Social value evaluation device, social value evaluation method, and program

Abstract

This social value evaluation device is provided with: a feature amount generation unit that generates a feature amount from textual information relating to the evaluation of social values; an input unit that inputs textual information relating to a subject for evaluation; an evaluation unit that evaluates relevance between the textual information input by the input unit and the feature amount; and an output unit that outputs a result of the evaluation by the evaluation unit.

Description

Technical Field

[0001] The present invention relates to a technology for evaluating the social value of enterprises and the like.

Background Art

[0002] Non-financial information of enterprises represented by ESG (Environment, Society, Governance) has been emphasized in evaluating the social value of enterprises.

[0003] So far, in the judgment of corporate value in investment and the like, the economic value of enterprises, that is, financial information (such as profit amount and profit rate), has been used as the main evaluation material. However, in recent years, not only financial information but also non-financial information such as efforts towards ESG and SDGs has become an important material for screening investment destinations in evaluating the social value of enterprises. On the other hand, as for the way of existence of enterprises themselves, there have also emerged corporate activities that attach more importance to social value than economic value, such as the corporate activities of Grameen Bank.

Prior Art Documents

Non-Patent Documents

[0004]

Non-Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] When evaluating the social value of a company, currently, an index of ESG investment brands (referred to as an ESG index) is generally used. However, the ESG index is evaluated based on the public information of the company (such as annual updates), and the details of the evaluation criteria and processes are not publicly available. Also, due to the low update frequency, the correlation with the daily financial information of the company is not clear.

[0006] In addition, it is assumed that the public information of companies updated once a year or so is often created with the involvement of external consulting and is information based on subjective evaluations. Therefore, it is unclear whether the ESG efforts (for example, efforts related to SDGs) in the public information of companies are in line with the actual situation of corporate activities.

[0007] The present invention has been made in view of the above points, and an object thereof is to provide a technology that enables the evaluation of the social value in line with the actual situation of the activities of the evaluation target of social value.

Means for Solving the Problems

[0008] According to the disclosed technology, a feature quantity generation unit that generates a plurality of sentences representing goals for activities to enhance social value Generate target feature quantities consisting of a plurality of keywords from a plurality of target texts of the SDGs, and from the target feature quantities, generate feature quantities consisting of vectors and, an input unit that inputs text information regarding the evaluation target, an evaluation unit that evaluates the relevance between the text information input by the input unit and the feature quantity, and an output unit that outputs the evaluation result by the evaluation unit are provided in a social value evaluation device.

Effects of the Invention

[0009] According to the disclosed technology, a technology is provided that enables the evaluation of the social value in line with the actual situation of the activities of the evaluation target of social value.

Brief Description of the Drawings

[0010]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Embodiments for Carrying Out the Invention

[0011] Hereinafter, embodiments of the present invention (the present embodiments) will be described with reference to the drawings. The embodiments described below are merely examples, and the embodiments to which the present invention is applied are not limited to the following embodiments.

[0012] For example, in the following embodiments, a company is assumed as the evaluation target of social value. However, the technology according to the present invention is applicable not only when the evaluation target is a company. For example, it is also possible to evaluate the social value of an individual, a group, a country, a local government, etc. by the technology according to the present invention. Further, in the following embodiments, processing is performed on Japanese text, but this is an example, and the technology according to the present invention is applicable to any language.

[0013] (Problems, Outline of Embodiments) As described above, in the prior art, an ESG index is often used when evaluating the social value of a company. However, it is uncertain whether the ESG index reflects the actual state of a company's activities that change every moment. First, regarding this point, an example of a conventional ESG rating method will be given and explained.

[0014] For example, as disclosed in "MSCI (2017) MSCI ESG Research ESG Rating Methodology Summary", in the conventional ESG rating, first, key issues are selected and weighted in each industry, and for each key issue, risk exposure and risk management are scored for each company. A key issue score is obtained from these two scores.

[0015] Then, in order to conduct a relative evaluation of companies in the industry, the weighted average score of each company is standardized to determine the Industry Adjusted Score, and based on that, the original ESG score is determined.

[0016] In the determination of the ESG score as described above, public information of the company is used as the information source, and annual reports such as sustainability reports are used as public information.

[0017] However, it is not possible to fully evaluate non-financial elements that change every moment, such as the efforts towards the SDGs, from the public information of the company that is issued only once a year. In addition, in many cases, external consulting and the like are involved in the preparation of the company's public information, and it is unclear whether the actual situation is accurately reflected.

[0018] The social value evaluation device 100 in the present embodiment solves the above problems and can accurately evaluate the social value in real time, reflecting the actual state of a company's activities that change every moment.

[0019] For comparison with the social value evaluation method according to this embodiment, FIG. 1 shows an evaluation method according to the prior art. As shown in FIG. 1, in the prior art, a rating company uses publicly available information, for example, published once a year, to conduct an evaluation in a process with unclear details and publishes the evaluation results. The evaluation results are, for example, rating information such as "A" + ".

[0020] FIG. 2 is a diagram showing an overview of the operation of the social value evaluation apparatus 100 according to this embodiment. As shown in FIG. 2, the feature generation module 130 calculates feature quantities related to social value from text information (a plurality of sentences) related to the evaluation of social value read from the text DB 132.

[0021] Subsequently, the evaluation module 120 inputs text information including a plurality of sentences such as news, press releases, and SNS obtained from, for example, a corporate non-financial DB, evaluates the relevance between the text information and the feature quantities generated by the feature generation module 130, and outputs an evaluation result. This evaluation result is the evaluation result of the social value of the target company.

[0022] For example, it is considered that the social value of a company that issues text information with a high relevance to the feature quantities obtained from text information describing the goals related to social value as its daily activity content in press releases, news, etc. is high.

[0023] Then, the correlation calculation unit 140 inputs a plurality of indicators related to corporate finance (e.g., sales, profit, PBR, ROE, stock price, etc.) from the financial DB 150, calculates the correlation between these indicators (financial information) and the evaluation result of social value obtained by the evaluation module 120, and outputs the calculation result.

[0024] Note that each of the evaluation result by the evaluation module 120 and the financial information may be output, for example, in time series so that the correlation between the two can be grasped visually.

[0025] Figure 3 shows an image of the output by the social value evaluation device 100. In the example shown in Figure 3, it has "social value" and "financial value" as evaluation axes, and the time-series changes of "social value 1", "social value 2", and "financial value" are shown respectively. With such output, "social value 1" and "social value 2" can be grasped, and the correlation between "social value 1" / "social value 2" and "financial value" can be grasped.

[0026] Hereinafter, as an example, the configuration and operation examples of the social value evaluation device 100 will be described in more detail.

[0027] (Configuration example of the social value evaluation device 100) First, a configuration example of the social value evaluation device 100 will be described. Figure 4 shows a detailed configuration example of the social value evaluation device 100. As shown in Figure 4, the social value evaluation device 100 includes an input unit 110, an evaluation module 120, a feature generation module 130, a correlation calculation unit 140, a financial DB 150, and an output unit 160.

[0028] The evaluation module 120 includes a text analysis unit 121 and an evaluation unit 122. Further, the feature generation module 130 includes a feature storage unit 131, a text DB 132, and a feature calculation unit 133.

[0029] The general operation outlines of the evaluation module 120, the feature generation module 130, and the correlation calculation unit 140 are as described with reference to Figure 2. The detailed operations of each part constituting the social value evaluation device 100 will be described in the operation example section below.

[0030] <Hardware configuration example> The social value evaluation device 100 can be realized, for example, by causing a computer to execute a program. This computer may be a physical computer or a virtual machine on the cloud.

[0031] That is, the social value evaluation device 100 can be realized by using hardware resources such as a CPU and a memory built into a computer to execute a program corresponding to the processing performed by the social value evaluation device 100. The above program can be recorded on a computer-readable recording medium (such as a portable memory), saved, or distributed. Further, it is also possible to provide the above program through a network such as the Internet or e-mail.

[0032] FIG. 5 is a diagram showing an example of the hardware configuration of the above computer. The computer in FIG. 5 includes a drive device 1000, an auxiliary storage device 1002, a memory device 1003, a CPU 1004, an interface device 1005, a display device 1006, an input device 1007, an output device 1008, etc., which are mutually connected by a bus BS.

[0033] A program for realizing the processing on the computer is provided, for example, by a recording medium 1001 such as a CD-ROM or a memory card. When the recording medium 1001 storing the program is set in the drive device 1000, the program is installed from the recording medium 1001 to the auxiliary storage device 1002 via the drive device 1000. However, the installation of the program does not necessarily have to be performed from the recording medium 1001, and it may be downloaded from another computer via a network. The auxiliary storage device 1002 stores the installed program and also stores necessary files, data, etc.

[0034] When there is an instruction to start a program, the memory device 1003 reads and stores the program from the auxiliary storage device 1002. The CPU 1004 realizes the functions related to the social value evaluation device 100 according to the program stored in the memory device 1003. The interface device 1005 is used as an interface for connecting to a network and functions as a transmission unit and a reception unit. The display device 1006 displays a GUI (Graphical User Interface) etc. according to the program. The input device 1007 is composed of a keyboard, a mouse, buttons, or a touch panel etc. and is used to input various operation instructions. The output device 1008 outputs the calculation result.

[0035] (Operation example of the social value evaluation device 100) Next, an operation example of the social value evaluation device 100 will be described. Hereinafter, it is assumed that the social value of a certain one company (referred to as "evaluation target company") is evaluated.

[0036] In the present embodiment, first, feature quantities are generated from text information related to the evaluation of social value and stored. This stage is called the feature quantity generation phase. Next, using the stored feature quantities, the social value of the evaluation target company is evaluated. This stage is called the evaluation phase. Hereinafter, each of the feature quantity generation phase and the evaluation phase will be described.

[0037] (Feature quantity generation phase) In the text DB 132 in the feature generation module 130, for example, a plurality of sentences representing the goals of activities for enhancing social value are stored. Specifically, for example, the 169 target sentences of the SDGs are stored. Note that the targets and goals may also be referred to as indicators of social value.

[0038] The feature calculation unit 133 inputs a plurality of sentences read from the text DB 132 and performs morphological analysis on each sentence. By morphological analysis, keywords and the like can be obtained from the input sentences. Any technique can be used for morphological analysis. For example, natural language processing techniques such as the Tf-idf method, co-occurrence degree analysis, and dependency analysis, as well as text mining techniques, can be utilized. Also, morphological analysis tools such as Mecab, JUMAN, and ChaSen may be used.

[0039] Note that instead of extracting keywords from text information as described above, arbitrary keywords may be set subjectively (manually).

[0040] In this embodiment, 109 feature quantities composed of a plurality of keywords are generated by the feature calculation unit 133 from 169 target sentences of the SDGs. Then, the feature calculation unit 133 generates a feature quantity composed of vectors from the feature quantities composed of a plurality of keywords (which may be called "targets") using pre-trained word embedding vectors such as word2Vec, GloVe, and fastText. At that time, averaging, normalization, etc. between multiple keywords may be appropriately performed.

[0041] FIG. 6 shows an image of the case of generating a vector of a feature quantity using word2Vec. In the example shown in FIG. 6, a feature quantity (a vector such as 200 dimensions) is obtained by word2Vec from one target feature quantity composed of three keywords: "safety", "security", and "health".

[0042] The feature quantities obtained by the feature calculation unit 133 are stored in the feature storage unit 131. For example, in the feature storage 131, for each target, the target feature quantity (a plurality of keywords) and the feature quantity (vector) obtained therefrom are stored.

[0043] (Evaluation phase) Next, an operation example of the social value evaluation device 100 in the evaluation phase will be described according to the procedure shown in the flowchart of FIG. 7. Note that the order of the processes shown in FIG. 7 is just an example, and the processes may be performed in any order as long as the evaluation results can be calculated. Also, a plurality of processes may be executed in parallel.

[0044] <s101> In S101 (step 101), the input unit 110 inputs text information about the enterprise to be evaluated. The text information is information obtained in real time about the enterprise to be evaluated. The text information may be any text information regarding the enterprise to be evaluated, such as press, news releases, SNS, etc. In this embodiment, it is assumed that a news release of the enterprise to be evaluated provided by a PR company is input.

[0045] <s102> In S102, the text analysis unit 121 in the evaluation module 120 performs text analysis on the text information (which may be referred to as "sentence" or "document") input in S101.

[0046] Specifically, for example, in the same manner as the method described in the "feature quantity generation phase", the text analysis unit 121 performs morphological analysis of the input text, and generates feature quantities for one or more keywords obtained by the morphological analysis using word embedding vectors or the like.

[0047] <s103> In S103, the evaluation unit 122 calculates the relevance (specifically, similarity) between the feature amount obtained by the text analysis unit 121 and the feature amount read from the feature storage unit 131.

[0048] For example, assuming that 109 feature amounts (vectors) corresponding to 109 targets are stored in the feature storage unit 131, the evaluation unit 122 calculates the similarity between each of the 109 feature amounts and the feature amount obtained by the text analysis unit 121.

[0049] In calculating the similarity, any method can be used as long as it can calculate the similarity between two pieces of information. For example, the cosine similarity can be used. When using the cosine similarity, the similarity between the feature amount x and the feature amount y can be calculated by the following formula.

[0050] cos(x,y)=x·y / |x|×|y| For example, the evaluation unit 122 can extract any number of keywords with particularly high similarity from the input text (news release) for each feature amount stored in the feature storage unit 131, and use the average of the similarities of that number, the normalized value, etc. as the similarity of the calculation result.

[0051] For example, the evaluation unit 122 may extract 10 keywords. In this case, if the feature amounts of each of the 10 keywords are feature amount 1, feature amount 2,...., feature amount 9, feature amount 10, and the feature amount corresponding to a specific target A stored in the feature storage unit 131 is feature amount A, the evaluation unit 122 calculates the similarity 1 between feature amount 1 and feature amount A, the similarity 2 between feature amount 2 and feature amount A,...., the similarity 9 between feature amount 9 and feature amount A, and the similarity 10 between feature amount 10 and feature amount A.

[0052] For example, the evaluation unit 122 calculates the average value, minimum value, and maximum value of the 10 similarities 1 to 10 for target A, and outputs these as the calculation result of the similarity regarding target A. Note that such a calculation method is just an example.

[0053] <s104> In S104, the correlation calculation unit 140 calculates the correlation between the evaluation result obtained by the evaluation unit 122 and the financial information of the enterprise to be evaluated read from the financial DB 150.

[0054] The financial DB 150 stores information such as industry type, sales amount, stock price, ROE, and PBR for various enterprises. For the sales amount, stock price, ROE, PBR, etc., information from the past to the present is stored as time-series data, and the most up-to-date information is always stored. FIG. 8 shows an example of the information stored in the financial DB 50.

[0055] Regarding the calculation of the correlation, for example, the correlation coefficient between the similarity, which is the evaluation result of social value, and the financial information may be calculated. Through the analysis of the correlation, for example, when the similarity to a certain target is high, a correlation such as a high stock price can be found.

[0056] <s105> The output unit 160 outputs the evaluation result. The output of the evaluation result may be, for example, graphically displayed on the screen of the UI (user interface), or may be the output of a list of numerical values. When outputting a list of numerical values, another device may perform a graphical display from the numerical values.

[0057] The evaluation result to be output may be the similarity (e.g., similarity for each target) calculated by the evaluation module 120, or may be the similarity and the financial information obtained from the financial DB 150, or may be the correlation value calculated by the correlation calculation unit 140, or may be the similarity, the correlation, and the financial information, or may be other information.

[0058] Also, at the time of output, the information may be output after being summarized. For example, 109 targets may be summarized and output to 17 SDGs goals.

[0059] Output examples are shown in FIGS. 10 to 13. Note that these show a part of the output screen for the convenience of illustration in the drawings.

[0060] FIG. 9 is an example in which 109 targets are summarized into 17 SDGs goals, and the average value (Avg), maximum value (Max), and minimum value (Min) of the similarity (multiple similarities corresponding to multiple keywords) for each goal are displayed.

[0061] 1 to 17 are the SDGs goals. For example, as shown as the provisional translation of the indicators of the Ministry of Internal Affairs and Communications, "Goal 1: End poverty in all its forms everywhere", etc. At the time of output in the social value evaluation device 100, instead of the full text of the goal, the title of the goal, such as "1: Eliminate poverty", may be displayed. Although the goal numbers are described in FIG. 9, the goal titles such as "1: Eliminate poverty" may be displayed for each goal number.

[0062] In the example of the left figure in Fig. 9, the average value (Avg), maximum value (Max), and minimum value (Min) for each goal are shown in a table-like format. However, the density or color of each column in this table may be changed according to the magnitude of the value. For example, it may be set that the higher the value, the darker the color.

[0063] Also, in the example of the right figure in Fig. 9, at the position corresponding to each goal (the position in the height direction), the average value is shown by a circle at the horizontal position. Note that the color used to fill the circle for each goal may be changed.

[0064] In the example of Fig. 10, for each target number shown on the horizontal axis, the ratio of values at multiple similarities is displayed. In Fig. 10, the range of the ratio is distinguished by hatching, but it may also be distinguished by color differences. Also, in the example of Fig. 10, on the horizontal axis, the goal numbers corresponding to multiple target numbers are shown. In the example of Fig. 11, the annual change in the evaluation results for each target is shown. In the example of Fig. 12, the evaluation results and financial information are displayed in a time series.

[0065] (Effects of the Embodiment) As described above, with the social value evaluation device 100 according to the present embodiment, it is possible to evaluate the social value of a company with respect to various evaluation axes from text information distributed daily such as press, news releases, and SNS, regardless of public information such as the company's annual report.

[0066] Also, various evaluation axes can be set, and for example, it becomes possible to evaluate the specific progress status of social value such as the implementation of SDGs.

[0067] Also, since the evaluation is performed by inputting text information related to the company, a real-time evaluation in line with the actual state of the company's activities becomes possible. By enabling such real-time evaluation, correlation analysis with financial information updated daily such as stock prices becomes possible. Furthermore, it becomes possible to evaluate and analyze the efforts of each industry type and scale of the company.

[0068] (Summary of the Embodiment) This specification discloses a social value evaluation device, a social value evaluation method, and a program, at least for each of the following items. (Item 1) A feature quantity generation unit that generates feature quantities from text information related to the evaluation of social value, An input unit that inputs text information regarding the evaluation target, An evaluation unit that evaluates the relevance between the text information input by the input unit and the feature quantity, An output unit that outputs the evaluation result by the evaluation unit A social value evaluation device comprising the above. (Item 2) The evaluation unit evaluates the relevance by calculating the similarity between the feature quantity generated from the text information input by the input unit and the feature quantity generated by the feature quantity generation unit The social value evaluation device according to Item 1. (Item 3) A correlation calculation unit that calculates the correlation between the evaluation result by the evaluation unit and the financial information of the evaluation target The social value evaluation device according to Item 1 or Item 2, further comprising the above. (Item 4) The output unit outputs information indicating the evaluation result by the evaluation unit for each index of social value The social value evaluation device according to any one of Items 1 to 3. (Item 5) The output unit outputs information showing the evaluation result by the evaluation unit and the financial information of the evaluation target in time series The social value evaluation device according to any one of Items 1 to 4. (Item 6) A social value evaluation method executed by a social value evaluation device, comprising: A feature quantity generation step of generating feature quantities from text information related to the evaluation of social value, An input step of inputting text information regarding the evaluation target, An evaluation step of evaluating the relevance between the text information input by the input step and the feature quantity, An output step for outputting the evaluation result by the evaluation step A social value evaluation method comprising the above. (Item 7) A program for causing a computer to function as each part in the social value evaluation device according to any one of Items 1 to 5.

[0069] As described above, the present embodiment has been described, but the present invention is not limited to such a specific embodiment, and various modifications and changes are possible within the scope of the gist of the present invention described in the claims.

Explanation of Reference Numerals

[0070] 100 Social value evaluation device 110 Input unit 120 Evaluation module 121 Text analysis unit 122 Evaluation unit 130 Feature generation module 131 Feature storage unit 132 Text DB 133 Feature calculation unit 133 140 Correlation calculation unit 150 Financial DB 160 Output unit 1000 Drive device 1001 Recording medium 1002 Auxiliary storage device 1003 Memory device 1004 CPU 1005 Interface device 1006 Display device 1007 Input device

Claims

1. A feature quantity generation unit that generates a target feature quantity composed of a plurality of keywords from a plurality of target sentences of SDGs, which are a plurality of sentences representing goals for activities to enhance social value, and generates a feature quantity composed of vectors from the target feature quantity; An input unit that inputs text information regarding an evaluation target; An evaluation unit that evaluates the relevance between the text information input by the input unit and the feature quantity; An output unit that outputs the evaluation result by the evaluation unit A social value evaluation device comprising the above.

2. The evaluation unit evaluates the relevance by calculating the similarity between the feature quantity generated from the text information input by the input unit and the feature quantity generated by the feature quantity generation unit The social value evaluation device according to Claim 1.

3. A correlation calculation unit that calculates the correlation between the evaluation result by the evaluation unit and the financial information of the evaluation target The social value evaluation device according to Claim 1 or 2, further comprising the above.

4. The output unit outputs information indicating the evaluation result by the evaluation unit for each indicator of social value The social value evaluation device according to any one of Claims 1 to 3.

5. The output unit outputs information showing the evaluation result by the evaluation unit and the financial information of the evaluation target in time series The social value evaluation device according to any one of Claims 1 to 4.

6. A social value evaluation method executed by a social value evaluation device, comprising: A feature quantity generation step of generating a target feature quantity composed of a plurality of keywords from a plurality of target sentences of SDGs, which are a plurality of sentences representing goals for activities to enhance social value, and generating a feature quantity composed of vectors from the target feature quantity; An input step of inputting text information regarding an evaluation target; An evaluation step of evaluating the relevance between the text information input in the input step and the feature quantity; An output step of outputting the evaluation result by the evaluation step A social value evaluation method comprising the above.

7. A program for causing a computer to function as each part in the social value evaluation device according to any one of Claims 1 to 5.

Images