreases, the clusters will gradually merge to form a hierarchical structure. The specific calculation formula is as follows:

[0103]

[0104] H={S(λ1),S(λ2),…,S(λ k )};

[0105] Among them, λ is the density threshold, S(λ) is the clustering formed by the secondary indicators when the density threshold is λ, and H is the hierarchical structure of the secondary indicators.

[0106] S44. Select clusters that persist at different density levels to evaluate the stability of clusters. This method can automatically determine the number of clusters and discover clusters of different scales and densities. The specific calculation formula is as follows:

[0107] S stable = argmax S∈H stability(S);

[0108] Among them, stability(S) is the stability measure of cluster S, which can be calculated based on the existence time and consistency of clusters at different density levels.

[0109] S45. According to the clustering result S stable , define different macro dimensions, such as natural dimension, policy dimension, company dimension and social dimension, etc. Each cluster represents a macro dimension, and classifies the secondary indicators into these dimensions. The specific calculation formula is as follows:

[0110] F i =S i ;

[0111] M j ∈F i if v j ∈S i ;

[0112] Among them, F i represents the i-th macroscopic dimension, S i Represents the set of secondary indicators associated with the i-th macro dimension.

[0113] The core content of the climate risk assessment described in step S5 includes: displaying the results through interactive visualization tools such as dashboards, allowing users to adjust parameters and browse different assessment results, outputting key information including distribution and trends, and providing explanations to help decision makers understand the sources and impacts of risks, so as to formulate corresponding strategies and measures.

[0114] Climate risk assessment not only reflects the dynamic impact of climate change on macro dimensions such as nature, government, companies and society, but also reveals its influence and impact trends in different dimensions. The climate risk comprehensive assessment indicator system is a series of representative climate risk indicators identified and defined through in-depth analysis of multi-source data (such as research reports, policy documents, news reports, social media content, and corporate sustainability reports). These indicators include, but are not limited to, climate physical risk indicators, climate policy uncertainty indicators, climate concern indicators, etc.

[0115] By combining Top2Vec automatic topic modeling technology and HDBSCAN clustering algorithm, efficient processing and in-depth understanding of climate risk data are achieved; multi-source data integration and preprocessing processes are optimized, data quality and availability are improved, and the multi-level structure of climate risk is effectively captured; the Top2Vec automatic topic modeling method is used to identify topics closely related to climate risk, and detailed secondary climate risk indicators are established based on these topics, and the keyword weights are determined by combining the probability distribution of Top2Vec and the TF-IDF value, which improves the accuracy and interpretability of indicator quantification; the HDBSCAN algorithm is used to cluster secondary indicators into macro dimensions such as nature, policy, company and society, and comprehensively capture the multi-level structure of climate risk; through interactive visualization tools, users can intuitively browse and analyze climate risks, support real-time monitoring and early warning mechanisms, and ensure the timeliness and accuracy of assessments; through automated processing and intelligent analysis, the construction of a data-driven indicator system simplifies the data preprocessing process, reduces costs, and provides intelligent solutions for climate risk management, policy formulation, and corporate strategic planning. The embodiments of the present invention not only deepen the understanding of various climate risk indicators, but also provide a reliable basis for observing the macro impact of climate risks, and promote the intelligence of climate risk management and decision-making.

[0116] In summary, the embodiment of the present invention proposes a method for constructing a climate risk comprehensive assessment index system with a multi-level structure, aiming to efficiently identify and deeply analyze the multi-dimensional characteristics of climate risks, and quantify the specific impact of each dimension. This method integrates Top2Vec automatic topic modeling technology and HDBSCAN clustering algorithm to adaptively and dynamically analyze the complexity and diversity of climate risk data, effectively process large-scale data sets from different sources, and capture the multi-level structure of climate risks. The embodiment of the present invention can not only identify a series of typical climate risk indicators, such as climate risk, climate physical risk, climate policy uncertainty, etc., but also reveal special climate risk indicators such as management climate concerns, climate risk impacts, climate change news, climate risk concerns, climate risk awareness, and corporate biodiversity; in addition, the embodiment of the present invention can also discover the specific manifestations of climate risks in macro dimensions such as nature, policy, company and society, and comprehensively capture the multi-level structure of climate risks, providing a comprehensive tool for climate risk management, relevant standard formulation, and corporate strategic planning, so that decision makers can better understand and respond to the challenges brought by climate change.

[0117] System Example

[0118] According to an embodiment of the present invention, a system for constructing a climate risk comprehensive assessment indicator system with a multi-level structure is provided. Figure 6 : is a schematic diagram of a system for constructing a multi-level climate risk comprehensive assessment index system according to an embodiment of the present invention. Figure 6As shown, the multi-level structure climate risk comprehensive assessment indicator system construction system according to an embodiment of the present invention specifically includes:

[0119] A data module 60 is used to obtain multi-source data related to climate risk, and pre-process the multi-source data to obtain a climate risk data set;

[0120] A quantification module 62, configured to construct a secondary climate risk indicator according to the climate risk data set, and quantify the secondary climate risk indicator to obtain a secondary climate risk indicator value; and

[0121] A construction module 64 is used to classify the secondary climate risk indicators into corresponding macro dimensions based on the values ​​of the secondary climate risk indicators, and to construct a comprehensive climate risk assessment indicator system through the macro dimensions and the secondary climate risk indicators;

[0122] The system further comprises:

[0123] The visualization module is used to display and analyze the climate risk comprehensive assessment indicator system using interactive visualization tools to obtain the climate risk comprehensive assessment results.

[0124] The embodiment of the present invention is a system embodiment corresponding to the above-mentioned method embodiment. The specific operations of each module can be understood by referring to the description of the method embodiment, which will not be repeated here.

[0125] In summary, the embodiment of the present invention provides a method for constructing a comprehensive assessment index system of climate risk based on multi-source data fusion. By integrating text information from different channels, the multi-dimensional characteristics of climate risk are comprehensively identified and analyzed, the specific impact of each dimension is quantified, and the hierarchical structure is revealed. The method is mainly implemented through the following five steps: (1) Multi-source data integration: collect and process text information related to climate risk and build a database; (2) Secondary index setting: identify topics closely related to climate risk through Top2Vec automatic topic modeling technology, and establish a series of secondary climate risk indicators based on these topics; (3) Climate risk quantification: combine the Top2Vec probability distribution and TF-IDF value of each keyword to determine the weight of the keyword, and perform weighted summation on the keywords under each secondary indicator to calculate the specific value of each secondary indicator; (4) Macro-dimensional induction: apply the HDBSCAN algorithm to cluster each secondary indicator into different macro-dimensions, such as natural dimension, policy dimension, company dimension and social dimension, revealing the multi-level structure of climate risk; (5) Climate risk assessment: use a comprehensive indicator system and visualization tools to conduct a comprehensive assessment and support real-time monitoring and early warning. The embodiment of the present invention combines topic modeling, cluster analysis and text quantification technology to achieve comprehensive identification and evaluation of the multi-dimensional characteristics of climate risks on the one hand, and to ensure the autonomous adjustment and efficient construction of the indicator system on the other hand. Through this method, a systematic and comprehensive analysis tool is provided for climate risk management, relevant standard formulation and corporate strategic planning, which helps decision makers better understand and respond to the challenges brought by climate change. This method not only improves the efficiency and accuracy of climate risk assessment, but also enhances the interpretability and practicality of the results, providing strong theoretical support for decision makers.

[0126] Device Example 1

[0127] An embodiment of the present invention provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the steps described in the method embodiment when executed by the processor.

[0128] Device Example 2

[0129] An embodiment of the present invention provides a computer-readable storage medium, on which a program for implementing information transmission is stored. When the program is executed by a processor, the steps described in the method embodiment are implemented.

[0130] The computer-readable storage medium in this embodiment includes, but is not limited to, ROM, RAM, magnetic disk or optical disk, etc.

[0131] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for constructing a multi-level climate risk comprehensive assessment indicator system, characterized by include: Acquire multi-source data related to climate risk, and pre-process the multi-source data to obtain a climate risk data set; Constructing a secondary climate risk indicator according to the climate risk data set, and quantifying the secondary climate risk indicator to obtain a secondary climate risk indicator value; as well as Based on the values ​​of the climate risk secondary indicators, the climate risk secondary indicators are summarized into corresponding macro dimensions, and a climate risk comprehensive assessment indicator system is constructed through the macro dimensions and the climate risk secondary indicators.

2. The method according to claim 1, characterized in that The method further comprises: Interactive visualization tools are used to display and analyze the climate risk comprehensive assessment indicator system to obtain the climate risk comprehensive assessment results.

3. The method according to claim 1, characterized in that Preprocessing the multi-source data to obtain a climate risk data set specifically includes: Preliminarily processing the multi-source data, and using natural language processing technology to perform word segmentation on the preliminarily processed multi-source data, and constructing a climate risk data set based on the word items after the word segmentation processing; Wherein, the multi-source data is text data; The preliminary processing includes removal of stop words, punctuation marks and numbers, and normalization.

4. The method according to claim 1, characterized in that The secondary indicators of climate risk constructed based on the climate risk data set specifically include: The climate risk dataset is trained using Doc2Vec word embedding technology to generate word vectors, which are mapped to vector space to obtain high-dimensional document vectors. The high-dimensional document vectors are reduced in dimension using the UMAP algorithm, and the reduced-dimensional document vectors are clustered using the Mean Shift algorithm. Keywords are extracted using the TF-IDF method based on the clustering results to obtain the theme of each cluster and its corresponding keywords, and the theme is used as a secondary indicator of climate risk. Among them, the secondary climate risk indicators include climate risk indicators, climate physical risk indicators, climate policy uncertainty indicators, climate management concern indicators, climate risk impact indicators, climate change news indicators, climate risk concern indicators, climate risk awareness indicators and corporate biodiversity indicators.

5. The method according to claim 4, characterized in that The secondary climate risk indicators are quantified to obtain the following secondary climate risk indicator values: The weight of each keyword is obtained using Formula 1 according to the Top2vec probability distribution and TF-IDF value of each keyword. The frequency of each keyword at each time point under each secondary indicator is weighted and summed according to the weight using Formula 2 to obtain the specific value of each secondary indicator at each time point. The specific value of each secondary indicator at each time point is subjected to Z-score normalization to obtain the final climate risk secondary indicator value; Weight(w,M j ) = α·p(w|M j )+(1-α)·TF-IDF(w,M j ) Formula 1; Among them, p(w|M j ) represents each keyword w∈K j Top2Vec probability distribution, TF-IDF(w,M j ) represents the TF-IDF value of each keyword w, M j is the jth secondary index, M j ={K j }, α is the adjustment factor, for each time point t∈T, T represents the set of time points, V(M j ,t) represents the jth secondary index M j The specific value at time point t, f(w,t,M j ) indicates the statistics of each secondary index M j Frequency of keywords.

6. The method according to claim 5, characterized in that Based on the values ​​of the secondary climate risk indicators, the secondary climate risk indicators are classified into corresponding macro dimensions, including: The characteristic vector of each secondary indicator is obtained through keyword weights, and the HDBSCAN algorithm is used for clustering based on the characteristic vector of each secondary indicator. Different macro dimensions are defined according to the optimal clustering results, and the climate risk secondary indicators are classified into corresponding macro dimensions.

7. A multi-level climate risk comprehensive assessment indicator system construction system, characterized by include: A data module, used to obtain multi-source data related to climate risks, and pre-process the multi-source data to obtain a climate risk data set; A quantification module, used to construct a climate risk secondary indicator based on the climate risk data set, and quantify the climate risk secondary indicator to obtain a climate risk secondary indicator value; as well as A construction module is used to summarize the climate risk secondary indicators into corresponding macro dimensions based on the values ​​of the climate risk secondary indicators, and to construct a climate risk comprehensive assessment indicator system through the macro dimensions and the climate risk secondary indicators.

8. The system according to claim 7, characterized in that The system further comprises: The visualization module is used to display and analyze the climate risk comprehensive assessment indicator system using interactive visualization tools to obtain the climate risk comprehensive assessment results.

9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the steps of the method for constructing a multi-level climate risk comprehensive assessment indicator system as described in any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores an implementation program for information transmission, and when the program is executed by the processor, the steps of the method for constructing a multi-level climate risk comprehensive assessment indicator system according to any one of claims 1 to 6 are implemented.

Images