An esg performance optimization method and system based on artificial intelligence

Abstract

The application relates to the technical field of enterprise management, and discloses an ESG performance optimization method and system based on artificial intelligence; the ESG performance optimization method comprises the following steps: acquiring evaluation indexes for evaluating the ESG performance of a target enterprise; acquiring data requirements of the evaluation indexes, and obtaining data types and data sources of enterprise data used for calculating evaluation index scores based on the data requirements; corresponding data collection nodes are respectively generated based on different enterprise data, and any data collection node is connected with a data source storing corresponding enterprise data; data collection node networks are integrated by using data collection nodes corresponding to different data types, and real-time enterprise data is synchronously acquired by using the data collection node networks; real-time scores of the evaluation indexes are calculated by using the real-time enterprise data, and real-time ESG performances of the target enterprise are evaluated by using the real-time scores. The application can provide real-time ESG performance evaluation feedback for enterprises, and is helpful to improving the sustainable development capability and social responsibility fulfillment effect of the enterprises.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of enterprise management, and in particular to an ESG performance optimization method and system based on artificial intelligence. BACKGROUND

[0002] With the global emphasis on sustainable development and social responsibility, ESG performance has become an important standard for measuring the long-term competitiveness and social responsibility fulfillment ability of enterprises. However, when conducting ESG assessment, enterprises face major challenges due to scattered data and lack of unified management mechanism.

[0003] Due to the use of different systems and data formats by different departments, traditional data management methods are difficult to effectively deal with a large amount of information from multiple heterogeneous data sources, thereby reducing the efficiency of data integration. At the same time, the lack of centralized management mechanism further exacerbates the severity of the problem, and some key ESG data may not be systematically recorded in some departments, or the records are incomplete, resulting in incomplete assessment results, affecting the accurate understanding of the enterprise's ESG performance. In addition, each department may update data at different time points, resulting in inconsistent data timeliness, further exacerbating the incompleteness problem. Moreover, each department independently manages its own data, which hinders the sharing and circulation of data, making it difficult for the enterprise to form a global perspective to assess its overall ESG performance. SUMMARY

[0004] In view of the deficiencies of the prior art and actual application requirements, the present application provides an ESG performance optimization method and system based on artificial intelligence, aiming to solve the problems of low data integration efficiency and insufficient assessment accuracy caused by scattered data, lack of unified management, inconsistent data timeliness and incomplete assessment results in the process of enterprise ESG assessment.

[0005] In a first aspect, the ESG performance optimization method based on artificial intelligence provided by the present application comprises the following steps:

[0006] Obtaining evaluation indicators for evaluating the ESG performance of the target enterprise, the evaluation indicators including environmental evaluation indicators, social evaluation indicators and management evaluation indicators;

[0007] Obtaining the data requirements of the evaluation indicators, and based on the data requirements, obtaining the data types and data sources of the enterprise data used to calculate the evaluation indicator scores;

[0008] Based on different enterprise data, corresponding data collection nodes are generated respectively, and any data collection node is connected with a data source storing corresponding enterprise data;

[0009] The data acquisition nodes corresponding to different data types are integrated into a data acquisition node network, and the real-time enterprise data used for calculating the evaluation index score is synchronously acquired by using the data acquisition node network.

[0010] The real-time score of the evaluation index is calculated by using the real-time enterprise data, and the real-time ESG performance of the target enterprise is evaluated by using the real-time score.

[0011] Optionally, the evaluation index used for evaluating the ESG performance of the target enterprise comprises the following steps:

[0012] The policy file for evaluating the ESG performance of the target enterprise is acquired.

[0013] The evaluation index is extracted from the policy file by using a natural language processing technology.

[0014] Optionally, the environmental evaluation index comprises, but is not limited to, carbon emission, energy consumption, water resource management, waste treatment, pollutant emission and ecological protection.

[0015] The social evaluation index comprises, but is not limited to, employee health and safety, employee diversity and inclusiveness, supply chain management, community investment and labor conditions.

[0016] The management evaluation index comprises, but is not limited to, board structure, anti-corruption and compliance, shareholder equity protection, information transparency and enterprise risk management.

[0017] Optionally, the data requirement of the evaluation index comprises the following steps:

[0018] The score calculation model of the evaluation index is acquired, and the definition of each calculation parameter in the score calculation model is acquired.

[0019] According to the definition of each calculation parameter in the score calculation model, the data type and data source of the enterprise data used for calculating the evaluation index score are obtained.

[0020] Optionally, the corresponding data acquisition nodes are respectively generated based on different enterprise data, and the method further comprises the following steps:

[0021] According to the data type of the enterprise data, the data acquisition mode of the corresponding data acquisition node is matched, and the matched data acquisition mode is loaded into the corresponding data acquisition node, wherein the data acquisition mode is a mode of acquiring enterprise data from a corresponding data source.

[0022] Optionally, the data acquisition node network is integrated by using the data acquisition nodes corresponding to different data types, and the method comprises the following steps:

[0023] From all data acquisition nodes, select one data acquisition node as a center data acquisition node;

[0024] Connect the remaining data acquisition nodes with the center data acquisition node respectively to obtain a data acquisition node network, and the data acquisition node network controls the remaining data acquisition nodes by using the center data acquisition node.

[0025] Optionally, the step of selecting one data acquisition node as a center data acquisition node from all data acquisition nodes comprises the following steps:

[0026] The average data transmission bandwidth of each data acquisition node and the remaining data acquisition nodes is obtained respectively;

[0027] The data acquisition node corresponding to the maximum average data transmission bandwidth is set as the center data acquisition node.

[0028] Optionally, the average data transmission bandwidth of any data acquisition node meets a calculation model.

[0029] Optionally, the step of synchronously obtaining real-time enterprise data used for calculating the evaluation index score by using the data acquisition node network comprises the following steps:

[0030] The center data acquisition node sends real-time enterprise data requests to the remaining data acquisition nodes at the same time node, and any data acquisition node in the remaining data acquisition nodes sends real-time enterprise data to the center data acquisition node based on the corresponding real-time enterprise data request.

[0031] The center data acquisition node integrates the real-time enterprise data from all data acquisition nodes.

[0032] In a second aspect, the ESG performance optimization system based on artificial intelligence provided by the application comprises an input device, a processor, a memory and an output device, which are connected with each other, wherein the memory is used for storing a computer program, the computer program comprises program instructions, and the processor is configured to call the program instructions to execute the ESG performance optimization method based on artificial intelligence provided in the first aspect.

[0033] The ESG performance optimization method and system based on artificial intelligence provided by the application have the following advantages:

[0034] 1. The automatic data acquisition node network solves the problems of scattered enterprise internal data and inconsistent formats, so that ESG data from different departments can be efficiently and centrally integrated and processed.

[0035] 2. Utilize the data collection node network to synchronize the ESG data of the enterprise in real time, ensure the timeliness and consistency of the data during evaluation, and avoid evaluation deviation caused by different departments due to delayed data updates.

[0036] 3. Through real-time data collection and score calculation, the ESG performance of the enterprise can be dynamically evaluated, providing instant feedback to the enterprise to help it quickly adjust its strategy and optimize its sustainable development and social responsibility performance.

[0037] 4. According to the policy documents of the industry in which the enterprise is located, combined with artificial intelligence to automatically extract evaluation indicators, and combined with real-time data to dynamically calculate ESG scores, helping the enterprise to understand its performance and make timely adjustments. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 A flowchart of the ESG performance optimization method based on artificial intelligence provided by the embodiments of the present application;

[0039] Figure 2 A schematic diagram of a data collection node provided by the embodiments of the present application;

[0040] Figure 3 A schematic diagram of a data collection node network provided by the embodiments of the present application;

[0041] Figure 4 A schematic diagram of the ESG performance optimization system based on artificial intelligence provided by the embodiments of the present application. DETAILED DESCRIPTION

[0042] The specific embodiments of the present application will be described in detail below. It should be noted that the embodiments described herein are only for illustration and do not limit the present application. In the following description, a large number of specific details are set forth in order to provide a thorough understanding of the present application. However, it is obvious to those skilled in the art that the present application does not necessarily have to be implemented with these specific details. In other examples, well-known circuits, software or methods are not specifically described in order not to obscure the present application.

[0043] In one embodiment, please refer to Figure 1 , Figure 1 A flowchart of the ESG performance optimization method based on artificial intelligence provided by the embodiments of the present application.

[0044] As Figure 1 shown, the ESG performance optimization method based on artificial intelligence provided by the embodiments of the present application includes the following steps:

[0045] S01, obtaining evaluation indicators for evaluating the ESG performance of the target enterprise, the evaluation indicators including environmental evaluation indicators, social evaluation indicators and management evaluation indicators.

[0046] In this embodiment, the target enterprise ESG performance refers to the actual actions and results of the target enterprise in the three dimensions of environment (Environmental), society (Social) and governance (Governance), representing its overall performance in sustainable development and social responsibility fulfillment.

[0047] It can be understood that the focus and requirements of enterprise ESG performance evaluation in different industries are different, such as energy enterprises pay more attention to environmental evaluation indicators such as carbon emissions and energy management, and technology enterprises pay more attention to management evaluation indicators such as data privacy and supply chain management.

[0048] Further, the evaluation indicators for evaluating the ESG performance of the target enterprise in step S01 include the following steps:

[0049] S011, obtaining the policy file of ESG performance evaluation of the industry where the target enterprise is located.

[0050] It can be known that the ESG evaluation system needs to be based on relevant policies, regulations and industry standards, which provide specific guidelines and requirements for ESG evaluation of different types of enterprises.

[0051] In the prior art, there are special ESG policy files or guidelines for different industries and regions, such as Global Reporting Initiative (GRI), Sustainability Accounting Standards Board (SASB) standards, industry-specific policy files, etc.

[0052] S012, using natural language processing technology to extract evaluation indicators from the policy file.

[0053] Specifically, one or more AI models such as ChatGPT, Wenxin Yiyang, etc. can be used to extract evaluation indicators for evaluating ESG performance from one or more policy files.

[0054] Further, the environmental evaluation indicators refer to the impact of enterprises on the natural environment during operation and related environmental protection measures, which are used to measure the environmental sustainability of enterprises and evaluate the performance of enterprises in resource use, pollution control and ecological protection, including but not limited to carbon emissions, energy consumption, water resource management, waste treatment, pollutant emissions and ecological protection.

[0055] Further, the social evaluation indicators mainly measure the performance of enterprises in social responsibility, employee welfare, community relations, etc., including but not limited to employee health and safety, employee diversity and inclusiveness, supply chain management, community investment and labor conditions.

[0056] Further, the management class evaluation indicators measure the performance of the enterprise in corporate governance, internal control and management transparency, including but not limited to board structure, anti-corruption and compliance, shareholder rights protection, information transparency and enterprise risk management.

[0057] S02, obtaining the data requirements of the evaluation indicators, and based on the data requirements, obtaining the data types and data sources of the enterprise data used to calculate the evaluation indicator scores.

[0058] It can be understood that step S02 obtains the data requirements of the evaluation indicators to ensure that the necessary data can be collected to calculate the performance scores of the enterprise in the three dimensions of environment, society and governance (ESG).

[0059] Further, step S02 obtains the data requirements of the evaluation indicators, including the following steps:

[0060] S021, obtaining the score calculation model of the evaluation indicators, and obtaining the definition of each calculation parameter in the score calculation model.

[0061] Similarly, step S021 can refer to the implementation of step S012 using natural language processing technology, that is, importing the relevant policies of ESG performance evaluation indicators into the AI model, and using the AI model to obtain one or more score calculation models of the evaluation indicators, and the definition of each calculation parameter in each score calculation model.

[0062] S022, obtaining the data types and data sources of the enterprise data used to calculate the evaluation indicator scores according to the definition of each calculation parameter in the score calculation model.

[0063] According to the GRI 305 standard, the score calculation model of the carbon emission management of the enterprise is , wherein, represents the carbon emission management score of the enterprise, represents the current carbon emission of the enterprise, with the unit of tons of carbon dioxide equivalent, represents the industry or historical benchmark emission, with the unit of tons of carbon dioxide equivalent.

[0064] Further, the enterprise data used to calculate the evaluation indicator scores respectively include the current carbon emission of the enterprise and the industry or historical benchmark emission, both of which are numerical data, the data source of the former is the enterprise environmental management system (EMS), and the data source of the latter is the industry database or the enterprise environmental management system (EMS).

[0065] S03, based on different enterprise data, respectively generating corresponding data collection nodes, and any data collection node is connected with the data source storing the corresponding enterprise data.

[0066] In the embodiment, in order to solve the problem of data dispersion of the ESG evaluation process, corresponding data collection nodes are set for different enterprise data, and the corresponding data collection nodes are connected with the data sources storing the corresponding enterprise data, so that the dispersed data can be collected.

[0067] Please refer to Figure 2 , Figure 2 the schematic diagram of the data collection node provided by the embodiment of the present application; as shown in Figure 2 , in order to evaluate the ESG performance of the target enterprise, enterprise data 1, enterprise data 2, …, enterprise data N-1 and enterprise data N and other enterprise data need to be collected, in order to ensure that each kind of enterprise data can be collected, based on enterprise data 1, enterprise data 2, …, enterprise data N-1 and enterprise data N, corresponding data collection nodes 1, data collection nodes 2, …, data collection nodes N-1 and data collection nodes N are set.

[0068] It can be understood that different enterprise data may come from different data sources, such as Figure 2 enterprise data 1 comes from data source 1 and enterprise data N comes from data source M; different enterprise data may come from the same data source, such as Figure 2 enterprise data 1 and enterprise data 2 come from data source 1.

[0069] Further, in order to solve the problem of data heterogeneity of the ESG evaluation process, since different enterprise data is obtained by corresponding data collection nodes respectively, each data collection node only needs to be responsible for the collection, transmission or processing of one type of data.

[0070] In the embodiment, the step S03 of generating corresponding data collection nodes based on different enterprise data further includes the following steps:

[0071] According to the data type of the enterprise data, the data collection mode of the corresponding data collection node is matched, and the matched data collection mode is loaded into the corresponding data collection node, and the data collection mode is a way of obtaining enterprise data from a corresponding data source.

[0072] Specifically, the data type includes but is not limited to numerical data, text data, document data and other conventional data types; further, each data type corresponds to a data collection mode, such as numerical data collected through API interface connection, text data collected through NLP technology, and document data collected through document analysis tools (such as OCR or PDF analysis tools).

[0073] As shown in Figure 2As shown, the data types of the enterprise data 1, the enterprise data 2, …, the enterprise data N-1 and the enterprise data N are matched with the corresponding data collection modes, and then the corresponding data collection modes are loaded into the data collection node 1, the data collection node 2, …, the data collection node N-1 and the data collection node N respectively. It can be understood that the data collection modes of different data collection nodes can be the same, such as the data collection node 1 and the data collection node 2 are both the data collection mode 1.

[0074] S04, using the data collection nodes corresponding to different data types, integrating the data collection node network, and using the data collection node network to synchronously acquire real-time enterprise data for calculating the evaluation index score.

[0075] In this embodiment, for the data dispersion problem of the ESG evaluation process, the integration of the data collection node network realizes the automation and centralized management of data collection.

[0076] Further, the step S04 of using the data collection nodes corresponding to different data types to integrate the data collection node network comprises the following steps:

[0077] S0411, selecting one data collection node from all data collection nodes as a central data collection node.

[0078] Further, the selection method of the central data collection node can be random selection, or the data collection node with the strongest data processing capability can be selected as the central data collection node according to the data processing capability of the data collection node, or the selection can be based on other methods.

[0079] In this embodiment, the step S0411 of selecting one data collection node from all data collection nodes as a central data collection node comprises the following steps:

[0080] S04111, acquiring the average data transmission bandwidth of each data collection node and the remaining data collection nodes respectively.

[0081] Specifically, the average data transmission bandwidth of any data collection node satisfies the following calculation model: wherein i is a positive integer from 1 to N, N is the total number of data collection nodes, i represents the number of data collection nodes, represents the average data transmission bandwidth of the data collection node i, represents the data transmission bandwidth between the data collection node i and the data collection node j.

[0082] S04112, setting the data collection node corresponding to the maximum average data transmission bandwidth as the central data collection node.

[0083] It can be understood that by selecting the node with the largest average data transmission bandwidth as the central node, the performance of the entire data collection network and the speed of data synchronization can be maximized, thereby ensuring higher efficiency in ESG data collection and processing.

[0084] S0412. Connect the remaining data acquisition nodes to the central data acquisition node respectively to obtain a data acquisition node network. The data acquisition node network uses the central data acquisition node to control the remaining data acquisition nodes.

[0085] See Figure 3 , Figure 3 A schematic diagram of a data acquisition node network provided by an embodiment of the present invention; Figure 3 As shown, data acquisition node 1 is set as the central data acquisition node O, and the central data acquisition node O is connected to data acquisition node 1, data acquisition node 2, ..., data acquisition node N-1 and data acquisition node N respectively, and then the central data acquisition node is used to control the remaining data acquisition nodes.

[0086] In this embodiment, in order to ensure the timeliness and uniformity of data in the ESG evaluation process, the problem of inconsistent data updates is solved by collecting all data at a unified time point, ensuring that data of all dimensions are collected within the same time period, thereby improving the accuracy and timeliness of the evaluation.

[0087] Furthermore, the step S04 of synchronously acquiring the real-time enterprise data used to calculate the evaluation index score by utilizing the data collection node network includes the following steps:

[0088] S0421. Utilize the central data collection node to send real-time enterprise data requests to the remaining data collection nodes at the same time node. Any of the remaining data collection nodes sends real-time enterprise data to the central data collection node based on the corresponding real-time enterprise data request.

[0089] In this embodiment, the central data collection node sends data requests to all other data collection nodes at the same time point, ensuring that all distributed data collection nodes can collect and return the latest enterprise data within the same period, avoiding data asynchrony problems caused by data collection times from different sources.

[0090] S0422. Utilize the central data collection node to integrate real-time enterprise data from all data collection nodes.

[0091] It can be understood that the central data collection node is responsible for corresponding real-time enterprise data collection, and is also responsible for summarizing and integrating data from all other nodes, ensuring that data sources are diverse and comprehensive, and that unified processing can be performed at the central node, further enhancing data consistency and availability, so that data of all dimensions can be analyzed within the same time frame.

[0092] S05, using the real-time enterprise data, calculating the real-time score of the evaluation index, and using the real-time score to evaluate the real-time ESG performance of the target enterprise.

[0093] In this embodiment, step S05 calculates the scores of the evaluation indexes of each dimension of ESG using real-time enterprise data, and dynamically evaluates the overall ESG performance of the enterprise based on these scores, thereby providing timely feedback to the target enterprise and helping it to discover potential problems and quickly adjust its operation strategy.

[0094] In one embodiment, please refer to Figure 4 , Figure 4 The ESG performance optimization system based on artificial intelligence provided by the embodiment of the present application.

[0095] As Figure 4 shown, the ESG performance optimization system based on artificial intelligence provided by the present application comprises an input device, a processor, a memory and an output device; wherein the input device, the processor, the memory and the output device are connected to each other.

[0096] Further, the memory is used to store a computer program, the computer program comprising program instructions, and the processor is configured to invoke the program instructions to execute the ESG performance optimization method based on artificial intelligence.

[0097] Further, the input device is used to input data and instructions related to enterprise ESG performance, allowing users to input content such as policy documents, access permissions of enterprise data sources, etc. Specifically, the input device can be a keyboard, a touch screen, a scanning device, etc., for manual input, or can be integrated with an API interface for automatic data acquisition from external systems (such as enterprise management systems, government databases).

[0098] Further, the output device is used to display and output the ESG performance evaluation results, providing users with real-time ESG performance scores, detailed evaluation reports, and feedback on system operation status. Specifically, the output device can be a display, a printer, a data interface, etc., through which users can view the evaluation results or export the results in the form of a report (such as PDF or spreadsheet). The output device can also feed back the evaluation results to the enterprise management system or regulatory agency through the API interface.

[0099] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than limiting them; although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the above embodiments, or make equivalent replacements for part or all of the technical features.

Claims

1. An ESG performance optimization method based on artificial intelligence, characterized in that: The steps include: Obtaining evaluation indicators for evaluating the ESG performance of target companies, including environmental, social, and management evaluation indicators; Obtaining data requirements for the evaluation indicators, and based on the data requirements, obtaining data types and data sources for enterprise data used to calculate the evaluation indicator scores; Based on different enterprise data, corresponding data collection nodes are generated respectively, and any data collection node is connected to the data source storing the corresponding enterprise data; Using data acquisition nodes corresponding to different data types, a data acquisition node network is integrated, and the data acquisition node network is used to synchronously acquire real-time enterprise data used to calculate the evaluation index score. By integrating the data acquisition node network, automation and centralized management of data acquisition are achieved. From all data acquisition nodes, one data acquisition node is selected as a central data acquisition node, and the remaining data acquisition nodes are respectively connected to the central data acquisition node to obtain a data acquisition node network. The data acquisition node network uses the central data acquisition node to control the remaining data acquisition nodes. The real-time enterprise data is used to calculate the real-time score of the evaluation indicator, and the real-time score is used to evaluate the real-time ESG performance of the target enterprise.

2. The ESG performance optimization method based on artificial intelligence according to claim 1, characterized in that: The step of obtaining the evaluation indicators for evaluating the ESG performance of the target enterprise includes the following steps: Obtain policy documents on ESG performance assessment for the target company’s industry; Natural language processing technology is used to extract evaluation indicators from the policy documents.

3. The ESG performance optimization method based on artificial intelligence according to claim 2, characterized in that: The environmental assessment indicators include carbon emissions, energy consumption, water resource management, waste disposal, pollutant emissions and ecological protection; The social evaluation indicators include employee health and safety, employee diversity and inclusion, supply chain management, community investment, and labor conditions; The management assessment indicators include board structure, anti-corruption and compliance, shareholder rights protection, information transparency and enterprise risk management.

4. The ESG performance optimization method based on artificial intelligence according to claim 1, characterized in that: The step of obtaining the data required for the evaluation indicator comprises the following steps: Obtaining a score calculation model for the evaluation indicator and obtaining definitions of various calculation parameters in the score calculation model; According to the definition of each calculation parameter in the score calculation model, the data type and data source of the enterprise data used to calculate the evaluation index score are obtained.

5. The artificial intelligence-based ESG performance optimization method according to claim 4, characterized in that: The step of generating corresponding data collection nodes based on different enterprise data also includes the following steps: According to the data type of the enterprise data, the data collection mode of the corresponding data collection node is matched, and the matched data collection mode is loaded into the corresponding data collection node. The data collection mode is a method of obtaining enterprise data from the corresponding data source.

6. The ESG performance optimization method based on artificial intelligence according to claim 5, characterized in that: The step of selecting a data acquisition node from all data acquisition nodes as a central data acquisition node comprises the following steps: Obtain the average data transmission bandwidth of each data acquisition node and the remaining data acquisition nodes respectively; The data collection node corresponding to the maximum average data transmission bandwidth is set as the central data collection node.

7. The artificial intelligence-based ESG performance optimization method according to claim 6, characterized in that: The average data transmission bandwidth of any data acquisition node satisfies the calculation model.

8. The artificial intelligence-based ESG performance optimization method according to claim 5, characterized in that: The method of synchronously acquiring real-time enterprise data for calculating the evaluation index score by utilizing the data acquisition node network comprises the following steps: Using the central data collection node, at the same time node, real-time enterprise data requests are sent to the remaining data collection nodes respectively, and any of the remaining data collection nodes sends real-time enterprise data to the central data collection node based on the corresponding real-time enterprise data request; The central data collection node is utilized to integrate real-time enterprise data from all data collection nodes.

9. An artificial intelligence-based ESG performance optimization system, characterized by: It includes an input device, a processor, a memory and an output device, wherein the input device, the processor, the memory and the output device are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute the artificial intelligence-based ESG performance optimization method as described in any one of claims 1 to 8.

Images