Enterprise ESG risk assessment method and enterprise ESG risk assessment system

Abstract

This application discloses an enterprise ESG risk assessment method and an enterprise ESG risk assessment system, which involve the field of comprehensive data analysis technology, realize comprehensive, systematic and dynamic ESG risk assessment, help enterprises better fulfill their social responsibilities, enhance their market competitiveness, and provide strong guarantees for the sustainable development of enterprises. The method includes: responding to an enterprise risk assessment request, determining the target enterprise to be risk assessed; obtaining the supply chain network to be assessed, and querying whether the target enterprise has upstream supplier enterprises in the supply chain network to be assessed; when the query determines that the target enterprise has upstream supplier enterprises, determining the first ESG individual risk value corresponding to the target enterprise, and determining the second ESG individual risk value corresponding to the upstream supplier enterprise; based on the first ESG individual risk value and the second ESG individual risk value, calculating the ESG relationship risk value of the target enterprise, and outputting the ESG relationship risk value.

Description

Technical Field

[0001] The present application relates to the technical field of comprehensive data analysis, and in particular to an enterprise ESG risk assessment method, device, electronic device and storage medium. Background Art

[0002] As global environmental and climate issues become increasingly prominent, countries are placing increasing emphasis on sustainable development and environmental protection. Against this backdrop, the concept of ESG (Environmental, Social, and Governance) investing has become a core issue in the investment landscape. ESG investing emphasizes that while pursuing economic returns, companies should actively fulfill their social responsibilities, prioritize environmental protection and social welfare, and implement effective corporate governance measures. The widespread adoption and promotion of this concept is crucial for promoting sustainable development in the global economy.

[0003] In recent years, as ESG investment concepts have gained widespread recognition, more and more investors have begun incorporating ESG factors into their investment decisions. The automotive industry, a key pillar of the national economy, has also attracted widespread attention for its ESG performance. As the automotive industry faces increasing pressures from economic volatility, climate change, and supply chain challenges, the demand for high-quality, sustainable development is becoming increasingly urgent. Consequently, a growing number of automotive companies are prioritizing ESG practices and increasing their investment.

[0004] However, amid the ESG investment craze, companies face numerous risks and challenges. First, ESG standards and requirements vary across countries and regions, requiring companies to adapt to diverse regulatory environments and market requirements. Second, as ESG investment concepts gain popularity, more and more investors are paying attention to a company's ESG performance, which will impact their financing costs and access to financing.

[0005] To address these issues, companies in the automotive industry need to accurately identify and effectively mitigate ESG risks. However, due to the industry's long supply chain, wide reach, and numerous upstream and downstream companies, companies face significant challenges in identifying and managing ESG risks. This is particularly true within the supply chain, where the numerous and complex suppliers involved, coupled with limited data transparency and traceability, often make it difficult for companies to accurately assess ESG risks within the supply chain. Therefore, to enhance the ESG risk identification and prevention capabilities of companies across the automotive supply chain and promote the sustainable development of the industry, a comprehensive, systematic, and dynamic ESG risk assessment approach is urgently needed. Summary of the Invention

[0006] In view of this, the present application provides an enterprise ESG risk assessment method and an enterprise ESG risk assessment system.

[0007] According to the first aspect of this application, a method for enterprise ESG risk assessment is provided, the method comprising:

[0008] In response to enterprise risk assessment requests, determine the target enterprises to be risk assessed;

[0009] Obtaining a supply chain network to be evaluated, and querying whether the target enterprise has an upstream supplier enterprise in the supply chain network to be evaluated, wherein the target enterprise is in the supply chain network to be evaluated, and the supply chain network to be evaluated includes multiple enterprises and records the supply relationships between the multiple enterprises;

[0010] When the query determines that the target enterprise has an upstream supplier enterprise, determine the first ESG individual risk value corresponding to the target enterprise and the second ESG individual risk value corresponding to the upstream supplier enterprise;

[0011] Calculate the ESG relationship risk value of the target enterprise based on the first ESG individual risk value and the second ESG individual risk value, and output the ESG relationship risk value.

[0012] According to the second aspect of the present application, a system for assessing ESG risks in an enterprise is provided, the system comprising:

[0013] A risk assessment module, configured to respond to enterprise risk assessment requests and determine target enterprises to be risk assessed;

[0014] The risk assessment module is further configured to obtain a supply chain network to be assessed and query whether the target enterprise has any upstream supplier enterprises in the supply chain network to be assessed, wherein the target enterprise is in the supply chain network to be assessed, and the supply chain network to be assessed includes multiple enterprises and records the supply relationships between the multiple enterprises;

[0015] The risk assessment module is further configured to determine a first ESG individual risk value corresponding to the target enterprise and a second ESG individual risk value corresponding to the upstream supplier enterprise when the query determines that the target enterprise has an upstream supplier enterprise;

[0016] The risk assessment module is further configured to calculate the ESG relationship risk value of the target enterprise based on the first ESG individual risk value and the second ESG individual risk value, and output the ESG relationship risk value.

[0017] By means of the above technical solution, the present application provides an enterprise ESG risk assessment method and an enterprise ESG risk assessment system. The present application responds to the enterprise risk assessment request, determines the target enterprise to be risk assessed, obtains the supply chain network to be assessed, and inquires whether the target enterprise has upstream supplier enterprises in the supply chain network to be assessed. When the query determines that the target enterprise has upstream supplier enterprises, the first ESG individual risk value corresponding to the target enterprise is determined, and the second ESG individual risk value corresponding to the upstream supplier enterprise is determined. Based on the first ESG individual risk value and the second ESG individual risk value, the ESG relationship risk value of the target enterprise is calculated, and the ESG relationship risk value is output. When conducting ESG risk assessment on the enterprise, the specific position of the enterprise in the supply chain network to be assessed and the connection with upstream and downstream enterprises can be referred to at the same time, so as to realize comprehensive, systematic and dynamic ESG risk assessment, accurately identify enterprise ESG risks and conduct effective early warning, which helps enterprises better fulfill their social responsibilities, enhance market competitiveness and provide strong guarantees for the sustainable development of enterprises.

[0018] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present application. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0020] Figure 1 A schematic diagram of a process for enterprise ESG risk assessment provided by an embodiment of the present application is shown;

[0021] Figure 2A A flow chart of another enterprise ESG risk assessment method provided in an embodiment of the present application is shown;

[0022] Figure 2B A flowchart of another enterprise ESG risk assessment method provided in an embodiment of the present application is shown;

[0023] Figure 3 A schematic diagram of the structure of an enterprise ESG risk assessment system provided by an embodiment of the present application is shown;

[0024] Figure 4 A schematic diagram of the device structure of a computer device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0025] The following describes exemplary embodiments of the present application in more detail with reference to the accompanying drawings. Although exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0026] In today's globalized and increasingly stringent business environment, the automotive industry faces unique challenges, particularly with its environmental, social, and governance (ESG) performance receiving unprecedented attention. Automakers must not only ensure product quality and safety, but also strive to reduce environmental impact, enhance social responsibility, and achieve efficient governance. Therefore, this application proposes a method for enterprise ESG risk assessment. Specifically, this method is an ESG risk assessment and early warning method for automotive companies based on risk grading and supply chain networks. This method comprehensively considers ESG factors across all links in the automotive industry chain and, through the construction of an indicator system module, a data mapping module, a risk assessment module, and a risk early warning module, accurately identifies and effectively warns of ESG risks for each company in the automotive industry chain. The method proposed in this application can effectively identify and manage ESG risks within the automotive industry chain, which will help automotive companies better fulfill their social responsibilities, enhance their market competitiveness, and provide strong support for their sustainable development.

[0027] This application embodiment provides an enterprise ESG risk assessment method, such as Figure 1 As shown, the method includes:

[0028] 101. In response to an enterprise risk assessment request, determine the target enterprise to be risk assessed.

[0029] The embodiments of the present application can be applied to an enterprise ESG risk assessment system, which relies on the computing power of a server to provide risk assessment services to the enterprise. The server can be a standalone server or a server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms, etc., to provide ESG risk assessment services to the enterprise.

[0030] Among them, the enterprise ESG risk assessment system can provide a front-end application for enterprises. When there is a need for risk assessment, the enterprise staff can specify which enterprise to assess in the front-end application, and send an application enterprise risk assessment request to the enterprise ESG risk assessment system based on the front-end application. In response to the enterprise risk assessment request, the enterprise ESG risk assessment system will determine the target enterprise to be risk assessed, which is the enterprise specified by the enterprise staff on the front end, and start to perform the risk assessment operation. It should be noted that the target enterprise can be an automobile enterprise, a financial enterprise, etc., and this application does not specifically limit the type of target enterprise; in addition, for the convenience of explanation, the embodiment of this application is described as an automobile enterprise.

[0031] 102. Obtain the supply chain network to be evaluated and check whether the target enterprise has any upstream supplier enterprises in the supply chain network to be evaluated.

[0032] Considering that the automotive industry's supply chain involves many suppliers and is complex, and the transparency and traceability of supply chain data are low, it is often difficult for companies to accurately assess ESG risks in the supply chain. Therefore, in order to better monitor and evaluate the ESG performance of each company in the automotive industry chain, the embodiment of this application will obtain the supply chain network to be evaluated, query whether the target company has upstream suppliers in the supply chain network to be evaluated, and then subsequently comprehensively consider the situation of the target company's upstream suppliers to assess the risks of the target company.

[0033] Among them, the target enterprise is in the supply chain network to be evaluated, which includes multiple enterprises and records the supply relationships between multiple enterprises. Therefore, it is possible to query whether the target enterprise has upstream supplier enterprises in the supply chain network to be evaluated.

[0034] 103. When the query determines that the target enterprise has an upstream supplier enterprise, determine the first ESG individual risk value corresponding to the target enterprise and determine the second ESG individual risk value corresponding to the upstream supplier enterprise.

[0035] In an embodiment of the present application, when a query determines that the target enterprise has an upstream supplier enterprise, in order to more scientifically and effectively monitor and evaluate the ESG risks of each enterprise in the supply chain network to be evaluated, it is not possible to only consider the ESG risk situation of a single enterprise during the calculation, but it is also necessary to refer to the specific position of the enterprise in the supply chain network to be evaluated and its connection with upstream and downstream enterprises. Therefore, the enterprise ESG risk assessment system will determine the first ESG individual risk value corresponding to the target enterprise and the second ESG individual risk value corresponding to the upstream supplier enterprise, and then take into account the potential ESG risks brought to the target enterprise by the upstream supplier enterprise in the risk assessment process of the target enterprise.

[0036] Among them, taking the target enterprise as an example, the enterprise itself has some violation information and public opinion information, which can bring risks to the enterprise itself. This risk is caused by the enterprise itself and should also be considered in the risk assessment process of the target enterprise. Therefore, this application proposes the concept of ESG individual risk value. The ESG individual risk value is a risk value calculated based on the enterprise's own violation information and public opinion information to indicate the enterprise's own risk situation. This risk value does not take into account the potential risks brought to the enterprise by the risk values of other suppliers. It is an independent risk value determined by the enterprise's own situation.

[0037] 104. Calculate the ESG relationship risk value of the target enterprise based on the first ESG individual risk value and the second ESG individual risk value, and output the ESG relationship risk value.

[0038] In an embodiment of the present application, after determining the first ESG individual risk value and the second ESG individual risk value, the enterprise ESG risk assessment system will calculate the ESG relationship risk value of the target enterprise based on the first ESG individual risk value and the second ESG individual risk value, and output the ESG relationship risk value so that the potential ESG risks brought to the target enterprise by the upstream supplier enterprises are also taken into account in the risk assessment process of the target enterprise. In this way, the output ESG relationship risk value is also the ESG risk of the target enterprise after considering the supply relationship in the supply chain network to be assessed.

[0039] The method provided in the embodiment of the present application responds to a risk assessment request from an enterprise, determines a target enterprise to be risk assessed, obtains a supply chain network to be assessed, queries whether the target enterprise has an upstream supplier enterprise in the supply chain network to be assessed, and when the query determines that the target enterprise has an upstream supplier enterprise, determines a first ESG individual risk value corresponding to the target enterprise, determines a second ESG individual risk value corresponding to the upstream supplier enterprise, calculates the ESG relationship risk value of the target enterprise based on the first ESG individual risk value and the second ESG individual risk value, and outputs the ESG relationship risk value. When conducting an ESG risk assessment on an enterprise, it can simultaneously refer to the specific position of the enterprise in the supply chain network to be assessed and its connection with upstream and downstream enterprises, thereby achieving a comprehensive, systematic and dynamic ESG risk assessment, accurately identifying the ESG risks of the enterprise and conducting effective early warnings, which will help enterprises better fulfill their social responsibilities, enhance their market competitiveness, and provide strong guarantees for the sustainable development of the enterprise.

[0040] Furthermore, as a refinement and extension of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, the embodiment of this application provides another enterprise ESG risk assessment method, such as Figure 2A As shown, the method includes:

[0041] 201. In response to an enterprise risk assessment request, determine a target enterprise for risk assessment.

[0042] The embodiments of the present application can be applied to an enterprise ESG risk assessment system, which relies on the computing power of a server to provide risk assessment services to the enterprise. The server can be a standalone server or a server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms, etc., to provide ESG risk assessment services to the enterprise.

[0043] Among them, the enterprise ESG risk assessment system can provide a front-end application for enterprises. When there is a need for risk assessment, the enterprise staff can specify which enterprise to assess in the front-end application, and send an application enterprise risk assessment request to the enterprise ESG risk assessment system based on the front-end application. In response to the enterprise risk assessment request, the enterprise ESG risk assessment system will determine the target enterprise to be risk assessed, which is the enterprise specified by the enterprise staff on the front end, and start to perform the risk assessment operation. It should be noted that the target enterprise can be an automobile enterprise, a financial enterprise, etc., and this application does not specifically limit the type of target enterprise; in addition, for the convenience of explanation, the embodiment of this application is described as an automobile enterprise.

[0044] 202. Obtain the supply chain network to be evaluated and query whether the target enterprise has any upstream supplier enterprises in the supply chain network to be evaluated. If the query determines that the target enterprise has an upstream supplier enterprise, execute the following steps 203 and 204; if the query determines that the target enterprise does not have an upstream supplier enterprise, execute the following step 205.

[0045] Considering that the automotive industry's supply chain involves many suppliers and is complex, and the transparency and traceability of supply chain data are low, it is often difficult for companies to accurately assess ESG risks in the supply chain. Therefore, in order to better monitor and evaluate the ESG performance of each company in the automotive industry chain, the embodiment of this application will obtain the supply chain network to be evaluated, query whether the target company has upstream suppliers in the supply chain network to be evaluated, and then subsequently comprehensively consider the situation of the target company's upstream suppliers to assess the risks of the target company.

[0046] Among them, the target enterprise is in the supply chain network to be evaluated, which includes multiple enterprises and records the supply relationships between multiple enterprises. Therefore, it is possible to query whether the target enterprise has upstream supplier enterprises in the supply chain network to be evaluated.

[0047] Furthermore, if the query determines that the target enterprise has upstream suppliers, in order to more scientifically and effectively monitor and assess the ESG risks of each enterprise in the supply chain network to be assessed, the calculation must not only consider the ESG risk of a single enterprise, but also consider the enterprise's specific position in the supply chain network to be assessed and its connections with upstream and downstream enterprises. Therefore, the process from steps 203 to 204 is executed. If the query determines that the target enterprise has no upstream suppliers, it means that no supplier enterprises pose potential ESG risks to the target enterprise. The target enterprise only needs to consider its own ESG risks. Therefore, step 205 is executed.

[0048] 203. When the query determines that the target enterprise has an upstream supplier enterprise, determine the first ESG individual risk value corresponding to the target enterprise and determine the second ESG individual risk value corresponding to the upstream supplier enterprise.

[0049] In an embodiment of the present application, when a query determines that the target enterprise has an upstream supplier enterprise, in order to more scientifically and effectively monitor and evaluate the ESG risks of each enterprise in the supply chain network to be evaluated, it is not possible to only consider the ESG risk situation of a single enterprise during the calculation, but it is also necessary to refer to the specific position of the enterprise in the supply chain network to be evaluated and its connection with upstream and downstream enterprises. Therefore, the enterprise ESG risk assessment system will determine the first ESG individual risk value corresponding to the target enterprise and the second ESG individual risk value corresponding to the upstream supplier enterprise, and then take into account the potential ESG risks brought to the target enterprise by the upstream supplier enterprise in the risk assessment process of the target enterprise.

[0050] Among them, taking the target enterprise as an example, the enterprise itself has some violation information and public opinion information, which can bring risks to the enterprise itself. This risk is caused by the enterprise itself and should also be considered in the risk assessment process of the target enterprise. Therefore, this application proposes the concept of ESG individual risk value. The ESG individual risk value is a risk value calculated based on the enterprise's own violation information and public opinion information to indicate the enterprise's own risk situation. This risk value does not take into account the potential risks brought to the enterprise by the risk values of other suppliers. It is an independent risk value determined by the enterprise's own situation.

[0051] Specifically, when querying the ESG individual risk values of the target enterprise and upstream supplier enterprises, considering that this risk assessment may not be the first assessment, the ESG individual risk value may have been calculated for each enterprise in the supply chain network to be assessed in the past. Alternatively, although this is the first risk assessment, the ESG individual risk value of each enterprise may have been calculated in advance, and enterprise individual risk information has been formed based on all the calculated ESG individual risk values. In other words, the enterprise individual risk information is used to record the ESG individual risk value corresponding to each enterprise in the supply chain network to be assessed. The ESG individual risk value of each enterprise can be directly queried in the enterprise individual risk information. Therefore, the enterprise ESG risk assessment system will query whether the supply chain network to be assessed has enterprise individual risk information. If the query determines that enterprise individual risk information exists in the supply chain network to be assessed, it means that the ESG individual risk value of each enterprise has been calculated and can be used directly without further calculation. Therefore, the enterprise ESG risk assessment system will directly obtain the ESG individual risk value corresponding to the target enterprise from the enterprise individual risk information as the first ESG individual risk value, and obtain the ESG individual risk value corresponding to the upstream supplier enterprise from the enterprise individual risk information as the second ESG individual risk value. When the query determines that there is no enterprise individual risk information in the supply chain network to be evaluated, it means that the ESG individual risk value has not been calculated for the enterprise yet and the calculation process needs to be performed. Therefore, the enterprise ESG risk assessment system will calculate the ESG individual risk value for each enterprise in the supply chain network to be evaluated, and obtain multiple ESG individual risk values corresponding to multiple enterprises. The multiple ESG individual risk values will be used as the enterprise individual risk information, and the enterprise individual risk information will be stored.

[0052] The enterprise ESG risk assessment system is divided into an indicator system module, a data mapping module, and a risk assessment module to calculate the ESG individual risk value. The following is a detailed introduction to these three modules:

[0053] 1. Indicator system module.

[0054] In today's globalized and increasingly stringently regulated business environment, the automotive industry faces unique challenges, particularly with environmental, social, and governance (ESG) performance receiving unprecedented attention. Automakers must not only ensure product quality and safety, but also strive to reduce environmental impact, enhance social responsibility, and achieve efficient governance. Therefore, this application embodiment constructs an ESG indicator system consisting of three primary indicators, six secondary indicators, and twelve tertiary indicators, aiming to provide automotive companies with a clear assessment and management framework to optimize their overall ESG performance.

[0055] First of all, environment (E) is the first first-level indicator of this system, covering two second-level indicators of "environmental risk management" and "resource use and efficiency". In terms of environmental risk management, the embodiment of this application pays special attention to the two third-level indicators of "environmental protection law and regulations compliance rate" and "frequency of major environmental incidents". This is because compliance with environmental protection laws and regulations is not only a legal requirement, but also a manifestation of corporate social responsibility, and effectively reducing the frequency of major environmental incidents is directly related to the sustainable development and public image of the enterprise. In terms of resource use and efficiency, the embodiment of this application focuses on the two third-level indicators of "energy efficiency and compliance" and "water resource management" to ensure that enterprises use energy and water resources efficiently in the production process and comply with relevant environmental protection standards and regulations.

[0056] Society (S) is the second first-level indicator, comprising two second-level indicators: "Supply Chain Responsibility" and "Employee Welfare and Safety." Regarding supply chain responsibility, this embodiment uses two third-level indicators: "Supplier Compliance Rate" and "Supply Chain Transparency" to ensure supply chain compliance and transparency. This not only helps reduce operational risks but also enhances consumer trust in the brand. Regarding employee welfare and safety, two third-level indicators: "Employee Health and Safety" and "Employee Development and Training" ensure that companies provide employees with a safe working environment and development opportunities, which directly impacts employee satisfaction and the company's long-term stability.

[0057] Governance (G) is the third first-level indicator, covering the two second-level indicators of "risk management and compliance" and "governance structure and transparency". In terms of risk management and compliance, the two third-level indicators of "risk identification and assessment" and "compliance management system" help companies identify potential internal and external risks and respond to these risks through an effective compliance management system. In terms of governance structure and transparency, the two third-level indicators of "board effectiveness" and "stakeholder communication" ensure the efficiency and transparency of corporate governance, and enhance the company's governance level and sustainability by improving the independence and effectiveness of the board of directors and strengthening communication with stakeholders. The ESG evaluation index system constructed in the embodiment of this application is shown in Table 1 below:

[0058] Table 1

[0059]

[0060] First, this ESG evaluation index system adopts a hierarchical design, dividing ESG evaluation into secondary and tertiary indicators, making the evaluation content more systematic and structured, helping companies better understand and evaluate their own ESG performance, while also providing a clear reference framework for investors and stakeholders; secondly, this system covers many important areas of enterprises in the three aspects of environment, society and governance. The secondary indicators are very specific, including all aspects of the company's ESG practices, ensuring the comprehensiveness of the evaluation of the company's ESG performance; finally, this indicator system is used to assist in the ESG risk assessment of the company's supply chain. By comparing the size of the company's ESG risks under different indicators, the company's shortcomings and advantages in ESG can be discovered, thereby promoting the continuous improvement of its ESG management.

[0061] 2. Data mapping module.

[0062] To scientifically, objectively, and comprehensively assess ESG risks within the automotive supply chain, it's crucial to first collect ESG-related information and data. Specifically, this data can be obtained through the established automotive supply chain ESG risk early warning platform. This platform, based on a comprehensive perspective across the entire automotive supply chain, encompasses extensive data across three dimensions: environmental, social, and corporate governance. These data resources provide the foundational risk data and supply chain network structure for conducting ESG risk assessments and early warnings, serving as sample data.

[0063] Next, we need to classify the above data information and complete the mapping of each sample data to the corresponding three-level indicators in the ESG evaluation index system. We also need to set indicator weights for each three-level indicator so that we can conduct a comprehensive ESG risk assessment of the enterprise supply chain based on multiple indicators. In this embodiment of the application, NLP (natural language processing) methods are used for data mapping. This process mainly includes the following steps:

[0064] Step 1: Text preprocessing: The enterprise ESG risk assessment system removes noise, standardizes the format, and performs word segmentation on the collected sample data.

[0065] Step 2: Feature Extraction. The enterprise ESG risk assessment system extracts features from multiple sample data points and converts them into numerical form. Specifically, the system counts the frequency of each word in the sample data, determines word order, identifies contextual information, or uses word embeddings to represent word semantics. This converts the textual sample data into numerical form that the model can understand.

[0066] Step 3: Model Selection and Training. The enterprise ESG risk assessment system acquires a text category prediction model. This model is a language recognition model trained using labeled training data to predict the category of a text. In practical applications, a modified BERT deep learning model can be selected for classification. This model is then trained using a labeled training dataset to obtain a text category prediction model. This model then learns how to predict the category of a text based on input features.

[0067] Step 4: Model Evaluation and Optimization. The enterprise ESG risk assessment system uses the test set to evaluate the performance of the text category prediction model. Based on the evaluation results, the parameters of the text category prediction model are adjusted or a different model is selected to optimize performance. Finally, the trained text category prediction model is applied in the embodiment of this application to complete the mapping of the sample data.

[0068] Step 5: Data mapping. The enterprise ESG risk assessment system will input the converted multiple sample data into the text category prediction model for identification, and obtain the data category of each sample data. Next, determine a preset number of evaluation indicators, and divide the multiple sample data into a preset number of evaluation indicators according to the mapping relationship between the evaluation indicators and the data categories, and count the number of samples of the sample data divided into each evaluation indicator, and use the entropy weight method to calculate the number of samples of each evaluation indicator to obtain the indicator weight corresponding to each evaluation indicator. Among them, the preset number of evaluation indicators in the actual application process can be the 12 three-level indicators described in Table 1 above. After the data mapping process, the amount of ESG risk information of each enterprise under each evaluation indicator can be statistically obtained, so the entropy weight method can be used to determine the weight of each evaluation indicator. Specifically, the weight of the evaluation indicator is expressed by W k , where k = 1, …, 12, representing 12 third-level indicators. In subsequent risk assessments, we first assess the ESG risk level of each company under different indicators, and then conduct a more comprehensive analysis and early warning of each company's ESG risks.

[0069] 3. Risk assessment module.

[0070] In the risk assessment module, this embodiment of the application conducts ESG risk assessments from both the enterprise side and the supply chain side. On the enterprise side, the ESG risk of the enterprise itself is quantified from the perspectives of risk quantity and risk level, resulting in the enterprise's individual ESG risk value. On the supply chain side, the effect of the supply chain on risk addition is considered from the perspective of the supply chain's complex network, resulting in the output of the enterprise's ESG relationship risk value.

[0071] Specifically, the risk assessment module utilizes collected and integrated information on ESG violations and public opinion from various companies in the automotive supply chain to quantify each company's ESG risk. When quantifying risk, the module considers both the volume of corporate violations and negative public opinion (i.e., the number of risks) and the severity of penalties imposed on the company (i.e., the level of risk). Furthermore, to fully account for the impact of upstream and downstream companies in the supply chain, the module analyzes the company's importance by referencing its specific location within the automotive supply chain network and its connections with upstream and downstream companies. Based on the quantified individual ESG risks of the company itself and drawing on existing automotive company supply chain network relationships, an ESG risk index assessment model is established based on the automotive supply chain network.

[0072] In the specific assessment calculation, it is first necessary to pre-set the risk levels corresponding to different violation penalties and assign different importance to different risk levels; then use text analysis technology and the text information of each violation data to classify the risk; finally, under each assessment indicator, quantify the ESG risk of each company according to the number and risk level of risks, and then combine multiple indicators to obtain the weighted summation to obtain the ESG individual risk value of different companies.

[0073] The enterprise violation data collected by the present embodiment mainly comes from public information on various official websites. The specific content of the violation information includes important information such as violation of laws and regulations, violation type, penalty category, penalty content, penalty announcement date, etc. The specific processing process of violation information includes the following three steps:

[0074] Step 1: Classify risk levels based on the severity of the penalty. The enterprise ESG risk assessment system will classify risk levels into three levels: high, medium, and low, based on the type of penalty and the relevant provisions of the penalty standards. The specific classification results are shown in Table 2 below:

[0075] Table 2

[0076]

[0077] Step 2: Use text parsing technology to determine the risk level of the violation information. For each company in the supply chain network to be assessed, the enterprise ESG risk assessment system obtains a preset number of assessment indicators, matches the company's multiple violation information under each assessment indicator, and uses text parsing technology to process the multiple violation information under each assessment indicator, determining the corresponding risk level of each violation information. Based on the corresponding risk level of each violation information, the multiple violation information under each assessment indicator is divided into high-risk, medium-risk, and low-risk groups, and the number of violations included in each risk group is counted.

[0078] Specifically, when conducting analysis, first, the enterprise ESG risk assessment system will collect violation processing keywords corresponding to high-risk levels, medium-risk levels, and low-risk levels, and use the collected violation processing keywords to build a violation information vocabulary.

[0079] Subsequently, considering that the terminology used to describe violations has specific domain characteristics, the enterprise ESG risk assessment system will access the Jieba word segmentation tool and update its vocabulary based on the violation information vocabulary, resulting in an improved Jieba word segmentation tool. In actual application, the enterprise ESG risk assessment will perform data cleansing on the multiple violation information collected, and then use the improved Jieba word segmentation tool to segment each violation information. Furthermore, the enterprise ESG risk assessment will use a Chinese stop word list to remove stop words from the violation information after word segmentation, thereby reducing the amount of information to be processed and speeding up the processing time.

[0080] Finally, the enterprise ESG risk assessment system will use the violation information vocabulary to perform keyword matching on each violation information after word segmentation, determine the target violation processing keyword corresponding to each violation information, and then determine the corresponding risk level for each violation information according to the risk level of the target violation processing keyword in the violation information vocabulary.

[0081] In this way, the risk level corresponding to each violation information can be obtained through the above process. For the i-th enterprise in the violation information, we can use 、 、 They represent the number of high-risk, medium-risk, and low-risk violation information of enterprise i on the k-th evaluation indicator.

[0082] Step 3: Determine the risk group weight corresponding to each risk group. To indicate the differences in ESG risks that penalties at different risk levels may bring to the enterprise, this embodiment of the application assigns different importance weights to different risk levels as risk group weights. The specific risk group weights corresponding to each risk group are shown in Table 3 below:

[0083] Table 3

[0084]

[0085] In the subsequent risk quantification, the embodiment of the present application will calculate the risk value brought by each violation information based on the risk group weights of different risk levels. Subsequently, w1 is used to represent the risk group weight corresponding to high risk, w2 is used to represent the risk group weight corresponding to medium risk, and w3 is used to represent the risk group weight corresponding to low risk.

[0086] The processing of violation information is completed through the above three steps. In the embodiment of the present application, not only the violation information of the enterprise is collected, but also the public opinion information of the enterprise is collected. Among them, the relevant public opinion information of the enterprise comes from the information published by the media, including positive or negative information such as major events, business operations, major transactions, and violations involving litigation. For public opinion information, the embodiment of the present application uses text sentiment analysis technology to divide the emotional polarity of each piece of public opinion information. Specifically, it is divided into three types of emotional polarity: positive, neutral, and negative. When quantifying the ESG risk of an enterprise, only information with negative emotional polarity is considered, and this part of information is included in the enterprise ESG risk assessment. The specific processing process of public opinion information includes the following two steps:

[0087] Step 1: Use text sentiment analysis to determine the sentiment polarity of public opinion information. The enterprise ESG risk assessment system matches multiple pieces of public opinion information for each assessment indicator and uses text sentiment analysis technology to determine the sentiment polarity of each piece of public opinion information.

[0088] Specifically, the enterprise ESG risk assessment system obtains a preset Chinese text classification toolkit and uses it as a training dataset, and all public opinion data under a preset number of assessment indicators as a test dataset. The preset Chinese text classification toolkit may be the THUCTC toolkit (THU Chinese Text Classification, a Chinese text classification toolkit).

[0089] The enterprise ESG risk assessment system then cleans the collected public opinion information, tokenizes the training and test datasets using a violation information lexicon, and removes stop words from the training and test datasets using a Chinese stop word list. The violation information lexicon can be the one constructed during the aforementioned violation information processing. After completing this processing, the enterprise ESG risk assessment system uses word embedding technology to vectorize the processed training and test datasets, generating the first word embedding for the training dataset and the second word embedding for the test dataset.

[0090] Finally, the enterprise ESG risk assessment system uses automated rules for sentiment analysis. Specifically, a model is trained based on the first word embedding and the corresponding sentiment polarity label to generate a sentiment polarity recognition model. The sentiment polarity labels are categorized as positive, neutral, or negative. The second word embedding is then input into the sentiment polarity recognition model to determine the sentiment polarity of each piece of public opinion information, categorized as positive, neutral, or negative.

[0091] Step 2: Count the number of public opinion information with negative sentiment polarity for each enterprise under each evaluation indicator. After completing the sentiment polarity analysis of public opinion information, the public opinion information with negative sentiment polarity will serve as the key to the subsequent quantification of the enterprise's ESG risk. The enterprise ESG risk assessment system will count the number of negative public opinion information for each enterprise under each evaluation indicator. For the sake of convenience, for the i-th enterprise in the automotive supply chain network, m ik It represents the amount of negative public opinion information under the k-th evaluation index.

[0092] The processing of public opinion information is completed through the above two steps. Next, it is necessary to integrate the company's violation information and public opinion information to quantify the company's own ESG risk and calculate the corresponding ESG individual risk value for each company.

[0093] Among them, when calculating the corresponding ESG individual risk value for each enterprise, for the ESG risk brought by the enterprise's violation information, the embodiment of this application will multiply the risk level corresponding to each violation information of the enterprise by the indicator weight and add them together; and for the ESG risk brought by the enterprise's negative public opinion information, the embodiment of this application will use the number of negative public opinion information of the enterprise to represent its risk. Specifically, in the embodiment of this application, the enterprise ESG risk assessment system will query the indicator weight corresponding to each evaluation indicator, and use the following formula 1 to calculate the indicator weight corresponding to each evaluation indicator, the risk group weight corresponding to each risk group, the number of violation information corresponding to each risk group under each evaluation indicator, and the number of public opinion information corresponding to each evaluation indicator to obtain the enterprise's ESG individual risk value,

[0094] Formula 1:

[0095] Among them, R i represents the ESG individual risk value of the i-th enterprise in the supply chain network to be evaluated; the value of i is any positive integer from 1 to N, where N represents the number of enterprises in the supply chain network to be evaluated; W k The kth evaluation indicator weight is represented by k, where the value of k is any positive integer from 1 to a preset number. It is used to indicate the number of violation information included in the high-risk group of the i-th enterprise in the supply chain network to be evaluated under the k-th evaluation indicator; It is used to indicate the number of violation information included in the medium risk group of the i-th enterprise in the supply chain network to be evaluated under the k-th evaluation indicator; It is used to indicate the number of violation information included in the low-risk group of the i-th enterprise in the supply chain network to be evaluated under the k-th evaluation indicator; w1 is used to indicate the risk group weight corresponding to the high-risk group; w2 is used to indicate the risk group weight corresponding to the medium-risk group; w3 is used to indicate the risk group weight corresponding to the low-risk group; mik Used to indicate the amount of public opinion information corresponding to the k-th evaluation indicator.

[0096] Through the above process, the ESG individual risk value of each enterprise can be calculated. After the ESG individual risk value of each enterprise is calculated, the multiple ESG individual risk values obtained can be used to form the enterprise individual risk information of the supply chain network to be evaluated, so that the enterprise individual risk information can be directly used to conduct a comprehensive risk assessment of the enterprise when necessary.

[0097] 204. Calculate the ESG relationship risk value of the target enterprise based on the first ESG individual risk value and the second ESG individual risk value, output the ESG relationship risk value, and execute the following step 206.

[0098] In order to more scientifically and effectively monitor and assess the ESG risks of each enterprise, the calculation should not only consider the ESG risk situation of a single enterprise, but also consider the specific position of the enterprise in the supply chain network to be evaluated and its relationship with upstream and downstream enterprises. Therefore, in the embodiment of this application, based on the quantified enterprise ESG individual risk value and based on the existing supply chain network relationships to be evaluated, the ESG relationship risk value of the target enterprise is calculated. The specific calculation process includes the following two steps:

[0099] Step 1: Calculation of edge weights.

[0100] For automotive companies, each has its own upstream and downstream companies. When ESG-related issues arise in upstream and downstream companies, it is very likely that the company itself will also face ESG-related risks. The ESG risk impact of upstream companies on downstream companies is more obvious. Therefore, in this embodiment of the application, each directed edge in the supply chain network to be evaluated is defined with a certain edge weight. This edge weight represents the degree of ESG risk impact brought by the upstream supplier company to the downstream company.

[0101] According to the ESG individual risk value of each enterprise calculated in the above process, the embodiment of the present application selects the maximum ESG individual risk value of all enterprises’ ESG individual risks, which is defined as R max , that is, the maximum ESG individual risk value is determined by the following formula 2:

[0102] Formula 2: R max =max{R1, R2, ..., R i ,…,R k}

[0103] After finding the maximum ESG individual risk value in the supply chain network to be assessed, the enterprise ESG risk assessment system will use the following formula 3 to calculate the maximum ESG individual risk value and the second ESG individual risk value to obtain the edge weight between the upstream supplier enterprise and the target enterprise:

[0104] Formula 3:

[0105] Among them, v ij It is used to represent the edge weight between enterprise i and enterprise j. Enterprise i is the upstream supplier of enterprise j. The value of i is any positive integer from 1 to M. M is a subset of N. M represents the number of enterprises in the supply chain network to be evaluated whose ESG individual risk values are not equal to 0. N represents the number of enterprises in the supply chain network to be evaluated. The value of i is not equal to the value of j. i Used to represent the ESG individual risk value of enterprise i; R max Used to represent the maximum ESG individual risk value. That is, when calculating the edge weight between the upstream supplier enterprise and the target enterprise, the second ESG individual risk value is brought into R i In the calculation, the v ij The value of is the edge weight of the edge between the upstream supplier enterprise and the target enterprise. It should be noted that from the above formula 3, it can be seen that v ij ∈[1, 2], represents the ESG risk that upstream supplier company i may bring to downstream company j. i =0 when v ij =1, then the upstream supplier company i itself has no ESG risk, so the potential ESG risk it brings to the downstream company j is minimal; and when R i =R max Time ij =2, at this time, the ESG risk of upstream supplier enterprise i itself is the greatest, and therefore the potential ESG risk it brings to downstream enterprise j is the greatest.

[0106] Step 2: Calculation of ESG relationship risk value.

[0107] After calculating the ESG individual risk value for each enterprise through the above process and calculating the edge weight of each directed edge in the supply chain network to be evaluated, the ESG relationship risk value of each enterprise can be calculated. Specifically, for the target enterprise, the enterprise ESG risk assessment system will use the following formula 4 to calculate the edge weight and the first ESG individual risk value to obtain the ESG relationship risk value of the target enterprise:

[0108] Formula 4:

[0109] Among them, RI jIt is used to represent the ESG relationship risk value, that is, the ESG risk level of enterprise j after considering the supply relationship in the supply chain network to be evaluated; v ij It is used to represent the edge weight between enterprise i and enterprise j. Enterprise i is the upstream supplier of enterprise j. The value of i is any positive integer from 1 to M. M is a subset of N. M represents the number of enterprises in the supply chain network to be evaluated whose ESG individual risk values are not equal to 0. N represents the number of enterprises in the supply chain network to be evaluated. The value of i is not equal to the value of j. j It is used to represent the ESG individual risk value of enterprise j. That is, when calculating the ESG relationship risk value of the target enterprise, the first ESG individual risk value of the target enterprise is brought into R j And the calculated edge weights between the target enterprise and the upstream supplier enterprise are brought into v ij In this way, the calculated RI j The value of is the target company's ESG risk score. A higher ESG risk score indicates a higher potential ESG risk. The ESG risk score indicates that a company's upstream suppliers with significant ESG risks will result in the company bearing more potential ESG risks.

[0110] 205. When the query determines that the target enterprise does not have an upstream supplier enterprise, the first ESG individual risk value is used as the ESG relationship risk value of the target enterprise, and the ESG relationship risk value is output, and the following step 206 is executed.

[0111] In the embodiment of the present application, when the query determines that the target enterprise does not have an upstream supplier enterprise, it means that there is no upstream supplier enterprise that can bring potential risks to the target enterprise, and the target enterprise only has its own ESG risks. ij The value of is set to 0. Thus, the ESG relationship risk value of the target enterprise calculated using Formula 4 is also the ESG individual risk value of the target enterprise. Therefore, the enterprise ESG risk assessment system uses the first ESG individual risk value as the ESG relationship risk value of the target enterprise, outputs the ESG relationship risk value, and executes step 206.

[0112] It should be noted that when outputting the ESG relationship risk value, the ESG relationship risk value can be output to the staff of the target enterprise by email, text message, etc. for the reference of the target enterprise.

[0113] 206. Based on the ESG individual risk values of all enterprises included in the supply chain network to be evaluated, determine the target ESG risk level to which the first ESG individual risk value belongs, query the early warning prompt strategy corresponding to the target ESG risk level, and perform early warning operations on the target enterprise according to the early warning prompt strategy.

[0114] In the embodiment of the present application, in order to better monitor the ESG risk level of each enterprise in the supply chain network to be evaluated and to promptly issue early warnings to high-risk enterprises, each enterprise is classified into an ESG risk level based on the ESG individual risk value calculated in the above process, and different early warning prompts are issued for enterprises of different risk levels. Specifically, the ESG risk level classification is shown in Table 4 below:

[0115] Table 4

[0116]

[0117] As can be seen from Table 4 above, each ESG risk level corresponds to a classification ratio. Therefore, for the target enterprise, the classification ratio can be used to evaluate the ESG individual risk values of all enterprises included in the supply chain network to be evaluated, thereby determining the target ESG risk level to which the first ESG individual risk value belongs. Furthermore, in the embodiment of the present application, the enterprise ESG risk assessment system provides different early warning prompt strategies based on the ESG risk level of each enterprise, as shown in Table 5 below:

[0118] Table 5

[0119]

[0120] In this way, through the above Table 5, it is possible to query the early warning strategy corresponding to the target ESG risk level and perform early warning operations on the target enterprise according to the early warning strategy.

[0121] In summary, the process of the enterprise ESG risk assessment method proposed in this application is as follows: Figure 2B ,The method proposed in this application includes four modules, namely the ,index system module, the data mapping module, the risk assessment module and ,the risk warning module.

[0122] Among them, under the indicator system module, an ESG evaluation indicator system including 3 first-level indicators and 12 third-level indicators was designed, and the 12 third-level indicators were used as evaluation indicators for subsequent risk assessment.

[0123] Under the data mapping module, data matching and collection are achieved based on the established automotive supply chain ESG risk warning platform; through natural language processing, the mapping of text data and indicators is completed to increase the comprehensiveness of ESG risk assessment.

[0124] In the risk assessment module, on the one hand, at the enterprise side, violation information and public opinion information are processed separately. Specifically, when processing violation information, the risk level is divided based on the violation information and the indicator weight of each assessment indicator is determined; text analysis is used to classify the risk of each violation information. When processing public opinion information, text sentiment analysis is used to divide the sentiment polarity of public opinion information, and the amount of negative public opinion information for the enterprise under each assessment indicator is used as the key to risk assessment. Furthermore, the enterprise ESG risk is quantified in combination with the indicator system. On the other hand, at the supply chain side, the edge weights of the edges in the supply chain network are used to represent the degree of ESG risk impact brought by upstream and downstream enterprises. Combined with the ESG individual risk value of each enterprise, the ESG relationship risk value of each enterprise is calculated.

[0125] Under the risk warning module, ESG risk levels are divided according to the company's individual ESG risk value, and different warning prompts are given.

[0126] The method provided in the embodiments of the present application can, when conducting ESG risk assessment on an enterprise, simultaneously refer to the enterprise's specific position in the supply chain network to be assessed and its connections with upstream and downstream enterprises, thereby achieving a comprehensive, systematic and dynamic ESG risk assessment, accurately identifying the enterprise's ESG risks and providing effective early warnings, helping enterprises to better fulfill their social responsibilities, enhance their market competitiveness, and provide strong guarantees for their sustainable development.

[0127] Further, as Figure 1 The specific implementation of the method, the embodiment of the present application provides an enterprise ESG risk assessment system, such as Figure 3 As shown, the system includes: a risk assessment module 301.

[0128] The risk assessment module 301 is used to respond to an enterprise risk assessment request and determine a target enterprise to be risk assessed;

[0129] The risk assessment module 301 is further configured to obtain a supply chain network to be assessed and query whether the target enterprise has any upstream supplier enterprises in the supply chain network to be assessed, wherein the target enterprise is in the supply chain network to be assessed, and the supply chain network to be assessed includes multiple enterprises and records the supply relationships between the multiple enterprises;

[0130] The risk assessment module 301 is further configured to determine a first ESG individual risk value corresponding to the target enterprise and a second ESG individual risk value corresponding to the upstream supplier enterprise when the query determines that the target enterprise has an upstream supplier enterprise;

[0131] The risk assessment module 301 is further configured to calculate the ESG relationship risk value of the target enterprise based on the first ESG individual risk value and the second ESG individual risk value, and output the ESG relationship risk value.

[0132] In a specific application scenario, the risk assessment module 301 is used to query whether there is enterprise individual risk information in the supply chain network to be evaluated, and the enterprise individual risk information is used to record the ESG individual risk value corresponding to each enterprise in the supply chain network to be evaluated; when the query determines that the enterprise individual risk information exists in the supply chain network to be evaluated, the ESG individual risk value corresponding to the target enterprise is obtained from the enterprise individual risk information as the first ESG individual risk value, and the ESG individual risk value corresponding to the upstream supplier enterprise is obtained from the enterprise individual risk information as the second ESG individual risk value; when the query determines that the enterprise individual risk information does not exist in the supply chain network to be evaluated, the ESG individual risk value is calculated for each enterprise in the supply chain network to be evaluated, and multiple ESG individual risk values corresponding to the multiple enterprises are obtained, the multiple ESG individual risk values are used as the enterprise individual risk information, and the enterprise individual risk information is stored.

[0133] In a specific application scenario, the system also includes: a data mapping module.

[0134] The data mapping module is used to obtain a preset number of evaluation indicators for each enterprise in the supply chain network to be evaluated, and match multiple violation information of the enterprise under each evaluation indicator; use text parsing technology to process the multiple violation information under each evaluation indicator, determine the risk level corresponding to each violation information, and divide the multiple violation information under each evaluation indicator into high-risk group, medium-risk group and low-risk group according to the risk level corresponding to each violation information, and count the number of violation information included in each risk group; match the multiple public opinion information of the enterprise under each evaluation indicator, use text sentiment analysis technology to determine the sentiment polarity of each public opinion information, and count the number of public opinion information whose sentiment polarity indicates negative public opinion information under each evaluation indicator;

[0135] The risk assessment module 301 is used to query the indicator weight corresponding to each evaluation indicator, and calculate the indicator weight corresponding to each evaluation indicator, the risk group weight corresponding to each risk group, the number of violation information corresponding to each risk group under each evaluation indicator, and the number of public opinion information corresponding to each evaluation indicator using the following formula to obtain the ESG individual risk value of the enterprise.

[0136]

[0137] Among them, R i represents the ESG individual risk value of the i-th enterprise in the supply chain network to be evaluated; the value of i is any positive integer from 1 to N, and N represents the number of enterprises in the supply chain network to be evaluated; W k represents the indicator weight of the kth evaluation indicator, where the value of k is any positive integer from 1 to the preset number; It is used to indicate the number of violation information included in the high-risk group of the i-th enterprise in the supply chain network to be evaluated under the k-th evaluation indicator; It is used to indicate the number of violation information included in the medium risk group of the i-th enterprise in the supply chain network to be evaluated under the k-th evaluation indicator; It is used to indicate the number of violation information included in the low-risk group of the i-th enterprise in the supply chain network to be evaluated under the k-th evaluation indicator; w1 is used to indicate the risk group weight corresponding to the high-risk group; w2 is used to indicate the risk group weight corresponding to the medium-risk group; w3 is used to indicate the risk group weight corresponding to the low-risk group; m ik Used to indicate the amount of public opinion information corresponding to the k-th evaluation indicator.

[0138] In a specific application scenario, the data mapping module is also used to obtain multiple sample data related to ESG, perform text preprocessing operations on the multiple sample data, wherein the text preprocessing operations include one or more of denoising operations, standardized format operations, and word segmentation operations; perform feature extraction on the multiple sample data, and convert the multiple sample data into numerical form; obtain a text category prediction model, which is a language recognition model trained using labeled training data for predicting the category to which the text belongs; input the converted multiple sample data into the text category prediction model for identification to obtain the data category of each sample data; determine the preset number of evaluation indicators, and divide the multiple sample data into the preset number of evaluation indicators according to the mapping relationship between the evaluation indicators and the data categories; count the number of samples of the sample data divided into each evaluation indicator, and calculate the number of samples of each evaluation indicator using the entropy weight method to obtain the indicator weight corresponding to each evaluation indicator.

[0139] In a specific application scenario, the risk assessment module 301 is used to collect violation processing keywords corresponding to high risk level, medium risk level and low risk level respectively, and use the collected violation processing keywords to build a violation information vocabulary; obtain the Jieba word segmentation tool, update the vocabulary of the Jieba word segmentation tool based on the violation information vocabulary, and obtain the improved Jieba word segmentation tool; use the improved Jieba word segmentation tool to perform word segmentation processing on each violation information, and use the violation information vocabulary to perform keyword matching on each violation information after word segmentation processing to determine the target violation processing keyword corresponding to each violation information; determine the corresponding risk level for each violation information according to the risk level of the target violation processing keyword in the violation information vocabulary.

[0140] In a specific application scenario, the risk assessment module 301 is used to obtain a preset Chinese text classification toolkit, use the Chinese text classification toolkit as a training data set, and use all public opinion data under the preset number of evaluation indicators as a test data set; use the violation information vocabulary to tokenize the training data set and the test data set, and use the Chinese stop word list to remove stop words in the training data set and the test data set; use word embedding technology to vectorize the processed training data set and the test data set to obtain the first word embedding of the training data set and the second word embedding of the test data set; perform model training based on the first word embedding and the sentiment polarity label corresponding to the first word embedding to obtain a sentiment polarity recognition model; input the second word embedding into the sentiment polarity recognition model to obtain the sentiment polarity of each public opinion information.

[0141] In a specific application scenario, the risk assessment module 301 is used to query the maximum ESG individual risk value in the supply chain network to be assessed, and calculate the maximum ESG individual risk value and the second ESG individual risk value using the following formula to obtain the edge weight of the edge between the upstream supplier enterprise and the target enterprise:

[0142]

[0143] Among them, v ij It is used to represent the edge weight between enterprise i and enterprise j, where enterprise i is the upstream supplier of enterprise j, the value of i is any positive integer from 1 to M, M is a subset of N, M represents the number of enterprises in the supply chain network to be evaluated whose ESG individual risk values are not equal to 0, N represents the number of enterprises in the supply chain network to be evaluated, and the value of i is not equal to the value of j; R i Used to represent the ESG individual risk value of enterprise i; R maxUsed to represent the maximum ESG individual risk value;

[0144] The following formula is used to calculate the edge weight and the first ESG individual risk value to obtain the ESG relationship risk value of the target enterprise:

[0145]

[0146] Among them, RI j Used to express ESG relationship risk value; v ij It is used to represent the edge weight between enterprise i and enterprise j, where enterprise i is the upstream supplier of enterprise j, the value of i is any positive integer from 1 to M, M is a subset of N, M represents the number of enterprises in the supply chain network to be evaluated whose ESG individual risk values are not equal to 0, N represents the number of enterprises in the supply chain network to be evaluated, and the value of i is not equal to the value of j; R j Used to represent the ESG individual risk value of enterprise j.

[0147] In a specific application scenario, the risk assessment module 301 is also used to use the first ESG individual risk value as the ESG relationship risk value of the target enterprise and output the ESG relationship risk value when the query determines that the target enterprise has no upstream supplier enterprise.

[0148] In a specific application scenario, the system also includes:

[0149] The risk warning module is used to determine the target ESG risk level to which the first ESG individual risk value belongs based on the ESG individual risk values of all enterprises included in the supply chain network to be evaluated; query the warning prompt strategy corresponding to the target ESG risk level, and perform warning operations on the target enterprise according to the warning prompt strategy.

[0150] The system provided in the embodiment of the present application responds to an enterprise risk assessment request, determines the target enterprise to be risk assessed, obtains the supply chain network to be assessed, and inquires whether the target enterprise has upstream supplier enterprises in the supply chain network to be assessed. When the query determines that the target enterprise has upstream supplier enterprises, the system determines the first ESG individual risk value corresponding to the target enterprise and the second ESG individual risk value corresponding to the upstream supplier enterprise. Based on the first ESG individual risk value and the second ESG individual risk value, the ESG relationship risk value of the target enterprise is calculated, and the ESG relationship risk value is output. When conducting ESG risk assessment on the enterprise, the system can simultaneously refer to the specific position of the enterprise in the supply chain network to be assessed and the connection with upstream and downstream enterprises, so as to realize comprehensive, systematic and dynamic ESG risk assessment, accurately identify the ESG risks of the enterprise and conduct effective early warning, which helps the enterprise to better fulfill its social responsibilities, enhance its market competitiveness, and provide strong guarantees for the sustainable development of the enterprise.

[0151] It should be noted that for other corresponding descriptions of the modules involved in the enterprise ESG risk assessment system provided in the embodiment of this application, please refer to Figure 1 and Figures 2A to 2B The corresponding description in will not be repeated here.

[0152] It should be noted that the enterprise information (including but not limited to enterprise equipment information, basic enterprise information, violation information, public opinion information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the enterprise or fully authorized by all parties.

[0153] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0154] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

[0155] In an exemplary embodiment, see Figure 4A computer device is also provided, comprising a bus, a processor, a memory, and a communication interface. The device may also include an input / output interface and a display device, wherein the various functional units can communicate with each other via the bus. The memory stores a computer program, and the processor is configured to execute the program stored in the memory and perform the enterprise ESG risk assessment method in the above-described embodiment.

[0156] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the enterprise ESG risk assessment method.

[0157] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented through hardware or by using software plus the necessary general hardware platform. Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, mobile hard disk, etc.) and includes a number of instructions for enabling a computer device (such as a personal computer, server, or network device) to execute the methods described in various implementation scenarios of the present application.

[0158] Those skilled in the art will understand that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily required to implement the present application.

[0159] Those skilled in the art will appreciate that the modules in the devices in the implementation scenario can be distributed in the devices of the implementation scenario according to the implementation scenario description, or can be modified accordingly and located in one or more devices different from the implementation scenario. The modules in the above implementation scenario can be combined into one module or further split into multiple submodules.

[0160] The above application serial numbers are for description only and do not represent the advantages or disadvantages of the implementation scenarios.

[0161] The above disclosure only describes several specific implementation scenarios of the present application. However, the present application is not limited thereto, and any changes that can be conceived by those skilled in the art should fall within the scope of protection of the present application.

Claims

1. A method for assessing corporate ESG risk, characterized by: include: In response to enterprise risk assessment requests, determine the target enterprises to be risk assessed; Obtaining a supply chain network to be evaluated, and querying whether the target enterprise has an upstream supplier enterprise in the supply chain network to be evaluated, wherein the target enterprise is in the supply chain network to be evaluated, and the supply chain network to be evaluated includes multiple enterprises and records the supply relationships between the multiple enterprises; When the query determines that the target enterprise has an upstream supplier enterprise, determining a first ESG individual risk value corresponding to the target enterprise and determining a second ESG individual risk value corresponding to the upstream supplier enterprise, wherein calculating the ESG individual risk value for each enterprise in the supply chain network to be evaluated includes: for each enterprise in the supply chain network to be evaluated, obtaining a preset number of evaluation indicators, and matching multiple violation information of the enterprise under each evaluation indicator; Processing the plurality of violation information under each evaluation indicator using text parsing technology to determine a risk level corresponding to each violation information, and dividing the plurality of violation information under each evaluation indicator into a high-risk group, a medium-risk group, and a low-risk group according to the risk level corresponding to each violation information, and counting the number of violation information included in each risk group; Match the multiple public opinion information of the enterprise under each of the evaluation indicators, determine the sentiment polarity of each public opinion information by using text sentiment analysis technology, and count the number of public opinion information whose sentiment polarity indicates negative under each evaluation indicator; wherein, determining the sentiment polarity of each public opinion information by using text sentiment analysis technology includes: obtaining a preset Chinese text classification toolkit, using the Chinese text classification toolkit as a training data set, and using all public opinion data under the preset number of evaluation indicators as a test data set; using a violation information vocabulary to tokenize the training data set and the test data set, and using a Chinese stop word list to remove stop words in the training data set and the test data set; using word embedding technology to vectorize the processed training data set and the test data set to obtain a first word embedding of the training data set and a second word embedding of the test data set; performing model training based on the first word embedding and the sentiment polarity label corresponding to the first word embedding to obtain a sentiment polarity recognition model; inputting the second word embedding into the sentiment polarity recognition model to obtain the sentiment polarity of each public opinion information; Query the indicator weight corresponding to each evaluation indicator, and use the following formula to calculate the indicator weight corresponding to each evaluation indicator, the risk group weight corresponding to each risk group, the number of violation information corresponding to each risk group under each evaluation indicator, and the number of public opinion information corresponding to each evaluation indicator to obtain the ESG individual risk value of the enterprise. Among them, R i represents the ESG individual risk value of the i-th enterprise in the supply chain network to be evaluated; the value of i is any positive integer from 1 to N, and N represents the number of enterprises in the supply chain network to be evaluated; W k represents the indicator weight of the kth evaluation indicator, where the value of k is any positive integer from 1 to the preset number; It is used to indicate the number of violation information included in the high-risk group of the i-th enterprise in the supply chain network to be evaluated under the k-th evaluation indicator; It is used to indicate the number of violation information included in the medium risk group of the i-th enterprise in the supply chain network to be evaluated under the k-th evaluation indicator; It is used to indicate the number of violation information included in the low-risk group of the i-th enterprise in the supply chain network to be evaluated under the k-th evaluation indicator; w1 is used to indicate the risk group weight corresponding to the high-risk group; w2 is used to indicate the risk group weight corresponding to the medium-risk group; w3 is used to indicate the risk group weight corresponding to the low-risk group; m ik It is used to indicate the number of public opinion information whose sentiment polarity corresponding to the k-th evaluation index is negative; Based on the first ESG individual risk value and the second ESG individual risk value, the ESG relationship risk value of the target enterprise is calculated, and the ESG relationship risk value is output. When calculating the ESG relationship risk value of the target enterprise, the maximum ESG individual risk value is queried in the supply chain network to be evaluated, and the maximum ESG individual risk value and the second ESG individual risk value are calculated using the following formula to obtain the edge weight of the edge between the upstream supplier enterprise and the target enterprise. Among them, v ij It is used to represent the edge weight between enterprise i and enterprise j, where enterprise i is the upstream supplier of enterprise j, the value of i is any positive integer from 1 to M, M is a subset of N, M represents the number of enterprises in the supply chain network to be evaluated whose ESG individual risk values are not equal to 0, N represents the number of enterprises in the supply chain network to be evaluated, and the value of i is not equal to the value of j; R i Used to represent the ESG individual risk value of enterprise i; R max Used to represent the maximum ESG individual risk value; the following formula is used to calculate the edge weight and the first ESG individual risk value to obtain the ESG relationship risk value of the target enterprise, Among them, RI j Used to express ESG relationship risk value; v ij It is used to represent the edge weight between enterprise i and enterprise j, where enterprise i is the upstream supplier of enterprise j, the value of i is any positive integer from 1 to M, M is a subset of N, M represents the number of enterprises in the supply chain network to be evaluated whose ESG individual risk values are not equal to 0, N represents the number of enterprises in the supply chain network to be evaluated, and the value of i is not equal to the value of j; R j Used to represent the ESG individual risk value of enterprise j.

2. The method according to claim 1, characterized in that Determining the first ESG individual risk value corresponding to the target enterprise and determining the second ESG individual risk value corresponding to the upstream supplier enterprise includes: Query whether there is any enterprise-specific risk information in the supply chain network to be assessed, where the enterprise-specific risk information is used to record the ESG individual risk value corresponding to each enterprise in the supply chain network to be assessed; When the query determines that the supply chain network to be evaluated has the enterprise individual risk information, the ESG individual risk value corresponding to the target enterprise is obtained from the enterprise individual risk information as the first ESG individual risk value, and the ESG individual risk value corresponding to the upstream supplier enterprise is obtained from the enterprise individual risk information as the second ESG individual risk value; When the query determines that the enterprise individual risk information does not exist in the supply chain network to be evaluated, the ESG individual risk value is calculated for each enterprise in the supply chain network to be evaluated, and multiple ESG individual risk values corresponding to the multiple enterprises are obtained. The multiple ESG individual risk values are used as the enterprise individual risk information, and the enterprise individual risk information is stored.

3. The method according to claim 1, characterized in that The method further comprises: Acquire multiple pieces of sample data related to the ESG, and perform text preprocessing operations on the multiple pieces of sample data, wherein the text preprocessing operations include one or more of a noise removal operation, a format standardization operation, and a word segmentation operation; Performing feature extraction on the plurality of sample data and converting the plurality of sample data into numerical form; Obtaining a text category prediction model, where the text category prediction model is a language recognition model trained using labeled training data for predicting the category to which a text belongs; Inputting the converted multiple sample data into the text category prediction model for identification to obtain the data category of each sample data; Determining the preset number of evaluation indicators, and classifying the plurality of sample data into the preset number of evaluation indicators according to a mapping relationship between the evaluation indicators and the data categories; The number of samples of the sample data divided into each evaluation indicator is counted, and the number of samples of each evaluation indicator is calculated using the entropy weight method to obtain the indicator weight corresponding to each evaluation indicator.

4. The method according to claim 1, wherein The text parsing technology is used to process the multiple violation information under each evaluation indicator to determine the risk level corresponding to each violation information, including: Collecting violation handling keywords corresponding to high-risk level, medium-risk level, and low-risk level respectively, and building a violation information vocabulary using the collected violation handling keywords; Obtaining a Jieba word segmentation tool, and updating the word library of the Jieba word segmentation tool based on the violation information word library to obtain an improved Jieba word segmentation tool; Using the improved Jieba word segmentation tool to perform word segmentation processing on each piece of violation information, and using the violation information vocabulary to perform keyword matching on each piece of violation information after word segmentation processing, to determine the target violation processing keyword corresponding to each piece of violation information; According to the risk level of the target violation processing keyword in the violation information vocabulary, a corresponding risk level is determined for each piece of violation information.

5. The method according to claim 1, characterized in that The method further comprises: When the query determines that the target enterprise has no upstream supplier enterprise, the first ESG individual risk value is used as the ESG relationship risk value of the target enterprise, and the ESG relationship risk value is output.

6. The method according to claim 1, wherein The method further comprises: Determining, based on the ESG individual risk values of all enterprises included in the supply chain network to be assessed, a target ESG risk level to which the first ESG individual risk value belongs; Query the early warning prompt strategy corresponding to the target ESG risk level, and perform early warning operations on the target enterprise according to the early warning prompt strategy.

7. An enterprise ESG risk assessment system, characterized by: include: A risk assessment module, configured to respond to enterprise risk assessment requests and determine target enterprises to be risk assessed; The risk assessment module is further configured to obtain a supply chain network to be assessed and query whether the target enterprise has any upstream supplier enterprises in the supply chain network to be assessed, wherein the target enterprise is in the supply chain network to be assessed, and the supply chain network to be assessed includes multiple enterprises and records the supply relationships between the multiple enterprises; The risk assessment module is further configured to, when the query determines that the target enterprise has an upstream supplier enterprise, determine a first ESG individual risk value corresponding to the target enterprise and a second ESG individual risk value corresponding to the upstream supplier enterprise, wherein calculating the ESG individual risk value for each enterprise in the supply chain network to be assessed includes: for each enterprise in the supply chain network to be assessed, obtaining a preset number of assessment indicators, and matching multiple violation information of the enterprise under each assessment indicator; Processing the plurality of violation information under each evaluation indicator using text parsing technology to determine a risk level corresponding to each violation information, and dividing the plurality of violation information under each evaluation indicator into a high-risk group, a medium-risk group, and a low-risk group according to the risk level corresponding to each violation information, and counting the number of violation information included in each risk group; Match the multiple public opinion information of the enterprise under each of the evaluation indicators, determine the sentiment polarity of each public opinion information by using text sentiment analysis technology, and count the number of public opinion information whose sentiment polarity indicates negative under each evaluation indicator; wherein, determining the sentiment polarity of each public opinion information by using text sentiment analysis technology includes: obtaining a preset Chinese text classification toolkit, using the Chinese text classification toolkit as a training data set, and using all public opinion data under the preset number of evaluation indicators as a test data set; using a violation information vocabulary to tokenize the training data set and the test data set, and using a Chinese stop word list to remove stop words in the training data set and the test data set; using word embedding technology to vectorize the processed training data set and the test data set to obtain a first word embedding of the training data set and a second word embedding of the test data set; performing model training based on the first word embedding and the sentiment polarity label corresponding to the first word embedding to obtain a sentiment polarity recognition model; inputting the second word embedding into the sentiment polarity recognition model to obtain the sentiment polarity of each public opinion information; Query the indicator weight corresponding to each evaluation indicator, and use the following formula to calculate the indicator weight corresponding to each evaluation indicator, the risk group weight corresponding to each risk group, the number of violation information corresponding to each risk group under each evaluation indicator, and the number of public opinion information corresponding to each evaluation indicator to obtain the ESG individual risk value of the enterprise. Among them, R i represents the ESG individual risk value of the i-th enterprise in the supply chain network to be evaluated; the value of i is any positive integer from 1 to N, and N represents the number of enterprises in the supply chain network to be evaluated; W k represents the indicator weight of the kth evaluation indicator, where the value of k is any positive integer from 1 to the preset number; It is used to indicate the number of violation information included in the high-risk group of the i-th enterprise in the supply chain network to be evaluated under the k-th evaluation indicator; It is used to indicate the number of violation information included in the medium risk group of the i-th enterprise in the supply chain network to be evaluated under the k-th evaluation indicator; It is used to indicate the number of violation information included in the low-risk group of the i-th enterprise in the supply chain network to be evaluated under the k-th evaluation indicator; w1 is used to indicate the risk group weight corresponding to the high-risk group; w2 is used to indicate the risk group weight corresponding to the medium-risk group; w3 is used to indicate the risk group weight corresponding to the low-risk group; m ik It is used to indicate the number of public opinion information whose sentiment polarity corresponding to the k-th evaluation index is negative; The risk assessment module is further configured to calculate the ESG relationship risk value of the target enterprise based on the first ESG individual risk value and the second ESG individual risk value, and output the ESG relationship risk value. When calculating the ESG relationship risk value of the target enterprise, the maximum ESG individual risk value is queried in the supply chain network to be assessed, and the maximum ESG individual risk value and the second ESG individual risk value are calculated using the following formula to obtain the edge weight of the edge between the upstream supplier enterprise and the target enterprise. Among them, v ij It is used to represent the edge weight between enterprise i and enterprise j, where enterprise i is the upstream supplier of enterprise j, the value of i is any positive integer from 1 to M, M is a subset of N, M represents the number of enterprises in the supply chain network to be evaluated whose ESG individual risk values are not equal to 0, N represents the number of enterprises in the supply chain network to be evaluated, and the value of i is not equal to the value of j; R i Used to represent the ESG individual risk value of enterprise i; R max Used to represent the maximum ESG individual risk value; the following formula is used to calculate the edge weight and the first ESG individual risk value to obtain the ESG relationship risk value of the target enterprise, Among them, RI j Used to express ESG relationship risk value; v ij It is used to represent the edge weight between enterprise i and enterprise j, where enterprise i is the upstream supplier of enterprise j, the value of i is any positive integer from 1 to M, M is a subset of N, M represents the number of enterprises in the supply chain network to be evaluated whose ESG individual risk values are not equal to 0, N represents the number of enterprises in the supply chain network to be evaluated, and the value of i is not equal to the value of j; R j Used to represent the ESG individual risk value of enterprise j.

Images