CN113240272A

CN113240272A - Enterprise ESG index determination method and related product

Info

Publication number: CN113240272A
Application number: CN202110519905.XA
Authority: CN
Inventors: 诸世卓; 周星浩; 陈寅杰
Original assignee: Ping An Technology Shenzhen Co Ltd
Current assignee: Ping An Technology Shenzhen Co Ltd
Priority date: 2021-05-12
Filing date: 2021-05-12
Publication date: 2021-08-10
Anticipated expiration: 2041-05-12
Also published as: CN113240272B

Abstract

The embodiment of the application discloses an enterprise ESG index determining method and a related product. The method comprises the following steps: acquiring an evaluation index set; acquiring historical ESG evaluation data of a plurality of enterprises and historical profitability of the plurality of enterprises; taking historical ESG evaluation data of a plurality of enterprises as training samples and historical profitability of the plurality of enterprises as training labels to perform model training to obtain the actual weight of each evaluation index under each evaluation dimension; determining an initial weight of each evaluation index under each evaluation dimension according to historical ESG evaluation data of a plurality of enterprises; fusing the actual weight and the initial weight of each evaluation index under each evaluation dimension to obtain the target weight of each evaluation index under each evaluation dimension; and obtaining the ESG index of the enterprise to be evaluated according to the enterprise data of the enterprise to be evaluated and the target weight of each evaluation index under each evaluation dimension.

Description

Enterprise ESG index determination method and related product

Technical Field

The application relates to the technical field of information processing, in particular to an enterprise ESG index determining method and a related product.

Background

The ESG evaluation of an enterprise is the comprehensive evaluation of the environment (E) of the enterprise, the Society (S) and the Governance (G). Some successful experience has been accumulated internationally and domestically in evaluating the ESG performance of enterprises, and internationally known rating agencies such as MSCI, FTSE, etc. have established respective scoring standards and have performed ESG scoring on internationally known enterprises. With the importance of various investment institutions and governments on enterprise responsibility in international society, particularly with the recent progress of global climate change cooperation, and the promise of carbon peak reaching in 2030 and carbon neutralization in 2060 made internationally by China, domestic investors and governments also have a great demand on the grading of enterprises in ESG.

Currently, the ESG scoring of enterprises needs to combine the grading scoring of enterprises in multiple dimensions into an overall scoring in a weighted manner, however, the weight of each scoring dimension is set manually in the weighting process at present, so that the setting subjectivity of the weight is relatively high, and the accuracy of the finally calculated overall ESG scoring is relatively low, so that the accuracy of decisions made based on the ESG scoring is relatively low.

Disclosure of Invention

The embodiment of the application provides an enterprise ESG index determination method and a related product, wherein the weight of each evaluation index is determined through historical ESG evaluation data of an enterprise, so that the calculated total ESG score has higher precision, and the decision making precision is improved.

In a first aspect, an embodiment of the present application provides a method for determining an enterprise ESG index, including:

acquiring an evaluation index set, wherein the evaluation index set comprises evaluation indexes in a plurality of evaluation dimensions, and each evaluation dimension comprises at least one evaluation index;

acquiring historical ESG evaluation data of a plurality of enterprises and historical profitability of the plurality of enterprises;

performing model training by using the historical ESG evaluation data of the plurality of enterprises as training samples and using the historical profitability of the plurality of enterprises as training labels to obtain the actual weight of each evaluation index under each evaluation dimension;

determining an initial weight of each evaluation index under each evaluation dimension according to historical ESG evaluation data of the plurality of enterprises;

fusing the actual weight and the initial weight of each evaluation index under each evaluation dimension to obtain the target weight of each evaluation index under each evaluation dimension;

according to enterprise data of an enterprise to be evaluated and the target weight of each evaluation index under each evaluation dimension, obtaining the ESG index of the enterprise to be evaluated;

and sending the ESG index of the enterprise to be evaluated to target equipment, so that a user of the target equipment can make a decision related to the enterprise to be evaluated according to the ESG index of the enterprise to be evaluated.

In a second aspect, an embodiment of the present application provides an apparatus for determining an enterprise ESG index, including:

the device comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring an evaluation index set, the evaluation index set comprises evaluation indexes in a plurality of evaluation dimensions, and each evaluation dimension comprises at least one evaluation index; acquiring historical ESG evaluation data of a plurality of enterprises under each evaluation dimension and historical profitability of the plurality of enterprises;

the processing unit is used for acquiring historical ESG evaluation data of a plurality of enterprises and historical profitability of the plurality of enterprises;

and the sending unit is used for sending the ESG index of the enterprise to be evaluated to the target equipment so that a user of the target equipment can make a decision related to the enterprise to be evaluated according to the ESG index of the enterprise to be evaluated.

In a third aspect, an embodiment of the present application provides an electronic device, including: a processor coupled to a memory, the memory configured to store a computer program, the processor configured to execute the computer program stored in the memory to cause the electronic device to perform the method of the first aspect.

In a fourth aspect, embodiments of the present application provide a computer-readable storage medium, which stores a computer program, where the computer program makes a computer execute the method according to the first aspect.

In a fifth aspect, embodiments of the present application provide a computer program product comprising a non-transitory computer-readable storage medium storing a computer program, the computer being operable to cause a computer to perform the method according to the first aspect.

The embodiment of the application has the following beneficial effects:

it can be seen that, in the embodiment of the application, based on historical ESG evaluation data of multiple enterprises, a target weight of each evaluation index in each evaluation dimension is determined, instead of manually setting the target weight, so that the obtained target weight can be supported by the historical data, the setting of the target weight is more accurate, the obtained ESG index is more accurate, and the precision of the made decision is improved.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a schematic flowchart of an enterprise ESG index determining method according to an embodiment of the present application;

fig. 2 is a schematic flowchart of a data completion method according to an embodiment of the present application;

fig. 3 is a schematic flowchart of an enterprise ESG index determining method according to an embodiment of the present application;

fig. 4 is a block diagram illustrating functional units of an enterprise ESG index determining apparatus according to an embodiment of the present disclosure;

fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some, but not all, embodiments of the present application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

The terms "first," "second," "third," and "fourth," etc. in the description and claims of this application and in the accompanying drawings are used for distinguishing between different objects and not for describing a particular order. Furthermore, the terms "include" and "have," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those steps or elements listed, but may alternatively include other steps or elements not listed, or inherent to such process, method, article, or apparatus.

Reference herein to "an embodiment" means that a particular feature, result, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein can be combined with other embodiments.

First, it is explained that the enterprise ESG index or ESG index, i.e. the enterprise ESG score, may also be referred to as ESG score, which are the same in nature and will not be distinguished later. Accordingly, the ESG evaluation of a business or the ESG scoring of a business, which are all essentially the same, are used to determine the ESG index of a business.

Referring to fig. 1, fig. 1 is a schematic flowchart of a method for determining an enterprise ESG index according to an embodiment of the present disclosure. The method is applied to the enterprise ESG index determining device. The method includes but is not limited to the following steps:

101: acquiring an evaluation index set, wherein the evaluation index set comprises evaluation indexes in a plurality of evaluation dimensions, and each evaluation dimension comprises at least one evaluation index.

Optionally, an evaluation index configuration table may be obtained, and a field in the evaluation index configuration table is identified to obtain an evaluation index set. For example, if the evaluation index configuration table is pre-entered with the evaluation indexes in each evaluation dimension, the evaluation index configuration table may be read from the storage space of the evaluation index configuration table. Then, the fields in the evaluation configuration table are identified, for example, character identification can be performed in a line field manner to obtain the evaluation index entered in each line, and the evaluation index entered in each line is synthesized to obtain the evaluation index set.

It should be understood that for ESG evaluation, evaluation is generally performed from three dimensions, namely, an environment dimension (E dimension), a social dimension (S dimension), and a governance dimension (G dimension). Therefore, the plurality of evaluation dimensions mentioned in the present application are essentially the three evaluation dimensions, and the evaluation indexes in the evaluation index set are the evaluation indexes related to the three evaluation dimensions. However, as the subsequent society advances, more evaluation dimensions may be involved, and thus the present application does not limit the types of evaluation dimensions.

For example, the evaluation index set may be preset. For enterprise ESG evaluation, at least one evaluation index may be set in the E dimension, the S dimension, and the G dimension, respectively. Generally, there are qualitative and quantitative evaluation indexes for each evaluation dimension. For example, quantitative evaluation indicators in the E dimension include, but are not limited to: the carbon emission per capita, the electricity consumption per capita and the water consumption per capita, and the qualitative evaluation indexes include but are not limited to: whether energy saving measures exist or not and whether an energy saving management system exists or not; quantitative evaluation indexes in the S dimension include, but are not limited to: employee welfare, coverage of annuals, participation in end-of-year prizes, and qualitative evaluation indexes are controlled but not limited to: whether there is a equity incentive plan and a work claims payment standard; quantitative evaluation indicators in the G dimension include, but are not limited to: stockholder distribution, qualitative assessment indicators include, but are not limited to: the protection system of middle and small shareholders, the situation that large shareholders are reduced, the perfect board architecture, and the board owners and the CEO are not the same person.

102: historical ESG evaluation data of a plurality of enterprises and historical profitability of the plurality of enterprises are obtained.

The historical ESG evaluation data of each enterprise in the plurality of enterprises comprises evaluation indexes used when ESG evaluation is carried out on each enterprise and historical ESG indexes of each evaluation index, and the historical ESG evaluation data of each enterprise is similar to a mode of acquiring enterprise data of the enterprise to be evaluated later and is not repeated. It should be understood that the historical ESG evaluation data and the historical profitability of each enterprise are related to each other, for example, if the historical ESG evaluation data of a certain enterprise in t years is obtained, the historical profitability of the enterprise is also in t years, so that the relationship between the profitability and the ESG evaluation can be constructed.

Illustratively, historical ESG rating data for each enterprise as well as historical profitability may be obtained through crawler technology. For example, historical ESG evaluation data and historical profitability of each enterprise can be obtained from a plurality of news media platforms through a crawler technology; or, the historical ESG evaluation data of each enterprise can be crawled from the official platform of the enterprise to be evaluated through a crawler technology, for example, the historical ESG evaluation data and the historical revenue rate of each enterprise can be obtained from the annual social responsibility report, yearly newspaper, semiannual newspaper, quarterly newspaper and official announcement of the company of each enterprise. The method for acquiring historical ESG evaluation data and historical profitability of the enterprise is not limited.

103: and performing model training by using the historical ESG evaluation data of the plurality of enterprises as training samples and using the historical profitability of the plurality of enterprises as training labels to obtain the actual weight of each evaluation index under each evaluation dimension.

Optionally, the evaluation indexes under each evaluation dimension are retested based on a model training mode to obtain the actual weight of each evaluation index under each evaluation dimension.

Illustratively, the historical ESG evaluation data of each enterprise includes a historical ESG index corresponding to each evaluation index in the evaluation indexes used for evaluating the ESG of each enterprise. Accordingly, a historical ESG index for each business under each of the set of rating indices may be determined. It should be understood that, for a certain enterprise without a historical ESG index under a certain evaluation index, the historical ESG index corresponding to the evaluation index is set to 0; and then, performing model training according to the historical ESG index of each enterprise under each evaluation index and the historical profitability to obtain the actual weight of each evaluation index under each evaluation dimension.

Exemplarily, a historical ESG index of each enterprise under each evaluation index of each evaluation dimension is used as a training sample, and a historical profitability of each enterprise is used as a training label and input into a pre-constructed model, so that the profitability of each enterprise is predicted; determining a predicted loss according to a training label (historical profitability) and a predicted profitability of each enterprise; and adjusting the weight value of each evaluation index under each evaluation dimension in a pre-constructed model according to the predicted loss to obtain the actual weight of each evaluation index under each evaluation dimension.

Wherein the pre-constructed model is constructed according to the corresponding relation between the yield and the ESG index of the evaluation index.

Illustratively, the correspondence relationship may be represented by formula (1):

r ═ α 1 × mark (t1) + α 2 × mark (t2) + … … + α n mark (tn) formula (1);

wherein, R is a profitability, mark is an evaluation operation, that is, an operation of determining an ESG index of each evaluation index, t1, t2, … …, tn is an evaluation index 1, an evaluation index 2, … …, and evaluation indexes tn, α 1, α 2, and.

Wherein the initial weight of each evaluation index in formula (1) is randomly generated. Thus, the initial model is built from randomly generated initial weights.

Therefore, based on the correspondence relationship shown in formula (1), the historical ESG index of each enterprise under each evaluation index of each evaluation dimension is input to formula (1), and the profitability of each enterprise can be predicted.

Then, determining a predicted loss based on the predicted profitability of each enterprise and the historical profitability of each enterprise; adjusting the weight value (namely the initial weight value) of each evaluation index under each evaluation dimension based on the prediction loss and a gradient descent method, and predicting the profitability of each enterprise again by using the adjusted weight value; and sequentially and circularly iterating until the predicted loss is smaller than a threshold value, and taking the weight value of each evaluation index under each evaluation dimension as the actual weight of each evaluation index under each evaluation dimension.

104: and determining the initial weight of each evaluation index under each evaluation dimension according to the historical ESG evaluation data of the plurality of enterprises.

Optionally, a preset score of each preset evaluation index under each evaluation dimension under the evaluation dimension is obtained, where the preset score is used to represent the importance degree of the evaluation index relative to the evaluation dimension, and the preset score of each evaluation index of each evaluation dimension is preset; and carrying out normalization processing on the preset score of each evaluation index under each evaluation dimension to obtain the initial weight of each evaluation index under each evaluation dimension.

For example, the initial weight of each evaluation index in any evaluation dimension can be represented by formula (2):

wherein, w_iIs the initial weight r of the i-th evaluation index in any evaluation dimension_iIs the preset score of the ith evaluation index in the evaluation dimension, n is the number of the evaluation indexes in the evaluation dimension, r_jAnd the preset score of the j-th evaluation index in the evaluation dimension is obtained.

Optionally, determining an evaluation index used when performing ESG evaluation on each enterprise of the multiple enterprises according to historical ESG evaluation data of the multiple enterprises; according to the evaluation indexes used when ESG evaluation is carried out on each enterprise in the plurality of enterprises, the number of times each evaluation index in each evaluation dimension is used is determined, namely the number of times each evaluation index in the evaluation index set is used is determined. It should be understood that if a certain evaluation index in the above evaluation index set is not used all the time when ESG evaluation is performed on the plurality of enterprises, the number of times that the evaluation index is used is set to zero.

And finally, normalizing the used times of each evaluation index under each evaluation dimension to obtain the initial weight of each evaluation index under each evaluation dimension.

For example, the initial weight of each evaluation index in any evaluation dimension can be represented by formula (3):

wherein, w_iIs the initial weight, k, of the i-th evaluation index in any evaluation dimension_iIs the number of times the ith evaluation index in the evaluation dimension is used, n is the number of evaluation indexes in the evaluation dimension, k_jThe number of times the j-th evaluation index in the evaluation dimension is used.

105: and fusing the actual weight and the initial weight of each evaluation index under each evaluation dimension to obtain the target weight of each evaluation index under each evaluation dimension.

Optionally, the method of fusing the initial weight and the actual weight of each evaluation index in each evaluation dimension may be implemented by the following steps:

acquiring the maximum value and the minimum value in the actual weight of each evaluation index under each evaluation dimension, and determining the average value of the actual weight of each evaluation index under each evaluation dimension; and fusing the initial weight and the actual weight of each evaluation index under each evaluation dimension based on the initial weight of each evaluation index under each evaluation dimension, the maximum value, the minimum value and the average value in the actual weight of each evaluation index and a preset hyper-parameter to obtain the weight of each evaluation index under each evaluation dimension.

For example, the weight of each evaluation index in any evaluation dimension can be represented by formula (4):

wi is the initial weight of the i-th evaluation index in any evaluation dimension, fi is the actual weight of the i-th index in the evaluation dimension, mean (F), max (F), min (F) are the average value, the maximum value and the minimum value of the actual weight of each evaluation index in the evaluation dimension respectively, and P is a preset hyper-parameter.

The fusion mode of the formula (4) has the following advantages:

when the initial weight of each evaluation index is calculated by a preset score, namely, the calculation mode of the formula (2) (namely, the weight determined based on expert experience), the initial weight can be adjusted by using the weight obtained by model training and taking the initial weight as the center, and the adjustment proportion is not more than p. Therefore, the initial weight is adjusted by taking the weight given by the expert in the real-time operation as a principle. 2. The optimal model is trained by using historical yield correlation, so that the yield is improved as much as possible, ESG evaluation is more in line with an investment scene, investors are attracted to invest in enterprises with higher ESG indexes to a greater extent, development of pursuing more in line with ESG standards is promoted, and sustainable development is achieved.

106: and obtaining the ESG index of the enterprise to be evaluated according to enterprise data of the enterprise to be evaluated and the target weight of each evaluation index under each evaluation dimension.

The enterprise data of the enterprise to be evaluated is disclosure data related to the evaluation index, for example, if the evaluation index is average human power consumption, the enterprise data includes a specific numerical value of the average human power consumption.

For example, enterprise data of an enterprise to be evaluated can be obtained through a crawler technology. For example, W pieces of news related to the enterprise to be evaluated can be acquired from a plurality of news media platforms through a crawler technology, and the W pieces of news are identified to obtain enterprise data of the enterprise to be evaluated; or the report of the enterprise to be evaluated can be crawled from the official platform of the enterprise to be evaluated through a crawler technology, for example, enterprise data of the enterprise to be evaluated can be obtained from the annual social responsibility report, yearly report, semiannual report, quarterly report and official notice of the enterprise to be evaluated. The method and the device for obtaining the enterprise data of the enterprise to be evaluated are not limited.

For example, the evaluation of the enterprise to be evaluated under each evaluation index in each evaluation dimension is determined based on enterprise data of the enterprise to be evaluated, for example, the evaluation under each evaluation index of each evaluation dimension may be determined according to a preset evaluation rule and disclosure data of each evaluation index in each evaluation dimension. For example, the evaluation index is the average human electricity consumption, and when the average human electricity consumption is 30 degrees, the corresponding evaluation of the average human electricity consumption is set to 10 points; according to the weight of each evaluation index of each evaluation dimension, carrying out weighting processing on the evaluation of the enterprise to be evaluated under each evaluation index of each evaluation dimension to obtain the comprehensive evaluation of the enterprise to be evaluated under each evaluation dimension; then, acquiring the enterprise type of the enterprise to be evaluated, and determining the weight of each evaluation dimension according to the mapping relation between the enterprise type and the weight and the enterprise type of the enterprise to be evaluated, wherein the mapping relation is preset; and according to the weight of each evaluation dimension, carrying out weighting processing on the comprehensive evaluation under each evaluation dimension to obtain the ESG evaluation of the enterprise to be evaluated.

107: and sending the ESG index of the enterprise to be evaluated to target equipment, so that a user of the target equipment can make a decision related to the enterprise to be evaluated according to the ESG index of the enterprise to be evaluated.

It can be seen that, in the embodiment of the present application, based on historical ESG evaluation data of multiple enterprises, an actual weight of each evaluation index in each evaluation dimension is determined, and an initial weight of each evaluation index is adjusted based on the actual weight of each evaluation index to obtain a target weight, instead of manually setting the target weight, so that the obtained target weight is supported by the historical data, the setting of the target weight is more accurate, and further, the obtained ESG index is more accurate, thereby improving the accuracy of the decision making.

In one embodiment of the present application, the ESG index of the present application may have different application scenarios according to different target devices.

Scene 1: when the target equipment is the equipment of the investment institution, the ESG index of the enterprise to be evaluated is sent to the investment institution, and the investment institution can make an investment decision related to the enterprise to be evaluated. For example, because the ESG index of an enterprise reflects the value and sustainable development capability of the enterprise, when the ESG index of an enterprise to be evaluated is high, the investment amount and investment period of the enterprise to be evaluated may be added; when the ESG index of the enterprise to be evaluated is low, the investment on the enterprise to be evaluated can be withdrawn or reduced, and the like. In general, the ESG index of the enterprise to be evaluated is sent to the investment institution, so that the investment of the investment institution can be guided, and the investment risk is reduced.

Scene 2: when the target device is a device of an enterprise to be evaluated, the ESG index of the enterprise to be evaluated is sent to the enterprise to be evaluated, and the enterprise to be evaluated can make a management decision related to the enterprise to be evaluated. Illustratively, as the ESG index of an enterprise reflects the value and sustainable development ability of the enterprise, higher enterprises willing to invest in the ESG index pay more attention to the social responsibility of the enterprise as the investor increases in acceptance of the ESG index. Therefore, when the ESG index of the enterprise to be evaluated is higher, the enterprise to be evaluated can conveniently make a measure for reinforcing management and continuously keep the performance in the aspect of ESG; when the ESG index of an enterprise to be evaluated is low, the enterprise to be evaluated needs to adjust the development strategy of the enterprise, so that the sustainable development of the enterprise is improved, and the ESG index is improved. Generally, the ESG index of the enterprise to be evaluated is sent to the enterprise to be evaluated, so that the enterprise to be evaluated is promoted to strive to improve the ESG evaluation condition of the enterprise to be evaluated, and the benign development of the enterprise to be evaluated is guided.

Scene 3: when the target device is a device of a government or a social organization, the ESG index of the enterprise to be evaluated is sent to the government or the social organization, and the government or the social organization can make a plan related to the enterprise to be evaluated. Illustratively, the ESG index of a business reflects the value and sustainable ability of the business. Therefore, when the ESG index of the enterprise to be evaluated is higher, the development potential of the enterprise to be evaluated is higher, and the enterprise to be evaluated can be greatly promoted so as to provide more development opportunities for the enterprise; when the ESG index of the enterprise to be evaluated is low, which indicates that the development potential of the enterprise to be evaluated is low, the enterprise can be instructed to adjust the development strategy of the enterprise to be evaluated, or the support is reduced, so as to guide the adjustment of the enterprise to be evaluated to the benign development direction.

In an embodiment of the application, because the problem of data acquisition loss or the fact that an enterprise does not want to disclose some data, the obtained enterprise data of the enterprise to be evaluated has some missing disclosure data of evaluation indexes, in order to improve the accuracy of ESG evaluation of the enterprise to be evaluated, the missing disclosure data needs to be complemented, and then the supplemented disclosure data is used for ESG evaluation of the enterprise to be evaluated.

With reference to the accompanying drawings, the data completion process will be described by taking the obtained enterprise data of the enterprise to be evaluated as the enterprise data of the enterprise to be evaluated in a preset time period.

Referring to fig. 2, fig. 2 is a schematic flow chart of a data completion method according to an embodiment of the present disclosure. The content of this embodiment that is the same as that shown in fig. 1 will not be described again. The method comprises the following steps:

201: the method comprises the steps of obtaining enterprise data of an enterprise to be evaluated in a preset time period, wherein the preset time period comprises N historical moments, the enterprise data of the enterprise to be evaluated comprises disclosure data of an evaluation index A in M historical moments, M is less than or equal to N, N is an integer larger than 1, and the evaluation index A is any one of a plurality of evaluation indexes used for ESG evaluation of the enterprise.

The preset time period may be any one of historical time periods, for example, the preset time period may be approximately ten days, last month, last year, or the like. The preset time period is not limited in the present application. That is to say, the ESG index determining method in the present application may perform ESG evaluation based on data in a recent time period, or may trace back a history, and perform ESG evaluation for an enterprise in a certain historical time period based on data in the historical time period.

202: and under the condition that M is smaller than N, the enterprise ESG index determining device completes the disclosure data of the evaluation index A according to the enterprise data of the enterprise to be evaluated in the preset time period and the enterprise data of a plurality of first enterprises in the preset time period to obtain the disclosure data of the evaluation index A at N historical moments, wherein the plurality of first enterprises are enterprises belonging to the same industry as the enterprise to be evaluated.

The obtaining mode of the enterprise data of the plurality of first enterprises in the preset time period is similar to the obtaining mode of the enterprise data of the enterprise to be evaluated in the preset time period, and is not repeated. For example, enterprise data relating to the ESG index for each first enterprise may be crawled from the official platform of each first enterprise. It should be understood that in order to make the completed disclosure data of the enterprise to be evaluated relatively accurate, the disclosure data of the first enterprises are made as complete as possible, and therefore the disclosure data of the first enterprises including the evaluation index a at N historical times is taken as an example in the present application.

Illustratively, when M is greater than 1 and less than N, that is, the enterprise to be evaluated discloses the disclosure data of the evaluation index a in N historical moments, but there is some historical moments that do not disclose the disclosure data of the evaluation index a. For example, when the disclosure data of the evaluation index a is not disclosed at the ith history time of the N history times, the disclosure data of the enterprise to be evaluated related to the evaluation index a at the ith history time may be complemented according to the disclosure data of the enterprise to be evaluated related to the evaluation index a at the history time adjacent to the ith history time, so as to obtain the disclosure data of the enterprise to be evaluated related to the evaluation index a at the N history times. And the historical time adjacent to the ith historical time comprises the (i-1) th historical time and/or the (i + 1) th historical time.

Illustratively, a discount ratio corresponding to an evaluation index a is acquired, and according to the discount ratio and disclosure data related to the evaluation index a of the enterprise to be evaluated at a history time adjacent to the ith history time, the disclosure data related to the evaluation index a of the enterprise to be evaluated at the ith history time is complemented to obtain the disclosure data related to the evaluation index a of the enterprise to be evaluated at the N history times.

Alternatively, the discount rate corresponding to the evaluation index a may be set in advance. It should be noted that, when the evaluation rule of the evaluation index a is that the larger the value of the evaluation index a is, the higher the score is, the value of the discount proportion corresponding to the evaluation index a may be set to be less than 1, so that the unrevealed data may be discounted, the ESG index of the enterprise to be evaluated is reduced, and the enterprise to be evaluated is encouraged to actively disclose the data; when the evaluation rule of the evaluation index A is that the value of the evaluation index A is larger and the score is lower, the value of the discount proportion corresponding to the evaluation index A can be set to be larger than 1, so that undisclosed data can be amplified, the ESG index of an enterprise to be evaluated is reduced, and the enterprise to be evaluated is encouraged to actively disclose data.

Alternatively, it is determined by the number of times of disclosure of the evaluation index a in a set period of time. Specifically, the total number of times that the specified evaluation index a is to be disclosed within a set time period is acquired, and the number of times that the evaluation index a is actually disclosed within the set time period is acquired; and determining the discount proportion corresponding to the evaluation index A according to the total times of the disclosure of the evaluation index A, the actual times of the disclosure of the evaluation index A and the evaluation rule of the evaluation index A. Specifically, when the evaluation rule of the evaluation index a is that the larger the value of the evaluation index a is, the higher the score is, the ratio of the number of times that the evaluation index a actually reveals to the total number of times that the evaluation index a reveals can be taken as the discount ratio; when the evaluation rule of the evaluation index a is that the larger the value of the evaluation index a is, the lower the score is, the ratio of the total number of times that the evaluation index a is to be disclosed to the number of times that the evaluation index a is actually disclosed may be used as the discount ratio.

Optionally, a first product of the disclosure data related to the evaluation index a at the ith-1 th historical time and the discount rate is used as the disclosure data related to the evaluation index a at the ith historical time; (ii) a

Optionally, a second product of the disclosure data related to the evaluation index a at the i +1 th historical time and the discount rate is used as the disclosure data related to the evaluation index a at the i th historical time;

optionally, a first product of the disclosure data related to the evaluation index a at the i-1 th historical time and the discount ratio and a second product of the disclosure data related to the evaluation index a at the i +1 th historical time and the discount ratio are obtained to synthesize the first product and the second product, so as to obtain the disclosure data related to the evaluation index a at the i-th historical time. For example, the average value or the maximum value or the minimum value of the first product and the second product is used as the disclosure data related to the evaluation index a at the ith historical time; or determining the ESG index corresponding to each product of the first product and the second product according to the evaluation rule of the evaluation index A, and taking the product corresponding to the highest ESG index as the disclosure data related to the evaluation index A at the ith historical time.

For example, if the evaluation index a is the average human power consumption, the larger the average human power consumption, the lower the ESG index. Therefore, the discount proportion of the evaluation index A is set or calculated to be 1.2, and when the per capita electricity consumption of the enterprise to be evaluated at the ith-1 historical moment is 30 degrees and the per capita electricity consumption at the ith-1 historical moment is used for complementing the disclosure data at the ith historical moment, the per capita electricity consumption at the ith historical moment is 36; when the per-capita electricity consumption of an enterprise to be evaluated at the (i + 1) th historical moment is 40 degrees, and the per-capita electricity consumption at the (i + 1) th historical moment is used for supplementing the disclosure data at the (i) th historical moment, the per-capita electricity consumption at the (i) th historical moment is 48 degrees; if the disclosure data corresponding to the highest ESG index is used, it can be known that the ESG index corresponding to the average electricity consumption of people of 36 degrees is the highest, and the average electricity consumption of people 36 is used as the average electricity consumption of people of the enterprise to be evaluated at the ith historical moment.

Optionally, if the evaluation indicator a continuously lacks disclosure data at two historical times, for example, there is no disclosure data at the ith historical time and the (i + 1) th historical time, the disclosure data at the ith historical time may be first complemented by using the disclosure data at the (i-1) th historical time, and then the disclosure data at the (i + 1) th historical time may be complemented by using the complemented disclosure data at the ith historical time, that is, the missing data at the two historical times may be complemented by continuously using the discount ratio twice. Of course, the disclosure data at the i +2 th historical time may be used to supplement the disclosure data. And finally, combining the supplemented disclosure data in the two directions to obtain the disclosure data of the evaluation index A at two historical moments, wherein the combination mode is similar to the combination mode and is not described again.

It should be understood that if the ith history time is the first history time in the N history times, the ith history time does not have the previous history time, so the disclosure data at the ith history time can be complemented by only the disclosure data at the next history time, i.e. the (i + 1) th history time; similarly, if the ith history time is the last history time of the N history times, there is no history time at the ith history time, so that the disclosure data at the ith history time can be complemented with the disclosure data at the last history time, i.e., the (i-1) th history time.

In an embodiment of the present application, when M is equal to 0, that is, the enterprise to be evaluated does not disclose the disclosure data related to the evaluation index a at any of the N historical times, the disclosure data of the enterprise to be evaluated at the N historical times cannot be complemented by only the enterprise data of the enterprise to be evaluated.

Optionally, enterprise data of a plurality of first enterprises are acquired, and the plurality of first enterprises disclose disclosure data of the evaluation index a at N historical times. And performing ESG evaluation on the enterprises to be evaluated according to the enterprise data of the enterprises to be evaluated to obtain the ESG index of the enterprises to be evaluated. In the process of carrying out ESG evaluation on an enterprise to be evaluated, because the enterprise to be evaluated lacks disclosure data related to an evaluation index a, the ESG index of the evaluation index may be set to zero for the evaluation index a, or the evaluation index a is not subjected to ESG evaluation, that is, ESG evaluation is roughly carried out on the enterprise to be evaluated based on the existing disclosure data (that is, existing disclosure data related to other evaluation indexes) of the enterprise to be evaluated, so as to obtain the ESG index of the enterprise to be evaluated.

Then, determining an average value of the disclosure data related to the evaluation index A at each historical time of the plurality of first enterprises; according to the average value of the disclosure data of the plurality of first enterprises at each historical time and the ESG indexes of the plurality of first enterprises and the ESG index of the enterprise to be evaluated, completing the disclosure data of the enterprise to be evaluated, which is related to the evaluation index A at the N historical times, to obtain the disclosure data of the enterprise to be evaluated, which is related to the evaluation index A at the N historical times.

The disclosure data related to the evaluation index a at each historical time of the enterprise to be evaluated can be represented by formula (5):

formula (5) is si ═ m (s/smax);

the method comprises the steps that si is the disclosure data of an enterprise to be evaluated and an evaluation index A at each historical time, m is the average value of the disclosure data of a plurality of first enterprises and the evaluation index A at each historical time, s is the ESG index of the enterprise to be evaluated, and smax is the maximum value of the ESG indexes of the plurality of first enterprises and the ESG index of the enterprise to be evaluated.

The data completion mode has the following advantages:

1. because the missing exposed data is complemented, the self-exposed behavior (whether the data is exposed) of the enterprise also reflects the management level and the management capability of the enterprise, so for the enterprise which does not expose the data of a certain evaluation index, the complemented estimation value should be greatly discounted. This step is to further encourage businesses to make the disclosure of data as complete as possible;

2. the business to be rated using the estimates is not necessarily the worst performing business, but simply because no data is disclosed. Therefore, the disclosure data using the default value of 0 as the evaluation index is also not reasonable. Specifically, there are some enterprises that do not disclose data under a certain evaluation index because the data is not good, and because the data is not disclosed to cause the overall performance of the enterprise to be better, even if the performance is good, the enterprise cannot give a good evaluation value; in addition, there are some enterprises that do not disclose data under a certain evaluation index because they forget to disclose it, and because this item of data is not good, they cannot give too low data (e.g., 0). Therefore, the reasonable estimation value should be reasonably estimated based on the overall performance of the enterprise to be evaluated, and taking mean as the highest estimation value is a reasonable option, namely, the estimation value of the enterprise to be evaluated is controlled to be smaller than or equal to the average value.

203: and obtaining the disclosure data of the enterprise to be evaluated, which is related to each evaluation index at the N historical moments, according to the disclosure data of the enterprise to be evaluated, which is related to the evaluation index A at the N historical moments.

Illustratively, the disclosed data of the evaluation index A is complemented to obtain the disclosed data related to the evaluation index A at N historical moments, and the evaluation index A is any one of a plurality of evaluation indexes, so that the data complementation is realized for any evaluation index with the missing disclosed data, and each index of the plurality of evaluation indexes has the disclosed data at N historical moments, namely the disclosed data related to each evaluation index of the enterprise to be evaluated at the N historical moments.

It can be seen that, in the embodiment of the present application, when the disclosure data of an enterprise to be evaluated under a certain evaluation index is missing, the disclosure data of the evaluation index can be complemented according to the enterprise data of the evaluation enterprise in a preset time period and the enterprise data of the same industry in the preset time period, instead of simply using an average value to perform data complementation, so that the complemented data is more matched with the enterprise to be evaluated, and then the complemented data is used to perform ESG evaluation, and the obtained ESG index is relatively more accurate, so that the formulated decision is more accurate.

The method of setting the weight and the method of data completion are described above. The following describes how to perform automated ESG evaluation using the set weights and the completed data.

Referring to fig. 3, fig. 3 is a schematic flowchart of another method for determining an enterprise ESG index according to an embodiment of the present application. In this embodiment, the same contents as those shown in fig. 1 and 2 will not be described again. The method comprises the following steps:

301: and acquiring the disclosure data of the enterprise to be evaluated under the evaluation index A.

The evaluation index a may be any one evaluation index in the evaluation index set. Namely, the disclosure data of the enterprise to be evaluated under the evaluation index A is determined according to the enterprise data of the enterprise to be evaluated.

302: and structuring the exposed data under the evaluation index A according to a preset data structure corresponding to the evaluation index A to obtain at least one piece of structured data, wherein each piece of structured data of the at least one piece of structured data corresponds to one evaluation item under the evaluation index A.

Illustratively, a data structure of each evaluation index is defined in advance. Because there may be multiple evaluation items required to participate in the ESG evaluation under each evaluation index, and each evaluation item may require multiple fields to be clearly described. Therefore, a data structure as shown in table 1 may be set for the evaluation index a:

table 1:

as shown in table 1, the data structure of the evaluation index a includes a first field index _ code, at least one second field, i.e., fields data _ label _1, data _ label _2, data _ label _3, a third field data _ text, and a fourth field data _ digit. Wherein the first field index _ code is used for indicating the identification of the evaluation index, and the at least one second field is used for indicating the identification of the evaluation item of the evaluation index; when the evaluation item of the evaluation index is a qualitative evaluation item, the third field data _ text is used for indicating the evaluation content of the evaluation item, and the fourth field data _ digit is emptied; when the evaluation item of the evaluation index is a quantitative evaluation item, the fourth field data _ digit is used to indicate quantitative data of the evaluation item, and the third field data _ text is nulled.

It should be understood that, after the exposed data of the evaluation index a is structured according to the data structure shown in table 1, the value of each piece of structured data in the at least one piece of structured data under each field is referred to as the attribute value of the piece of structured data under each field.

In one embodiment of the present application, as shown in table 2, the data structure corresponding to the evaluation index a may further include the following optional fields: a field company _ id, a field rl _ second _ level _ index, a field data _ source, a field model _ version, a field remap, and a field data _ day.

Table 2:

the field company _ id is used for indicating the identification of an enterprise, and the field attribute value is mainly used for indicating which enterprise each structured data belongs to when the enterprise data of a plurality of enterprises is structured; the field rl _ second _ level _ index is used for indicating the industry level to which the enterprise belongs; the field data _ source is used for indicating the source of enterprise data of the enterprise; the field model _ version is used to indicate the item version; the field remark is used for marking each piece of structured data, i.e. similar to remarks; the field data _ day is used to indicate time, i.e., an attribute value of the field data _ day is used to indicate a disclosure date of the enterprise data.

In one embodiment of the present application, the bullets data _ label _1, data _ label _2, and data _ label _3 may or may not have a hierarchical relationship, and need to be determined according to the specific type of the evaluation item. For example, when the evaluation item is a winning prize, the data _ label _1 is used for recording that the evaluation item is a winning prize, and the data _ label _2 is used for indicating that the evaluation item is a national prize, namely, the attribute value of the data _ label _1 and the attribute value of the data _ label _2 have a hierarchical relationship.

It should be noted that designing a plurality of second fields is merely an exemplary illustration, and a greater or lesser number of second fields may be designed in practical applications. In addition, for the case where the evaluation item needs a plurality of second fields for description, one second field description may be used, and the evaluation item may be described by way of hierarchical description in one second field. For example, data _ label _1 may be used directly to indicate that the evaluation item is a winning and a national prize. Of course, in the case where only one second field description is needed for one evaluation item, the redundant second fields may be left blank.

It should be noted that the above refers to setting the field, i.e. the attribute value of the field, to "NULL".

The process of structuring the enterprise data of the enterprise to be evaluated is illustrated below by tables 3 and 4:

table 3: the enterprise data is structured by a second field.

As shown in table 3, based on the definition of the data structure, the identifier of the enterprise to be evaluated is as follows: 12137736, the industry level of the enterprise to be evaluated is: in the printing industry, the marks of the evaluation indexes are as follows: g015, namely the 15 th evaluation index under the G dimension, the evaluation item under the evaluation index is indicated by the first second field, and the evaluation item is length, the second field and the third second field are empty, that is, the value is NULL, and the quantitative value of the evaluation item is: 3.1, qualitatively evaluating the blank (NULL); the sources of the data of the enterprise to be evaluated are as follows: raw _ fr _ directors; the project version is: 1.0; remark is: NULL (NULL), the date of disclosure of the data of the enterprise to be evaluated is: 2020-08-26.

Table 4 the enterprise data is structured by a plurality of second fields.

As shown in table 4, based on the definition of the data structure, the identifier of the enterprise to be evaluated is as follows: 140485480, the industry level of the enterprise to be evaluated is: general purpose, special purpose equipment manufacturing; the evaluation index is identified as: s121, namely a 121 th evaluation index under the dimension S; the evaluation items under the evaluation indexes are represented by three second fields, namely, the evaluation items are indicated as winning prizes, and the prizes of the three provinces are obtained, namely, the prizes of the three provinces of Zhejiang province, scientific and technical, and the like are obtained; the qualitative evaluation content is as follows: the prize is won through the consistent efforts of scientific research, technical personnel and the like; the quantitative inner container is empty; the sources of the data of the enterprise to be evaluated are as follows: raw _ fr _ directors; the project version is: 1.0; the remarks are as follows: NULL, the date of disclosure of the data of the enterprise to be evaluated is: 2020-08-26.

It should be understood that although the evaluation items of each evaluation index are defined in advance, if the disclosure data of a certain evaluation item is not acquired at present, the structured data corresponding to the evaluation item is missing, and it can be understood that the structured data corresponding to the evaluation item is not structured.

That is to say, a plurality of evaluation items are predefined in the data structure of each evaluation index, and if the currently acquired data of an evaluation item is not disclosed in the disclosure data of the enterprise to be evaluated under the evaluation index, the evaluation item is not structured, and the structured data corresponding to the evaluation item does not exist. Or, the data of the evaluation item is supplemented first by a data supplementing mode, and then the data of the evaluation item is subjected to structuring processing to obtain the structured data of the evaluation item. The data completion process may refer to the completion process shown in fig. 2, and is not described again.

Furthermore, the exposed data under each evaluation index is identified through an NLP technology to obtain the related data of the evaluation item under each evaluation index, and the related data of each evaluation item is structured based on the defined data structure, that is, the value of each evaluation item under each field is assigned to the field to obtain the structured data corresponding to each evaluation item.

It should be understood that there may be multiple pieces of disclosure data for each evaluation item, for example, the enterprise data of the enterprise to be evaluated includes disclosure data of the evaluation item at N historical times, however, when the ESG evaluation is performed on the enterprise, only one piece of disclosure data may be needed. Therefore, the disclosure data of the N historical moments can be integrated, for example, averaged, and the disclosure data obtained after fusion is subjected to structuring processing to obtain the structured data corresponding to the evaluation item.

303: and configuring a target evaluation code corresponding to each evaluation item in the at least one evaluation item according to the at least one piece of structured data.

For example, all evaluation indexes have 12 evaluation rules at present. For example, for a quantitative evaluation index, the evaluation rule is: ESG evaluation is carried out according to industry ranking; for another example, for qualitative evaluation indexes, the evaluation rule is: whether there is a corresponding description for ESG evaluation.

Therefore, for these 12 evaluation rules, a general evaluation code is compiled for each evaluation rule in advance.

Generally, evaluation rules can be classified into two main categories, namely qualitative evaluation rules and quantitative evaluation rules. For quantitative evaluation rules, generally speaking, the meaning described based on quantitative data is different, and the corresponding evaluation manner is different, so that the quantitative evaluation rules are relatively more.

For example, the following generic evaluation code may be compiled for an industry-ranked evaluation rule:

{“index_code@data_label”：{

“[0，25)”：1，

“[25，50)”：2，

“[50，75)”：3，

“[75，100)”：4

}

wherein, index _ code is a first parameter in the code and is used for indicating the identification of the evaluation index, and data _ label is a second parameter in the code and is used for indicating the identification of the evaluation item under the evaluation index. "[0, 25)": 1 is used for indicating that when the evaluation item value, namely the industry rank is between [0, 25 ], the ESG index of the evaluation item data _ label is 1; by analogy, "[ 25, 50)": 2,"[50, 75)": 3,"[75, 100)": 4 indicate ESG indices of 2, 3, 4, respectively.

It should be noted that the generality of the code is embodied in that the first parameter index _ code and the second data _ label do not take specific values. When the ESG evaluation is performed on a specific evaluation index, if the evaluation rule of a certain evaluation item in the evaluation index is to perform the ESG evaluation through industry ranking, only the code is required to be called, the first entry index _ code and the second data _ label are endowed with the value corresponding to the evaluation index, namely, the target evaluation code corresponding to the evaluation item under the evaluation index is configured, and the ESG evaluation is performed on the evaluation item through the target evaluation code to obtain the ESG index of the evaluation item.

For qualitative evaluation, the evaluation means is to score a point if there is a corresponding disclosure, and to score 0 if there is no disclosure. Therefore, for such an evaluation rule of qualitative evaluation, the following general evaluation code may be defined:

{“index_code@data_label”：1}

wherein index _ code is an identifier of an evaluation index, data _ label is an identifier of an evaluation item of the evaluation index, "index _ code @ data _ label": 1 denotes that if there is a description of an evaluation item in the evaluation index denoted by "index _ code @ data _ label", the evaluation item is given a score.

Similarly, the generality of the code is embodied in that the first input parameter index _ code and the second input parameter data _ label do not take specific values. When the ESG evaluation is performed on a specific evaluation index, if the evaluation rule of a certain evaluation item in the evaluation index is to perform the ESG evaluation through industry ranking, only the code is called, the first entry index _ code and the second data _ label are given to the value corresponding to the evaluation index, namely, the target evaluation code corresponding to the evaluation item under the evaluation index is configured, and the ESG evaluation is performed on the evaluation item through the target evaluation code to obtain the ESG index of the evaluation item.

Therefore, a general evaluation code corresponding to each evaluation rule can be configured in advance for each evaluation rule, and the value of the input parameter in the general evaluation code corresponding to each evaluation rule is not assigned.

Thus, an evaluation rule of each evaluation item in the evaluation index a is acquired; calling a configured general evaluation code corresponding to each evaluation item according to the evaluation rule of each evaluation item in the evaluation index A; and assigning the parameters in the general evaluation codes corresponding to each evaluation item according to at least one piece of structured data to obtain target evaluation codes corresponding to each evaluation item. Specifically, a first input parameter in the general evaluation code corresponding to each evaluation item is assigned as an attribute value of a first field in the structured data corresponding to each evaluation item, and a second input parameter in the general evaluation code corresponding to each evaluation item is assigned as an attribute value of a target second field, so as to obtain a target evaluation code corresponding to each evaluation item; and the target second field is a second field which is used for participating in ESG evaluation in at least one second field in the structured data corresponding to each evaluation item. Since each evaluation item may be described by a plurality of second fields, however, the second fields used for participating in the ESG evaluation may be only a part, for example, the evaluation item described in table 4 above is a winning prize situation, and the evaluation items are described by three second fields, when the ESG evaluation is actually performed, only the attribute value of the first second field may be needed, that is, whether there is a winning prize record in the first second field is checked, if there is a winning prize record in the first second field, a point is marked, and the ESG evaluation on the evaluation item is completed. Therefore, according to the evaluation rule of each evaluation item, a target second field in at least one second field corresponding to each evaluation item is determined, and the attribute value of the target second field is assigned to the second input parameter. And then, after the assignment of the first input parameter and the second input parameter is completed, the target evaluation code corresponding to each evaluation item is obtained.

304: and determining the ESG index of the at least one evaluation item according to the target evaluation code corresponding to each evaluation item and the at least one piece of structured data.

For example, since the first entry and the second entry in the target evaluation code under each evaluation item have specific values, the specific value (i.e., callback) of the second entry can be determined from the structured data corresponding to each evaluation item through the target evaluation code under each evaluation item. Specifically, according to the assignment of the first input parameter and the assignment of the second input parameter in the target evaluation code corresponding to each evaluation item, the structured data corresponding to each evaluation item in at least one piece of structured data is determined, that is, according to the assignment of the first parameter and the assignment of the second parameter, the assignment of the first parameter and the assignment of the second parameter and the attribute values of the first field and the second field in at least one piece of structured data are used as the structured data corresponding to each evaluation item; then, the parameter values of the second entry in the target evaluation code corresponding to each evaluation item are obtained from the structured data corresponding to each evaluation item. Specifically, when the second parameter is a quantitative evaluation item, the attribute value of the fourth field in the structured data, that is, the attribute value of data _ didit, may be used as the parameter value of the second parameter; and when the second entry is a qualitative evaluation item, taking the attribute value of the third field in the structured data, namely the attribute value of the data _ text, as the parameter value of the second entry. Further, after the parameter value of the second parameter is obtained, executing the target evaluation code of each evaluation item to obtain the ESG index of each evaluation item; finally, according to the ESG index of each evaluation item, an ESG index of at least one evaluation item is obtained, that is, the ESG index of each evaluation item is integrated, for example, weighted, to obtain the ESG index of at least one evaluation item.

It should be noted that there may be a plurality of qualitative evaluation items and/or a plurality of quantitative evaluation items for one evaluation index. Therefore, if the evaluation rules corresponding to the two evaluation items under one evaluation index are the same, the common evaluation code corresponding to the evaluation rule can be multiplexed. For example, if there are multiple qualitative evaluations under one evaluation index, the common code corresponding to the qualitative evaluation may be multiplexed to complete the ESG evaluation of the evaluation index.

Next, the ESG evaluation procedure for the evaluation item under the evaluation index a will be described, taking the evaluation index a as a quantitative evaluation index and a qualitative evaluation index as an example.

Quantitative evaluation:

the evaluation index A is power consumption density, and the evaluation item under the evaluation index A is 'per capita power consumption', and then after the calling of the general evaluation code is completed, the first input parameter and the second input parameter in the general evaluation code can be assigned to obtain the following target evaluation code:

{“E012@density”：{

“[0，25)”：1，

“[25，50)”：2，

“[50，75)”：3，

“[75，100)”：4

}

where E012 is a sign of "power density" as an evaluation index, and density is a sign of "electricity consumption per person" as an evaluation item under the "power density" as the evaluation index.

It can be seen that, in the ESG evaluation process of the "average electricity consumption for people", the structured data corresponding to the "average electricity consumption for people" is obtained from the at least one structured data through the identifier "E012" of the electricity density and the identifier "sensitivity" of the "average electricity consumption for people", and then, the attribute value of the fourth field data _ didit of the structured data is used as the parameter value of the second parameter, that is, the ranking of the average electricity consumption for people of the enterprise to be evaluated in the industry. Determining an ESG index for the evaluation item based on the parameter values of the second parameter. For example, if the average power consumption of the enterprise to be evaluated is 30 in rank in the industry, the ESG score of the enterprise to be evaluated under the evaluation item "average power consumption of people" is determined to be 2.

And (3) qualitative evaluation:

for example, for the evaluation index a being "energy consumption condition disclosure", and there are a plurality of qualitative evaluation items, for example, whether there is an energy use condition disclosed, "whether there is an objective describing energy saving," whether there is an objective describing energy saving quantitatively, "" whether there is a measure describing energy saving qualitatively, "" whether there is quantitative measure effect data, "whether there is a management system describing energy saving," and for the plurality of qualitative evaluation items, the evaluation rules are: a corresponding description is present and then scored as 1, and if not present, then scored as 0. Therefore, for the multiple qualitative ratings, after the general-purpose evaluation code corresponding to the qualitative rating is called and multiplexed, the following target evaluation codes corresponding to the multiple evaluation items can be obtained:

{“E003@disclosed”：1，

“E003@goal_”：1

“E003@quantity_goal_”：1

“E003@certified”：1

“E003@action”：1

“E003@action_result”：1

“E003@management”：1，

}

wherein, "E003E @ discrete": 1, "E003 @ coarse _": 1 "E003 @ quality _ good _": 1, "E003 @ certified": 1, "E003 @ action": 1, "E003 _ @ action _ result": 1, "E003 @ management": the method is used for carrying out ESG evaluation on 'whether the energy use condition is disclosed', 'whether an energy saving target is described', 'whether a quantitative energy saving target is described', 'whether a measure for describing energy saving qualitatively exists', 'whether quantitative measure effect data exists' and 'whether a management system for describing energy saving exists' respectively. And if the corresponding description exists in a certain evaluation item, scoring the evaluation item, otherwise, setting the score of the evaluation item to zero.

It can be seen that the above-described manner of configuring the code starts from the essence of the evaluation rule, and a common evaluation code is written for each evaluation rule. Thus, aiming at 12 evaluation rules, only 12 general evaluation codes need to be written, and when ESG evaluation is performed on each evaluation index, the ESG evaluation on the evaluation index can be completed only by calling the corresponding general evaluation code according to the type of the evaluation index. Therefore, when ESG evaluation is carried out on a plurality of evaluation indexes, corresponding evaluation codes do not need to be written for each evaluation index, the redundancy of the codes is reduced, and the grading efficiency is improved. In addition, from the essence of the evaluation rule, when a new evaluation rule appears subsequently, a code segment can be written for the evaluation rule, so that the evaluation code is convenient to modify, or when a certain evaluation rule is no longer applicable, the code segment corresponding to the evaluation rule is directly deleted, or when the evaluation content of a certain evaluation rule needs to be modified, the code segment of the evaluation rule is directly found and modified, the evaluation code of each evaluation index is not required to be modified, so that the maintenance of the general evaluation code of the new and old evaluation rules is convenient.

305: and determining the ESG index of the enterprise to be evaluated according to the ESG index of the at least one evaluation item.

For example, the ESG index of each evaluation item under each evaluation index may be summed up as the ESG index under each evaluation index. In addition, a corresponding weight may be set for each evaluation item, and the ESG index of each evaluation item may be weighted and summed by the preset weight to serve as the ESG index under each evaluation index. Then, for a plurality of evaluation indexes, weighting and summing the ESG indexes of the plurality of evaluation indexes according to the weights of the plurality of evaluation indexes to obtain the ESG index of the enterprise to be evaluated. The setting method of the weights of the plurality of evaluation indexes can be referred to the setting method shown in fig. 1, and will not be described.

Referring to fig. 4, fig. 4 is a block diagram illustrating functional units of an enterprise ESG index determination apparatus 400 according to an embodiment of the present application. The enterprise ESG index determining apparatus 400 includes: an obtaining unit 401, a processing unit 402 and a sending unit 403, wherein:

an obtaining unit 401, configured to obtain an evaluation index set, where the evaluation index set includes evaluation indexes in multiple evaluation dimensions, and each evaluation dimension includes at least one evaluation index; acquiring historical ESG evaluation data of a plurality of enterprises and historical profitability of the plurality of enterprises;

a processing unit 402, configured to perform model training using historical ESG evaluation data of the multiple enterprises as training samples and historical profitability of the multiple enterprises as training labels, to obtain an actual weight of each evaluation index in each evaluation dimension;

the sending unit 403 is configured to send the ESG index of the enterprise to be evaluated to a target device, so that a user of the target device makes a decision related to the enterprise to be evaluated according to the ESG index of the enterprise to be evaluated.

In some possible embodiments, the historical ESG evaluation data for the plurality of enterprises includes a historical ESG index for the plurality of enterprises at each evaluation index for each evaluation dimension; in terms of performing model training by using the historical ESG evaluation data of the multiple enterprises as training samples and using the historical profitability of the multiple enterprises as training labels to obtain an actual weight of each evaluation index in each evaluation dimension, the processing unit 402 is specifically configured to:

taking a historical ESG index evaluation of each enterprise in the plurality of enterprises under each evaluation index of each evaluation dimension as a training sample;

inputting the training samples into a pre-constructed model to predict the profitability of each enterprise, wherein the pre-constructed model is constructed according to the corresponding relation between the profitability and the ESG index of the evaluation index;

determining a predicted loss according to the predicted profitability of each enterprise and the predicted training label of each enterprise, wherein the label of each enterprise is the historical profitability of each enterprise;

and adjusting the weight value of each evaluation index under each evaluation dimension in the pre-constructed model according to the predicted loss to obtain the actual weight of each evaluation index under each evaluation dimension.

In some possible embodiments, the historical ESG evaluation data for the plurality of enterprises includes a historical ESG index for the plurality of enterprises at each evaluation index for each evaluation dimension; in terms of determining an initial weight of each evaluation index in each evaluation dimension according to the historical ESG evaluation data of the plurality of enterprises, the processing unit 402 is specifically configured to:

determining an evaluation index used when ESG evaluation is carried out on each enterprise in the plurality of enterprises according to the historical ESG indexes of the plurality of enterprises under each evaluation index of each evaluation dimension;

determining the number of times each evaluation index in each evaluation dimension is used according to the evaluation index used when ESG evaluation is performed on each enterprise in the plurality of enterprises;

and performing normalization processing on the number of times that each evaluation index under each evaluation dimension is used to obtain the initial weight of each evaluation index under each evaluation dimension.

In some possible embodiments, in terms of obtaining a target weight of each evaluation index in each evaluation dimension by fusing the initial weight and the actual weight of each evaluation index in each evaluation dimension, the processing unit 402 is specifically configured to:

acquiring the maximum value, the minimum value and the average value of the initial weight of at least one evaluation index under each evaluation dimension;

and fusing the initial weight and the actual weight of each evaluation index in each evaluation dimension according to the initial weight of each evaluation index in each evaluation dimension, the maximum value, the minimum value and the average value of the initial weight of at least one evaluation index in each evaluation dimension, and a preset hyper-parameter to obtain the target weight of each evaluation index in each evaluation dimension.

In some possible embodiments, the actual weight and the initial weight of each evaluation index in each evaluation dimension are fused by the following formula;

wi is an initial weight of the i-th evaluation index in each evaluation dimension, fi is an actual weight of the i-th index in each evaluation dimension, mean (F), max (F), and min (F) are respectively an average value, a maximum value and a minimum value of the actual weight of each evaluation index in each evaluation dimension, and P is a preset hyper-parameter.

In some possible embodiments, in terms of obtaining the ESG index of the enterprise to be evaluated according to enterprise data of the enterprise to be evaluated and the target weight of the evaluation index in each evaluation dimension, the processing unit 402 is specifically configured to:

obtaining an evaluation rule of each evaluation index under each evaluation dimension;

determining an ESG index corresponding to each evaluation index in each evaluation dimension according to the enterprise data of the enterprise to be evaluated and the evaluation rule of each evaluation index in each evaluation dimension;

according to the target weight of each evaluation index in each evaluation dimension, weighting the ESG index corresponding to each evaluation index in each evaluation dimension to obtain the ESG index in each evaluation dimension;

and acquiring preset weight of each evaluation dimension, and performing weighting processing on the ESG index under each evaluation dimension to obtain the ESG index of the enterprise to be evaluated.

In some possible embodiments, in terms of obtaining the evaluation index set, the processing unit 402 is specifically configured to: acquiring an evaluation index configuration table, wherein evaluation indexes under the multiple evaluation dimensions are recorded in the evaluation index configuration table; and identifying fields in the evaluation index configuration table to obtain the evaluation index set.

Referring to fig. 5, fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present disclosure. As shown in fig. 5, the electronic device 500 includes a transceiver 501, a processor 502, and a memory 503. Connected to each other by a bus 504. The memory 503 is used to store computer programs and data, and may transfer the data stored by the memory 503 to the processor 502.

The processor 502 is configured to read the computer program in the memory 503 to perform the following operations:

the control transceiver 501 obtains an evaluation index set, where the evaluation index set includes evaluation indexes in multiple evaluation dimensions, and each evaluation dimension includes at least one evaluation index;

the control transceiver 501 sends the ESG index of the enterprise to be evaluated to the target device, so that the user of the target device makes a decision related to the enterprise to be evaluated according to the ESG index of the enterprise to be evaluated.

Specifically, the transceiver 501 may be configured to implement the functions implemented by the obtaining unit 401 and the sending unit 403 of the enterprise ESG index determination apparatus 400 in the embodiment shown in fig. 4, and the processor 502 may be configured to implement the functions implemented by the processing unit 402 of the enterprise ESG index determination apparatus 400 in the embodiment shown in fig. 4. Therefore, the specific functions of the transceiver 501 may be referred to as the specific functions of the acquiring unit 401 and the transmitting unit 403 and the specific functions of the processor 502 may be referred to as the processing unit 402, which is not described again.

Embodiments of the present application also provide a computer-readable storage medium, which stores a computer program, where the computer program is executed by a processor to implement part or all of the steps of any one of the enterprise ESG index determination methods as described in the above method embodiments.

Embodiments of the present application also provide a computer program product comprising a non-transitory computer-readable storage medium storing a computer program, the computer program being operable to cause a computer to perform part or all of the steps of any one of the enterprise ESG index determination methods as set forth in the above method embodiments.

It should be noted that, for simplicity of description, the above-mentioned method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the present application is not limited by the order of acts described, as some steps may occur in other orders or concurrently depending on the application. Further, those skilled in the art should also appreciate that the embodiments described in the specification are exemplary embodiments and that the acts and modules referred to are not necessarily required in this application.

In the foregoing embodiments, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.

In the embodiments provided in the present application, it should be understood that the disclosed apparatus may be implemented in other manners. For example, the above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only one type of division of logical functions, and there may be other divisions when actually implementing, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not implemented. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of some interfaces, devices or units, and may be an electric or other form.

The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit may be implemented in the form of hardware, or may be implemented in the form of a software program module.

The integrated units, if implemented in the form of software program modules and sold or used as stand-alone products, may be stored in a computer readable memory. Based on such understanding, the technical solution of the present application may be substantially implemented or a part of or all or part of the technical solution contributing to the prior art may be embodied in the form of a software product stored in a memory, and including several instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method described in the embodiments of the present application. And the aforementioned memory comprises: a U-disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic or optical disk, and other various media capable of storing program codes.

Those skilled in the art will appreciate that all or part of the steps in the methods of the above embodiments may be implemented by associated hardware instructed by a program, which may be stored in a computer-readable memory, which may include: flash Memory disks, Read-Only memories (ROMs), Random Access Memories (RAMs), magnetic or optical disks, and the like.

The foregoing detailed description of the embodiments of the present application has been presented to illustrate the principles and implementations of the present application, and the above description of the embodiments is only provided to help understand the method and the core concept of the present application; meanwhile, for a person skilled in the art, according to the idea of the present application, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present application.

Claims

1. An enterprise ESG index determination method, comprising:

2. The method of claim 1, wherein the historical ESG rating data for the plurality of businesses comprises a historical ESG index for the plurality of businesses at each rating index for each rating dimension;

performing model training by using the historical ESG evaluation data of the plurality of enterprises as training samples and using the historical profitability of the plurality of enterprises as training labels to obtain the actual weight of each evaluation index under each evaluation dimension, including:

3. The method of claim 1, wherein the historical ESG rating data for the plurality of businesses comprises a historical ESG index for the plurality of businesses at each rating index for each rating dimension;

the determining an initial weight of each evaluation index under each evaluation dimension according to the historical ESG evaluation data of the plurality of enterprises includes:

4. The method according to any one of claims 1 to 3, wherein the fusing the actual weight and the initial weight of each evaluation index in each evaluation dimension to obtain the target weight of each evaluation index in each evaluation dimension comprises:

5. The method of claim 4,

fusing the actual weight and the initial weight of each evaluation index under each evaluation dimension through the following formula;

6. The method of claim 5, wherein the obtaining the ESG index of the enterprise to be evaluated according to enterprise data of the enterprise to be evaluated and a target weight of the evaluation index in each evaluation dimension comprises:

7. The method according to claim 1, wherein the obtaining of the set of evaluation metrics comprises:

obtaining an evaluation index configuration table, wherein evaluation indexes under the multiple evaluation dimensions are recorded in the evaluation index configuration table;

and identifying fields in the evaluation index configuration table to obtain the evaluation index set.

8. An apparatus for determining an enterprise ESG index, comprising:

9. An electronic device, comprising: a processor coupled to the memory, and a memory for storing a computer program, the processor being configured to execute the computer program stored in the memory to cause the electronic device to perform the method of any of claims 1-7.

10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program which is executed by a processor to implement the method according to any one of claims 1-7.