WO2021004550A1 - Esg评分体系的生成方法、装置、设备及可读存储介质 - Google Patents

Esg评分体系的生成方法、装置、设备及可读存储介质 Download PDF

Info

Publication number
WO2021004550A1
WO2021004550A1 PCT/CN2020/107325 CN2020107325W WO2021004550A1 WO 2021004550 A1 WO2021004550 A1 WO 2021004550A1 CN 2020107325 W CN2020107325 W CN 2020107325W WO 2021004550 A1 WO2021004550 A1 WO 2021004550A1
Authority
WO
WIPO (PCT)
Prior art keywords
esg
historical
popularity
current
indicators
Prior art date
Application number
PCT/CN2020/107325
Other languages
English (en)
French (fr)
Inventor
程善钿
李超
伍德意
殷磊
吴海山
Original Assignee
深圳前海微众银行股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳前海微众银行股份有限公司 filed Critical 深圳前海微众银行股份有限公司
Publication of WO2021004550A1 publication Critical patent/WO2021004550A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/3331—Query processing
    • G06F16/334—Query execution
    • G06F16/3344—Query execution using natural language analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/35—Clustering; Classification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/06—Asset management; Financial planning or analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/01—Social networking

Definitions

  • This application relates to the technical field of financial technology (Fintech), and in particular to methods, devices, equipment and readable storage media for generating ESG scoring systems in the financial industry.
  • ESG Environment, Social Responsibility, corporate governance
  • the ESG evaluation system currently used in the market relies on the experience of industry experts and analysts to formulate ESG indicators and weights. This traditional model has the following defects:
  • the data used in ESG analysis comes from the company's active disclosure, and the scoring results depend on the degree of information disclosure
  • the main purpose of this application is to propose a method, device, device, and readable storage medium for generating an ESG scoring system, which aims to solve the problem that the ESG scoring system in the prior art relies too much on human analysis and causes unstable scoring results.
  • the present application provides a method for generating an ESG scoring system.
  • the method for generating an ESG scoring system includes the following steps:
  • the latest ESG score card is obtained.
  • the ESG scoring system generation method before the step of putting historical data into the ESG indicator generator for processing to obtain historical popularity corresponding to the historical indicator, the ESG scoring system generation method includes:
  • the subject extractor is used to iteratively process the secondary indicators and secondary popularity to obtain an ESG indicator generator.
  • the step of classifying data information in the massive corpus according to preset industry standards to obtain the industry corpus includes:
  • the data information in the massive corpus is classified and processed using text analysis technology to obtain the industry corpus.
  • the step of inputting the industry corpus and the first-level seed vocabulary into a topic extractor to obtain the second-level index and second-level popularity includes:
  • the industry corpus and the first-level seed vocabulary are input into a topic extractor for processing to obtain a second-level index and a second-level popularity.
  • the step of obtaining the latest ESG score card based on the historical popularity, current popularity and current indicators includes:
  • the historical weight is adjusted to obtain the latest ESG score card.
  • the method for generating the ESG scoring system includes:
  • the method for generating the ESG scoring system includes:
  • the present application also provides a device for generating an ESG scoring system, the device for generating an ESG scoring system includes:
  • the first processing module is configured to put the historical data and the first seed vocabulary corresponding to the historical data into the ESG indicator generator for processing to obtain the historical popularity corresponding to the historical indicator;
  • the second module is used to put the currently acquired first public opinion data and the second seed vocabulary corresponding to the first public opinion data into the ESG indicator generator for processing to obtain current indicators and current popularity;
  • the scoring module is used to obtain the latest ESG score card based on the historical heat, current heat and current indicators.
  • the present application also provides a device for generating an ESG scoring system.
  • the device for generating an ESG scoring system includes: a memory, a processor, and stored in the memory and running on the processor.
  • the ESG scoring system generation program is executed by the processor to implement the steps of the ESG scoring system generation method described above.
  • the present application also provides a readable storage medium that stores an ESG scoring system generation program, and the ESG scoring system generation program is executed by a processor to achieve the above The steps of the ESG scoring system generation method described.
  • the ESG scoring system generation method proposed in this application puts historical data and the first seed vocabulary corresponding to the historical data into the ESG indicator generator for processing to obtain the historical popularity corresponding to the historical indicator; obtain the current first public opinion data, first 1.
  • Public opinion data can be obtained using AI technology and does not rely on experts and analyst teams; use a wide range of public opinion data to replace company disclosures, and the data is comprehensive and independent; put the first public opinion data and the second seed vocabulary corresponding to the first public opinion data
  • the ESG indicator generator processes to obtain the current indicator and current heat, and obtains the latest ESG score card based on the historical heat, current heat and current indicator, so that the score card can be updated in time and effectively, and the scoring result can achieve the effect of early warning.
  • the method of this application reduces the influence of human factors, and makes the result score more objective, consistent and stable.
  • FIG. 1 is a schematic diagram of a device structure of a hardware operating environment involved in a solution of an embodiment of the present application
  • FIG. 2 is a schematic flowchart of a first embodiment of a method for generating an ESG scoring system of this application
  • FIG. 3 is a flowchart of a first embodiment of a method for generating an ESG scoring system of this application
  • FIG. 4 is a schematic flowchart of a second embodiment of a method for generating an ESG scoring system of this application
  • FIG. 5 is a schematic flowchart of a third embodiment of a method for generating an ESG scoring system of this application.
  • FIG. 1 is a schematic diagram of the device structure of the hardware operating environment involved in the solution of the embodiment of the present application.
  • the generating device of the ESG scoring system in the embodiment of the present application may be a PC or a server device.
  • the device for generating the ESG scoring system may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, and a communication bus 1002.
  • the communication bus 1002 is used to implement connection and communication between these components.
  • the user interface 1003 may include a display screen (Display) and an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface.
  • the network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface).
  • the memory 1005 may be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as a magnetic disk memory.
  • the memory 1005 may also be a storage device independent of the foregoing processor 1001.
  • FIG. 1 does not constitute a limitation on the device, and may include more or fewer components than those shown in the figure, or a combination of certain components, or different component arrangements.
  • the memory 1005 which is a computer-readable storage medium, may include an operating system, a network communication module, a user interface module, and an ESG scoring system generation program.
  • the network interface 1004 is mainly used to connect to a back-end server and communicate with the back-end server;
  • the user interface 1003 is mainly used to connect to a client and communicate with the client;
  • the processor 1001 can be used to Call the ESG scoring system generation program stored in the memory 1005, and execute the operations in each embodiment of the following ESG scoring system generation method.
  • FIG. 2 is a schematic flowchart of a first embodiment of a method for generating an ESG scoring system according to this application. The method includes:
  • Step S10 Put the historical data and the first seed vocabulary corresponding to the historical data into the ESG indicator generator for processing to obtain historical popularity corresponding to the historical indicator;
  • ESG Environmental, Social and governance
  • ESG usually represents the three major factors of environment, society, and corporate governance. It is an important consideration factor in investment decision-making in socially responsible investment.
  • the investment object of socially responsible investment is in addition to corporate fundamentals. In addition to performance, companies that generally perform better under these three factors.
  • the construction process of the ESG indicator generator is: collect data information from the Internet, and build a massive corpus based on the data information; according to preset industry standards, classify and process the data information in the massive corpus to obtain an industry corpus; determine The first-level seed vocabulary corresponding to the industry, the industry corpus and the first-level seed vocabulary are input into the topic extractor for processing to obtain the second-level index and the second-level popularity; Iterative processing is performed on the level of heat to obtain an ESG indicator generator.
  • the first-level indicators are the three dimensions of environment, society and corporate governance.
  • the secondary indicators are 13 classified topics under the environment, society and corporate governance.
  • the secondary indicators under the environment include environmental goals, environmental management, environmental disclosure and negative events.
  • the three-level indicators will cover specific ESG indicators, with a total of 127 three-level indicators.
  • the social three-level indicators include more than 30 indicators such as labor policy, employee policy, female employees, diversification, and supply chain responsibility management.
  • the evaluation system is divided into general indicators and industry-specific indicators. General indicators are applicable to all listed companies, and industry-specific indicators refer to indicators specific to each industry, which are only applicable to companies within this industry classification.
  • the historical indicators in this case refer to the indicators at all levels when constructing the basic ESG score card, for example, N-level indicators f 1 , f 2 ??f n , that is, historical indicators include historical first-level indicators f 1 and historical second-level indicators f Level 2 , the corresponding historical heat, that is, the historical first-degree heat h 1 , the historical second-degree heat h 2 and so on.
  • the N-level index is used as the seed vocabulary, that is, the historical first-level index f 1 corresponds to the first-level seed vocabulary, and the corresponding first-level seed vocabulary exists in the historical data.
  • Popularity refers to the number of occurrences of indicators, the number of occurrences of historical indicators corresponding to historical heat, such as the number of occurrences of historical first-level indicators f 1 in a historical time period.
  • the historical data corresponding to the ESG basic score card can be obtained as The news and announcement data on and before August 1, 2019 are classified by industry.
  • the first seed vocabulary of the historical data chemical industry is "sewage discharge”. Then, put the chemical data and "sewage discharge” before August 2019.
  • the historical heat corresponding to the historical indicator is obtained. Since the historical indicator and its corresponding historical heat will not change anymore, it is possible to build a stable ESG indicator system.
  • Step S20 Put the currently acquired first public opinion data and the second seed vocabulary corresponding to the first public opinion data into the ESG indicator generator for processing to obtain current indicators and current popularity;
  • the first public opinion data and the second seed vocabulary corresponding to the first public opinion data currently acquired are put into the ESG indicator generator for processing, and the current indicators and current indicators are obtained.
  • Public opinion data can be new media data, online public opinion data, government affairs data, public security political law data, financial supervision data, stock and bond market data, business operation data, corporate financial data, business data, etc.
  • the second seed vocabulary is also a seed vocabulary.
  • the second seed vocabulary corresponding to the first public opinion data is distinguished from the first seed vocabulary corresponding to historical data.
  • the scoring system needs to be updated before scoring. Therefore, under the stable ESG indicator system corresponding to historical indicators, the latest first public opinion data and corresponding The second seed vocabulary.
  • the historical data is the chemical industry data before August 1, 2019, which is September 1, 2019.
  • major news websites publish news about chemical engineering, corresponding to more news related to air quality.
  • the first public opinion data is chemical industry data
  • the corresponding second seed vocabulary is air quality. Therefore, in order to update to the latest scoring results, the chemical data between August 1, 2019 and August 31, 2019, that is the first public opinion data, and the second seed vocabulary of air quality corresponding to the chemical data , Put it into the ESG indicator generator for processing to get the current indicator and current popularity.
  • the ESG indicator generator By putting the first public opinion data and the second seed vocabulary corresponding to the first public opinion data into the ESG indicator generator for processing, the current indicators and current popularity are obtained, thereby avoiding the lack of public opinion data and ensuring the accuracy of the scoring.
  • Step S30 Obtain the latest ESG score card based on the historical heat, current heat and current indicators.
  • this step specifically: after obtaining the historical heat, normalize the historical heat to obtain the historical weight; obtain the ESG basic score card based on the historical indicator, historical heat, and historical weight; Put the basic score card into the ESG indicator system library; obtain the historical heat and historical weight based on the ESG indicator system library; adjust the historical weight based on the historical heat and current heat to obtain the latest ESG score card, that is, match ESG indicators
  • the system adjusts the weight of the index according to the historical heat of the index and the heat change of each period, so as to obtain the latest ESG score card.
  • Historical weight refers to the weight of historical indicators.
  • the ESG basic score card is obtained, and then the ESG basic score card is put into the ESG index system library.
  • the ESG index system library stores the time, index, popularity, and weight for calculating historical weights.
  • the historical weights are adjusted according to historical heat, current heat and current indicators, so as to obtain the latest ESG score card, which effectively guarantees the high accuracy of the score card, and the entire process does not require human involvement and is highly intelligent. And the score card can be updated in time and effectively, and the score result can achieve the effect of early warning.
  • the ESG scoring system generation method proposed in this application puts historical data and the first seed vocabulary corresponding to the historical data into the ESG indicator generator for processing to obtain the historical popularity corresponding to the historical indicator; obtain the current first public opinion data, first 1.
  • Public opinion data can be obtained using AI technology and does not rely on experts and analyst teams; use a wide range of public opinion data to replace company disclosures, and the data is comprehensive and independent; put the first public opinion data and the second seed vocabulary corresponding to the first public opinion data
  • the ESG indicator generator processes to obtain the current indicator and current heat, and obtains the latest ESG score card based on the historical heat, current heat and current indicator, so that the score card can be updated in time and effectively, and the scoring result can achieve the effect of early warning.
  • the method of this application reduces the influence of human factors, and makes the result score more objective, consistent and stable.
  • a second embodiment of the method for generating the ESG scoring system of the present application is proposed; as shown in FIG. 4, before step S10, the method for generating the ESG scoring system can be include:
  • the subject extractor is used to iteratively process the secondary indicators and secondary popularity to obtain an ESG indicator generator.
  • an ESG indicator generator before putting historical data and the corresponding first seed vocabulary into the ESG indicator generator, an ESG indicator generator needs to be constructed.
  • the construction process of the ESG indicator generator is: collecting data information from the network and based on all The data information constructs a massive corpus; according to preset industry standards, the data information in the massive corpus is classified and processed to obtain an industry corpus; the industry corresponding first-level seed vocabulary is determined, and the industry corpus and the first-level seed
  • the vocabulary is input and processed in the topic extractor to obtain the secondary index and the secondary popularity; the topic extractor is used to iteratively process the secondary indicator and the secondary popularity to obtain an ESG indicator generator.
  • the network can be a government agency website, or a major official media website, and it can use data crawlers, or ask government agencies’ websites to actively provide data to obtain data information.
  • the construction of ESG indicator generator is 2019 On June 1, get the data and information at that time, for example, get a report on the chemical industry.
  • the data information in the massive corpus is classified and processed, and the industry corpus is obtained, for example, financial Industry corpus, technology industry corpus, transportation industry corpus, etc.
  • the industry corpus is obtained, for example, financial Industry corpus, technology industry corpus, transportation industry corpus, etc.
  • first-level seed vocabulary For example, for the chemical industry, sewage discharge can be used as a first-level seed vocabulary.
  • the industry corpus and the first-level seed vocabulary into the AI technology topic extractor to get the second-level indicators and their popularity.
  • the AI technology topic extractor can be used iteratively, using N-level indicators as seed vocabulary, and N+1-level indicators and their popularity can be obtained through the extractor, thereby obtaining an ESG indicator generator.
  • an ESG indicator generator By constructing an ESG indicator generator to facilitate the processing of historical data and the corresponding first seed vocabulary to obtain a stable ESG indicator system, and to facilitate the comparison of the currently acquired first public opinion data and the corresponding first public opinion data
  • the second seed vocabulary is processed to obtain the latest index and current popularity.
  • to classify the data information in the massive corpus can use text analysis technology to classify the data information in the massive corpus to obtain an industry corpus.
  • the text analysis method refers to going from the surface layer of the text to the deep layer of the text, so as to discover the deep meanings that cannot be grasped by ordinary reading.
  • the data information in the massive corpus is classified and processed to obtain the industry corpus, for example, the obtained industry corpus such as the financial industry corpus, the technology industry corpus, and the transportation industry corpus.
  • the obtained industry corpus such as the financial industry corpus, the technology industry corpus, and the transportation industry corpus.
  • the step of inputting the industry corpus and the first-level seed vocabulary into the topic extractor for processing to obtain the second-level index and the second-level popularity may include:
  • the industry corpus and the first-level seed vocabulary are input into a topic extractor for processing to obtain a second-level index and a second-level popularity.
  • Natural Language Processing is a sub-field of artificial intelligence (AI).
  • AI artificial intelligence
  • Deep learning methods have the ability to learn feature representations. Experts are required to manually specify and extract features from natural language, and to continuously and rapidly improve on challenging problems.
  • Topic Model is a modeling method for the hidden topics in the text. First, you need to define what the topic is. The topic is a concept and an aspect, which is expressed as a series of related words, which are words on the vocabulary. The conditional probability distribution of the word, the more closely related to the topic, the greater its conditional probability, and vice versa.
  • the first-level indicator can be ESG
  • the second-level indicators can get environment, society, and corporate governance through the theme model, and then put environment as a keyword to get the third-level indicators, water quality, air quality, etc.
  • the topic extractor is determined to realize the intelligent processing of the industry corpus and the first-level seed vocabulary, thereby obtaining the second-level indicators and the second-level popularity.
  • step S30 may include:
  • the historical weight is adjusted to obtain the latest ESG score card.
  • obtaining the latest ESG score card based on the historical heat, current heat, and current indicators specifically includes: obtaining the historical heat and historical weight based on the ESG index system library; based on the historical heat and current heat, Adjust the historical weight to obtain the latest ESG score card.
  • the ESG indicator system library is equivalent to a database for storing historical indicators, historical heat, historical weights, and even current time.
  • the method for generating the ESG scoring system includes:
  • the ESG basic score card is obtained, and then the ESG basic score card is put into the ESG index system library.
  • the ESG index system library stores the time, index, popularity, and weight for calculating historical weights.
  • a third embodiment of the method for generating the ESG scoring system of the present application is proposed; as shown in FIG. 5, after step S30, the method for generating the ESG scoring system Can include:
  • ESG scoring of a single company can be performed, specifically: obtaining the current second public opinion data of the enterprise; obtaining the historical indicators based on the ESG indicator system database; The language processing technology obtains the ESG score of the enterprise based on the second public opinion data, historical indicators, and the latest ESG score card; visualizes the ESG score to obtain a visual graph.
  • NLP technology natural language processing technology
  • match the indicators in the ESG index system library match the latest scorecard weight to obtain the ESG score of a single company.
  • This application also provides a device for generating an ESG scoring system.
  • the generating device of the ESG scoring system described in this application includes:
  • the first processing module is configured to put the historical data and the first seed vocabulary corresponding to the historical data into the ESG indicator generator for processing to obtain the historical popularity corresponding to the historical indicator;
  • the second module is used to put the currently acquired first public opinion data and the second seed vocabulary corresponding to the first public opinion data into the ESG indicator generator for processing to obtain current indicators and current popularity;
  • the scoring module is used to obtain the latest ESG score card based on the historical heat, current heat and current indicators.
  • the ESG scoring system generation device also includes a building module for:
  • the subject extractor is used to iteratively process the secondary indicators and secondary popularity to obtain an ESG indicator generator.
  • building module is also used for:
  • the data information in the massive corpus is classified and processed using text analysis technology to obtain the industry corpus.
  • building module is also used to:
  • the industry corpus and the first-level seed vocabulary are input into a topic extractor for processing to obtain a second-level index and a second-level popularity.
  • scoring module is also used for:
  • the historical weight is adjusted to obtain the latest ESG score card.
  • the scoring module is further used to:
  • the ESG scoring system generation device also includes a scoring module for:
  • the application also provides a computer-readable storage medium.
  • the computer-readable storage medium of the present application stores an ESG scoring system generation program, and when the ESG scoring system generation program is executed by a processor, the steps of the aforementioned ESG scoring system generation method are realized.
  • the method implemented when the ESG scoring system generation program running on the processor is executed can refer to the various embodiments of the ESG scoring system generation method of the present application, which will not be repeated here.

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Development Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Databases & Information Systems (AREA)
  • Finance (AREA)
  • Game Theory and Decision Science (AREA)
  • Accounting & Taxation (AREA)
  • Educational Administration (AREA)
  • Operations Research (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Quality & Reliability (AREA)
  • Computing Systems (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Artificial Intelligence (AREA)
  • Computational Linguistics (AREA)
  • Technology Law (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种ESG评分体系的生成方法、装置、设备及可读存储介质，涉及金融科技领域，所述方法包括：将历史数据及第一种子词汇放入ESG指标生成器处理，得到历史指标对应的历史热度(S10)；将当前获取的第一舆情数据及第二种子词汇放入所述ESG指标生成器处理，得到当前指标及当前热度(S20)；基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡(S30)。

Description

ESG评分体系的生成方法、装置、设备及可读存储介质
优先权信息
本申请要求于2019年9月4日申请的、申请号为201910835812.0、名称为“ESG评分体系的生成方法、装置、设备及可读存储介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及金融科技(Fintech)技术领域，尤其涉及金融行业的ESG评分体系的生成方法、装置、设备及可读存储介质。
背景技术
随着计算机技术的发展，越来越多的技术(如分布式、区块链Blockchain、人工智能等)应用在金融领域，传统金融业正在逐步向金融科技(Fintech)转变，越来越多的技术应用于金融行业。金融行业通常会用到ESG评价体系，ESG(Environment、Social Responsibility、Corporate Governance)，即环境，社会和公司治理，是指衡量公司或企业投资的可持续性和道德影响的三个核心因素。目前市场应用的ESG评价体系，都依赖于行业专家和分析师的经验制定ESG指标和权重，这种传统的模式有以下几个缺陷：
1、ESG分析所用的数据来自公司的主动披露，评分结果依赖信息披露程度；
2、ESG分析所用的数据有滞后性，评分不能及时反映公司当下的状态；
3、不同的ESG分析机构或者同一ESG分析机构在不同时间段，使用的评分指标集不同，相同的指标也会有不同的定义，从而导致同一家公司会有不一致、不稳定的评分结果。
发明内容
本申请的主要目的在于提出一种ESG评分体系的生成方法、装置、设备及可读存储介质，旨在解决现有技术中ESG评分体系过于依赖人为分析而导致评分结果不稳定的问题。
为实现上述目的，本申请提供一种ESG评分体系的生成方法，所述ESG评分体系的生成方法包括如下步骤：
将历史数据及所述历史数据对应的第一种子词汇放入ESG指标生成器处理，得到历史指标对应的历史热度；
将当前获取的第一舆情数据及所述第一舆情数据对应的第二种子词汇放入所述ESG指标生成器处理，得到当前指标及当前热度；
基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡。
在一实施例中，所述将历史数据放入ESG指标生成器处理，得到历史指标对应的历史热度的步骤之前，所述ESG评分体系的生成方法包括：
从网络上采集数据信息，且基于所述数据信息构建海量语料库；
根据预设行业标准，对所述海量语料库中的数据信息进行分类处理，得到行业语料库；
确定行业对应的一级种子词汇，将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度；
通过所述主题提取器对所述二级指标及二级热度进行迭代处理，得到ESG指标生成器。
在一实施例中，所述根据预设行业标准，对所述海量语料库中的数据信息进行分类处理，得到行业语料库的步骤包括：
根据预设行业标准，利用文本分析技术对所述海量语料库中的数据信息进行分类处理，得到行业语料库。
在一实施例中，所述将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度的步骤包括：
利用自然语言处理技术及主题模型，确定主题提取器；
将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度。
在一实施例中，所述基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡的步骤包括：
基于ESG指标体系库，获取所述历史热度及历史权重；
基于所述历史热度、当前热度，调节所述历史权重，得到ESG最新评分卡。
在一实施例中，所述基于ESG指标体系库，获取所述历史热度及历史权重的步骤之前，所述ESG评分体系的生成方法包括：
将所述历史热度进行归一化处理，得到历史权重；
基于所述历史指标、历史热度及历史权重，得到ESG基础评分卡；
将所述ESG基础评分卡放入ESG指标体系库。
在一实施例中，所述基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡的步骤之后，所述ESG评分体系的生成方法包括：
获取企业当前的第二舆情数据；
基于所述ESG指标体系库，获取所述历史指标；
通过自然语言处理技术，基于所述第二舆情数据、历史指标及所述ESG最新评分卡，得到所述企业的ESG得分；
将所述ESG得分进行可视化处理。
此外，为实现上述目的，本申请还提供一种ESG评分体系的生成装置，所述ESG评分体系的生成装置包括：
第一处理模块，用于将历史数据及所述历史数据对应的第一种子词汇放入ESG指标生成器处理，得到历史指标对应的历史热度；
第二模块，用于将当前获取的第一舆情数据及所述第一舆情数据对应的第二种子词汇放入所述ESG指标生成器处理，得到当前指标及当前热度；
评分模块，用于基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡。
此外，为实现上述目的，本申请还提供一种ESG评分体系的生成设备，所述ESG评分体系的生成设备包括：存储器、处理器及存储在所述存储器上并可在所述处理器上运行的ESG评分体系的生成程序，所述ESG评分体系的生成程序被所述处理器执行时实现如上所述的ESG评分体系的生成方法的步骤。
此外，为实现上述目的，本申请还提供一种可读存储介质，所述可读存储介质上存储有ESG评分体系的生成程序，所述ESG评分体系的生成程序被处理器执行时实现如上所述的ESG评分体系的生成方法的步骤。
本申请提出的ESG评分体系的生成方法，将历史数据及所述历史数据对应的第一种子词汇放入ESG指标生成器处理，得到历史指标对应的历史热度； 获取当前的第一舆情数据，第一舆情数据可以利用AI技术获得，不依赖于专家和分析师团队；使用广泛的舆情数据代替公司披露，数据全面而且独立；将第一舆情数据及第一舆情数据对应的第二种子词汇放入所述ESG指标生成器处理，得到当前指标及当前热度，并基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡，使得评分卡能够及时有效更新，评分结果可以达到预警的效果。本申请的方法，减少了人为因素的影响，使得结评分果更客观、一致及稳定。
附图说明
图1是本申请实施例方案涉及的硬件运行环境的设备结构示意图；
图2为本申请ESG评分体系的生成方法第一实施例的流程示意图；
图3为本申请ESG评分体系的生成方法第一实施例的流程框图；
图4为本申请ESG评分体系的生成方法第二实施例的流程示意图；
图5为本申请ESG评分体系的生成方法第三实施例的流程示意图。
本申请目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。
具体实施方式
应当理解，此处所描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。
如图1所示，图1是本申请实施例方案涉及的硬件运行环境的设备结构示意图。
本申请实施例ESG评分体系的生成设备可以是PC机或服务器设备。
如图1所示，该ESG评分体系的生成设备可以包括：处理器1001，例如CPU，网络接口1004，用户接口1003，存储器1005，通信总线1002。其中，通信总线1002用于实现这些组件之间的连接通信。用户接口1003可以包括显示屏(Display)、输入单元比如键盘(Keyboard)，可选用户接口1003还可以包括标准的有线接口、无线接口。网络接口1004可选的可以包括标准的有线接口、无线接口(如WI-FI接口)。存储器1005可以是高速RAM存储器，也可以是稳定的存储器(non-volatile memory)，例如磁盘存储器。存储器1005可选的还可以是独立于前述处理器1001的存储装置。
本领域技术人员可以理解，图1中示出的设备结构并不构成对设备的限定，可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件布置。
如图1所示，作为一种计算机可读存储介质的存储器1005中可以包括操作系统、网络通信模块、用户接口模块以及ESG评分体系的生成程序。
在图1所示的设备中，网络接口1004主要用于连接后台服务器，与后台服务器进行数据通信；用户接口1003主要用于连接客户端，与客户端进行数据通信；而处理器1001可以用于调用存储器1005中存储的ESG评分体系的生成程序，并执行下述ESG评分体系的生成方法各个实施例中的操作。
基于上述硬件结构，提出本申请ESG评分体系的生成方法实施例。
参照图2，图2为本申请ESG评分体系的生成方法第一实施例的流程示意图，所述方法包括：
步骤S10，将历史数据及所述历史数据对应的第一种子词汇放入ESG指标生成器处理，得到历史指标对应的历史热度；
本实施例中，ESG(Environmental,Social and Governance)通常代表环境、社会和公司治理三大因素，是社会责任投资中投资决策的重要考量因子，而社会责任投资选择的投资对象就是除了企业基本面表现外，一般在这三大因素下表现更为优秀的企业。
ESG指标生成器的构建过程为：从网络上采集数据信息，且基于所述数据信息构建海量语料库；根据预设行业标准，对所述海量语料库中的数据信息进行分类处理，得到行业语料库；确定行业对应的一级种子词汇，将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度；通过所述主题提取器对所述二级指标及二级热度进行迭代处理，得到ESG指标生成器。
通过将历史数据及历史数据对应的第一种子词汇放入ESG指标生成器中，得到历史指标对应的历史热度，在ESG评分体系中，一级指标为环境、社会和公司治理三个维度。二级指标为环境、社会和公司治理下的13项分类议题，如环境下的二级指标包括环境目标、环境管理、环境披露及负面事件等。三级指标将会涵盖具体的ESG指标，共有127项三级指标，例如社会方面的三级指标包括劳工政策、员工政策、女性员工、多样化、供应链责任管 理等30多项指标。评估体系分为通用指标和行业特定指标。通用指标适用于所有上市公司，行业特定指标是指各行业特有的指标，只适用于本行业分类内的公司。
本案中的历史指标，是指构建ESG基础评分卡时的各级指标，例如，N级指标f 1，f 2……f n，即历史指标包括历史一级指标f 1，历史二级指标f 2等，对应的历史热度，即历史一级热度h 1，历史二级热度h 2等。并且，将N级指标作为种子词汇，即，历史一级指标f 1对应一级种子词汇，且历史数据中存在对应的一级种子词汇。
热度是指指标出现次数，历史热度对应的历史指标出现的次数，如历史一级指标f 1在历史时间段出现的次数。
例如，在构建ESG基础评分卡的时间是2019年8月1日，采用数据爬虫或者是政府机构网站提供的方式，获取到一些公告、新闻，即可得到构建ESG基础评分卡对应的历史数据为2019年8月1日及以前的新闻和公告数据，按行业分类，比如历史数据化工的第一种子词汇为“污水排放”，那么，将2019年8月以前的化工数据及“污水排放”放入ESG指标生成器处理，得到历史指标对应的历史热度，即污水排放的热度。
通过将历史数据及历史数据对应的第一种子词汇放入ESG指标生成器中，得到历史指标对应的历史热度，由于历史指标及其对应的历史热度，不会再改变，因此，能够构建一个稳定的ESG指标体系。
步骤S20，将当前获取的第一舆情数据及所述第一舆情数据对应的第二种子词汇放入所述ESG指标生成器处理，得到当前指标及当前热度；
该步骤中，将当前(例如每期中的天、周、月、季度等)获取的第一舆情数据及第一舆情数据对应的第二种子词汇放入ESG指标生成器处理，得到当前指标及当前热度。舆情数据，可以是新媒体数据、网络舆情数据、政务数据、公安政法数据、金融监管数据、股债行情数据、企业经营数据、企业财务数据、工商数据等，第二种子词汇，也是种子词汇，为第一舆情数据对应的第二种子词汇，与历史数据对应的第一种子词汇区别开。
为了保证得到的指标及热度是最新的，在评分前，需要对评分体系进行更新，因此，在得到历史指标对应的历史热度这一稳定的ESG指标体系下，获取最新的第一舆情数据及对应的第二种子词汇。
比如，历史数据是2019年8月1日之前的化工数据，此时为2019年9月1号，然后，各大新闻网站刊登关于化工方面的新闻，对应更多的新闻与空气质量相关，此时，第一舆情数据为化工数据，对应的第二种子词汇为空气质量。因此，为了更新到最新的评分结果，将2019年8月1日至2019年8月31日之间的化工数据，即第一舆情数据，以及，化工数据对应的空气质量这一第二种子词汇，放入到ESG指标生成器处理，从而得到当前指标及当前热度。
通过将第一舆情数据及所述第一舆情数据对应的第二种子词汇放入所述ESG指标生成器处理，得到当前指标及当前热度，避免舆情数据缺失，保证评分的准确度。
步骤S30，基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡。
该步骤中，具体为：在得到历史热度后，将所述历史热度进行归一化处理，得到历史权重；基于所述历史指标、历史热度及历史权重，得到ESG基础评分卡；将所述ESG基础评分卡放入ESG指标体系库；基于ESG指标体系库，获取所述历史热度及历史权重；基于所述历史热度、当前热度，调节所述历史权重，得到ESG最新评分卡，即匹配ESG指标体系，根据指标历史热度和每期的热度变化、调节指标的权重，如此获取最新ESG评分卡。
历史权重，是指历史指标的权重。
例如N级指标f 1，f 2……f n及历史热度h 1，h 2……h n，对于每一个i＝1,……,n，热度归一化计算成历史权重
Figure PCTCN2020107325-appb-000001
从而得到ESG基础评分卡，再将ESG基础评分卡放进ESG指标体系库中，很明显，ESG指标体系库中存储有计算历史权重的时间，指标，热度，权重。
如图3所示，根据历史热度、当前热度及当前指标，对历史权重进行调整，从而得到ESG最新评分卡，有效保证评分卡的高度精准，并且，整个过程无需人为参与，智能化程度高，且评分卡能够及时有效更新，评分结果可以达到预警的效果。
本申请提出的ESG评分体系的生成方法，将历史数据及所述历史数据对应的第一种子词汇放入ESG指标生成器处理，得到历史指标对应的历史热度； 获取当前的第一舆情数据，第一舆情数据可以利用AI技术获得，不依赖于专家和分析师团队；使用广泛的舆情数据代替公司披露，数据全面而且独立；将第一舆情数据及第一舆情数据对应的第二种子词汇放入所述ESG指标生成器处理，得到当前指标及当前热度，并基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡，使得评分卡能够及时有效更新，评分结果可以达到预警的效果。本申请的方法，减少了人为因素的影响，使得结评分果更客观、一致及稳定。
进一步地，基于本申请ESG评分体系的生成方法的第一实施例，提出本申请ESG评分体系的生成方法的第二实施例；如图4所示，步骤S10之前，ESG评分体系的生成方法可以包括：
从网络上采集数据信息，且基于所述数据信息构建海量语料库；
根据预设行业标准，对所述海量语料库中的数据信息进行分类处理，得到行业语料库；
确定行业对应的一级种子词汇，将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度；
通过所述主题提取器对所述二级指标及二级热度进行迭代处理，得到ESG指标生成器。
本实施例中，在将历史数据及对应的第一种子词汇放入ESG指标生成器之前，需要构建ESG指标生成器，ESG指标生成器的构建过程为：从网络上采集数据信息，且基于所述数据信息构建海量语料库；根据预设行业标准，对所述海量语料库中的数据信息进行分类处理，得到行业语料库；确定行业对应的一级种子词汇，将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度；通过所述主题提取器对所述二级指标及二级热度进行迭代处理，得到ESG指标生成器。
具体地，网络可以是政府机构网站，也可以是各大官方媒体网站，可以采用数据爬虫方式，或者是请政府机构网站主动提供的方式，获取数据信息，比如，构建ESG指标生成器是2019年6月1号，则获取当时的数据信息，比如获取对一篇化工方面的报道。
该信息数据的获取，不依赖于公司的主动披露，主动从各大公开网站获得。
在获得数据信息后，基于数据信息构建海量语料库，然后，制定行业分类标准，根据制定的标准，即预设行业标准，对海量语料库中的数据信息进行分类处理，得到行业语料库，例如，得到金融行业语料库、科技行业语料库、交通运输业语料库等。接着，对每一个行业，制定一级种子词汇，例如对于化工行业，污水排放可以作为一级种子词汇。再将行业语料库和一级种子词汇放进AI技术主题提取器中处理，得到二级指标及其热度。
最后，AI技术主题提取器可以迭代使用，将N级指标作为种子词汇，通过提取器可以得到N+1级指标及其热度，从而得到ESG指标生成器。
通过构建ESG指标生成器，以便于对历史数据及对应的第一种子词汇进行处理，得到稳定的ESG指标体系，并且，以便于对当前获取的第一舆情数据及所述第一舆情数据对应的第二种子词汇进行处理，得到最新的指标及当前热度。
进一步地，对所述海量语料库中的数据信息进行分类处理可以利用文本分析技术，对海量语料库中的数据信息进行分类处理，得到行业语料库。
在本实施例中，文本分析法是指从文本的表层深入到文本的深层，从而发现那些不能为普通阅读所把握的深层意义。
通过文本分析技术，对海量语料库中的数据信息进行分类处理，得到行业语料库，例如，得到的行业语料库如金融行业语料库、科技行业语料库、交通运输业语料库等。
进一步地，所述将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度的步骤可以包括：
利用自然语言处理技术及主题模型，确定主题提取器；
将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度。
本实施例中，自然语言处理技术(Natural Language Processing，NLP)是人工智能(AI)的一个子领域，采用深度学习方法有助于开发全新的模型，深度学习方法具有学习特征表示的能力，不必要求专家从自然语言中人工指定和提取特征，并且，在挑战性问题上持续快速改进。
主题模型(Topic Model)，就是对文字中隐含主题的一种建模方法，首先，需要定义主题是什么，主题是一个概念，一个方面，表现为一系列相关 的词语，是词汇表上词语的条件概率分布，与主题关系越密切的词语，它的条件概率越大，反之则越小。
比如，一级指标可以是ESG，二级指标通过主题模型可以得到环境、社会、公司治理，然后把环境作为关键词放进去，得到三级指标、水质、空气质量等。
通过自然语言处理技术及主题模型，确定主题提取器，实现对行业语料库及一级种子词汇的智能处理，从而得到二级指标及二级热度。
进一步地，步骤S30可以包括：
基于ESG指标体系库，获取所述历史热度及历史权重；
基于所述历史热度、当前热度，调节所述历史权重，得到ESG最新评分卡。
在本实施例中，基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡具体为：基于ESG指标体系库，获取所述历史热度及历史权重；基于所述历史热度、当前热度，调节所述历史权重，得到ESG最新评分卡。
其中，ESG指标体系库，相当于一个数据库，用于存储历史指标、历史热度及历史权重，甚至是当前时间。
由于需要参考历史热度，因此，需要从ESG指标体系库中提取出历史热度及历史权重，并根据历史热度、当前热度，对历史权重进行调节，更新历史权重，从而得到ESG最新评分卡。
进一步地，所述基于ESG指标体系库，获取所述历史热度及历史权重的步骤之前，所述ESG评分体系的生成方法包括：
将所述历史热度进行归一化处理，得到历史权重；
基于所述历史指标、历史热度及历史权重，得到ESG基础评分卡；
将所述ESG基础评分卡放入ESG指标体系库。
本实施例中，例如N级指标f 1，f 2……f n及历史热度h 1，h 2……h n，对于每一个i＝1,……,n，热度归一化计算成历史权重
Figure PCTCN2020107325-appb-000002
从而得到ESG基础评分卡，再将ESG基础评分卡放进ESG指标体系库中，很明显，ESG指标体系库中存储有计算历史权重的时间，指标，热度，权重。
进一步地，基于本申请ESG评分体系的生成方法的第二实施例，提出本 申请ESG评分体系的生成方法的第三实施例；如图5所示，在步骤S30之后，ESG评分体系的生成方法可以包括：
获取企业当前的第二舆情数据；
基于所述ESG指标体系库，获取所述历史指标；
通过自然语言处理技术，基于所述第二舆情数据、历史指标及所述ESG最新评分卡，得到所述企业的ESG得分；
将所述ESG得分进行可视化处理。
本实施例中，在得到ESG最新评分卡后，可进行对单个公司的ESG评分，具体为：获取企业当前的第二舆情数据；基于所述ESG指标体系库，获取所述历史指标；通过自然语言处理技术，基于所述第二舆情数据、历史指标及所述ESG最新评分卡，得到所述企业的ESG得分；将所述ESG得分进行可视化处理，得到可视化图形。
对单个公司/企业的评分，首先汇集公司/企业的舆情数据，通过NLP技术(自然语言处理技术)，匹配ESG指标体系库中的指标，结合最新的评分卡权重，得到单个公司的ESG得分。
可视化公司得分，提供预警信号、趋势图和影响得分的关键舆情数据，方便用户直观地认识风险并做出决策。
本申请还提供一种ESG评分体系的生成装置。本申请所述ESG评分体系的生成装置包括：
第一处理模块，用于将历史数据及所述历史数据对应的第一种子词汇放入ESG指标生成器处理，得到历史指标对应的历史热度；
第二模块，用于将当前获取的第一舆情数据及所述第一舆情数据对应的第二种子词汇放入所述ESG指标生成器处理，得到当前指标及当前热度；
评分模块，用于基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡。
进一步地，ESG评分体系的生成装置还包括构建模块，用于：
从网络上采集数据信息，且基于所述数据信息构建海量语料库；
根据预设行业标准，对所述海量语料库中的数据信息进行分类处理，得到行业语料库；
确定行业对应的一级种子词汇，将所述行业语料库及所述一级种子词汇 输入主题提取器中处理，得到二级指标及二级热度；
通过所述主题提取器对所述二级指标及二级热度进行迭代处理，得到ESG指标生成器。
进一步地，所述构建模块还用于：
根据预设行业标准，利用文本分析技术对所述海量语料库中的数据信息进行分类处理，得到行业语料库。
进一步地，构建模块还用于：
利用自然语言处理技术及主题模型，确定主题提取器；
将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度。
进一步地，所述评分模块还用于：
基于ESG指标体系库，获取所述历史热度及历史权重；
基于所述历史热度、当前热度，调节所述历史权重，得到ESG最新评分卡。
进一步地评分模块还用于：
将所述历史热度进行归一化处理，得到历史权重；
基于所述历史指标、历史热度及历史权重，得到ESG基础评分卡；
将所述ESG基础评分卡放入ESG指标体系库。
进一步地，ESG评分体系的生成装置还包括得分模块，用于：
获取企业当前的第二舆情数据；
基于所述ESG指标体系库，获取所述历史指标；
通过自然语言处理技术，基于所述第二舆情数据、历史指标及所述ESG最新评分卡，得到所述企业的ESG得分；
将所述ESG得分进行可视化处理。
本申请还提供一种计算机可读存储介质。
本申请计算机可读存储介质上存储有ESG评分体系的生成程序，所述ESG评分体系的生成程序被处理器执行时实现如上所述的ESG评分体系的生成方法的步骤。
其中，在所述处理器上运行的ESG评分体系的生成程序被执行时所实现的方法可参照本申请ESG评分体系的生成方法各个实施例，此处不再赘述。
需要说明的是，在本文中，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者系统不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者系统所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括该要素的过程、方法、物品或者系统中还存在另外的相同要素。
上述本申请实施例序号仅仅为了描述，不代表实施例的优劣。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在如上所述的一个可读存储介质(如ROM/RAM、磁碟、光盘)中，包括若干指令用以使得一台ESG评分体系的生成设备(可以是手机，计算机，服务器，空调器，或者网络设备等)执行本申请各个实施例所述的方法。
以上仅为本申请的优选实施例，并非因此限制本申请的专利范围，凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本申请的专利保护范围内。

Claims (20)

  1. 一种ESG评分体系的生成方法，其中，所述ESG评分体系的生成方法包括如下步骤：
    将历史数据及所述历史数据对应的第一种子词汇放入ESG指标生成器处理，得到历史指标对应的历史热度；
    将当前获取的第一舆情数据及所述第一舆情数据对应的第二种子词汇放入所述ESG指标生成器处理，得到当前指标及当前热度；
    基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡。
  2. 如权利要求1所述的ESG评分体系的生成方法，其中，所述将历史数据放入ESG指标生成器处理，得到历史指标对应的历史热度的步骤之前，所述ESG评分体系的生成方法包括：
    从网络上采集数据信息，且基于所述数据信息构建海量语料库；
    根据预设行业标准，对所述海量语料库中的数据信息进行分类处理，得到行业语料库；
    确定行业对应的一级种子词汇，将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度；
    通过所述主题提取器对所述二级指标及二级热度进行迭代处理，得到ESG指标生成器。
  3. 如权利要求2所述的ESG评分体系的生成方法，其中，所述根据预设行业标准，对所述海量语料库中的数据信息进行分类处理，得到行业语料库的步骤包括：
    根据预设行业标准，利用文本分析技术对所述海量语料库中的数据信息进行分类处理，得到行业语料库。
  4. 如权利要求2所述的ESG评分体系的生成方法，其中，所述将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度的步骤包括：
    利用自然语言处理技术及主题模型，确定主题提取器；
    将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度。
  5. 如权利要求1至4中任一项所述的ESG评分体系的生成方法，其中，所述基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡的步骤包括：
    基于ESG指标体系库，获取所述历史热度及历史权重；
    基于所述历史热度、当前热度，调节所述历史权重，得到ESG最新评分卡。
  6. 如权利要求5任一项所述的ESG评分体系的生成方法，其中，所述基于ESG指标体系库，获取所述历史热度及历史权重的步骤之前，所述ESG评分体系的生成方法包括：
    将所述历史热度进行归一化处理，得到历史权重；
    基于所述历史指标、历史热度及历史权重，得到ESG基础评分卡；
    将所述ESG基础评分卡放入ESG指标体系库。
  7. 如权利要求5所述的ESG评分体系的生成方法，其中，所述基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡的步骤之后，所述ESG评分体系的生成方法包括：
    获取企业当前的第二舆情数据；
    基于所述ESG指标体系库，获取所述历史指标；
    通过自然语言处理技术，基于所述第二舆情数据、历史指标及所述ESG最新评分卡，得到所述企业的ESG得分；
    将所述ESG得分进行可视化处理。
  8. 一种ESG评分体系的生成装置，其中，所述ESG评分体系的生成装置包括：
    第一处理模块，用于将历史数据及所述历史数据对应的第一种子词汇放入ESG指标生成器处理，得到历史指标对应的历史热度；
    第二模块，用于将当前获取的第一舆情数据及所述第一舆情数据对应的第二种子词汇放入所述ESG指标生成器处理，得到当前指标及当前热度；
    评分模块，用于基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡。
  9. 一种ESG评分体系的生成设备，其中，所述ESG评分体系的生成设备包括：存储器、处理器及存储在所述存储器上并可在所述处理器上运行的 ESG评分体系的生成程序，所述ESG评分体系的生成程序被所述处理器执行时实现如下步骤：
    将历史数据及所述历史数据对应的第一种子词汇放入ESG指标生成器处理，得到历史指标对应的历史热度；
    将当前获取的第一舆情数据及所述第一舆情数据对应的第二种子词汇放入所述ESG指标生成器处理，得到当前指标及当前热度；
    基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡。
  10. 如权利要求9所述的ESG评分体系的生成设备，其中，所述将历史数据放入ESG指标生成器处理，得到历史指标对应的历史热度的步骤之前，所述ESG评分体系的生成方法包括：
    从网络上采集数据信息，且基于所述数据信息构建海量语料库；
    根据预设行业标准，对所述海量语料库中的数据信息进行分类处理，得到行业语料库；
    确定行业对应的一级种子词汇，将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度；
    通过所述主题提取器对所述二级指标及二级热度进行迭代处理，得到ESG指标生成器。
  11. 如权利要求10所述的ESG评分体系的生成设备，其中，所述根据预设行业标准，对所述海量语料库中的数据信息进行分类处理，得到行业语料库的步骤包括：
    根据预设行业标准，利用文本分析技术对所述海量语料库中的数据信息进行分类处理，得到行业语料库。
  12. 如权利要求10所述的ESG评分体系的生成设备，其中，所述将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度的步骤包括：
    利用自然语言处理技术及主题模型，确定主题提取器；
    将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度。
  13. 如权利要求9至12中任一项所述的ESG评分体系的生成设备，其中，所述基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡的 步骤包括：
    基于ESG指标体系库，获取所述历史热度及历史权重；
    基于所述历史热度、当前热度，调节所述历史权重，得到ESG最新评分卡。
  14. 如权利要求13任一项所述的ESG评分体系的生成设备，其中，所述基于ESG指标体系库，获取所述历史热度及历史权重的步骤之前，所述ESG评分体系的生成方法包括：
    将所述历史热度进行归一化处理，得到历史权重；
    基于所述历史指标、历史热度及历史权重，得到ESG基础评分卡；
    将所述ESG基础评分卡放入ESG指标体系库。
  15. 如权利要求13所述的ESG评分体系的生成设备，其中，所述基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡的步骤之后，所述ESG评分体系的生成方法包括：
    获取企业当前的第二舆情数据；
    基于所述ESG指标体系库，获取所述历史指标；
    通过自然语言处理技术，基于所述第二舆情数据、历史指标及所述ESG最新评分卡，得到所述企业的ESG得分；
    将所述ESG得分进行可视化处理。
  16. 一种可读存储介质，其中，所述可读存储介质上存储有ESG评分体系的生成程序，所述ESG评分体系的生成程序被处理器执行时实现如下步骤：
    将历史数据及所述历史数据对应的第一种子词汇放入ESG指标生成器处理，得到历史指标对应的历史热度；
    将当前获取的第一舆情数据及所述第一舆情数据对应的第二种子词汇放入所述ESG指标生成器处理，得到当前指标及当前热度；
    基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡。
  17. 如权利要求16所述的可读存储介质，其中，所述将历史数据放入ESG指标生成器处理，得到历史指标对应的历史热度的步骤之前，所述ESG评分体系的生成方法包括：
    从网络上采集数据信息，且基于所述数据信息构建海量语料库；
    根据预设行业标准，对所述海量语料库中的数据信息进行分类处理，得到行业语料库；
    确定行业对应的一级种子词汇，将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度；
    通过所述主题提取器对所述二级指标及二级热度进行迭代处理，得到ESG指标生成器。
  18. 如权利要求17所述的可读存储介质，其中，所述根据预设行业标准，对所述海量语料库中的数据信息进行分类处理，得到行业语料库的步骤包括：
    根据预设行业标准，利用文本分析技术对所述海量语料库中的数据信息进行分类处理，得到行业语料库。
  19. 如权利要求17所述的可读存储介质，其中，所述将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度的步骤包括：
    利用自然语言处理技术及主题模型，确定主题提取器；
    将所述行业语料库及所述一级种子词汇输入主题提取器中处理，得到二级指标及二级热度。
  20. 如权利要求16至19中任一项所述的可读存储介质，其中，所述基于所述历史热度、当前热度及当前指标，获得ESG最新评分卡的步骤包括：
    基于ESG指标体系库，获取所述历史热度及历史权重；
    基于所述历史热度、当前热度，调节所述历史权重，得到ESG最新评分卡。
PCT/CN2020/107325 2019-09-04 2020-08-06 Esg评分体系的生成方法、装置、设备及可读存储介质 WO2021004550A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910835812.0 2019-09-04
CN201910835812.0A CN110532357B (zh) 2019-09-04 2019-09-04 Esg评分体系的生成方法、装置、设备及可读存储介质

Publications (1)

Publication Number Publication Date
WO2021004550A1 true WO2021004550A1 (zh) 2021-01-14

Family

ID=68667259

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/107325 WO2021004550A1 (zh) 2019-09-04 2020-08-06 Esg评分体系的生成方法、装置、设备及可读存储介质

Country Status (2)

Country Link
CN (1) CN110532357B (zh)
WO (1) WO2021004550A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113610645A (zh) * 2021-08-11 2021-11-05 平安信托有限责任公司 项目数据的风险评估方法、装置、设备及存储介质
CN113837527A (zh) * 2021-08-02 2021-12-24 深圳前海微众银行股份有限公司 企业评级方法、装置、设备以及存储介质
WO2023050652A1 (zh) * 2021-09-28 2023-04-06 平安科技（深圳）有限公司 基于文本识别的区域内esg指数确定方法及相关产品
US20230162146A1 (en) * 2021-11-25 2023-05-25 Hitachi, Ltd. Value chain providing apparatus, value chain providing system, and value chain provision method

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110532357B (zh) * 2019-09-04 2024-03-12 深圳前海微众银行股份有限公司 Esg评分体系的生成方法、装置、设备及可读存储介质
CN111460260A (zh) * 2020-03-31 2020-07-28 上海智芝全智能科技有限公司 多类型数据的数据处理系统、方法及介质
TWI747421B (zh) * 2020-08-05 2021-11-21 元大證券投資信託股份有限公司 投資組合選擇系統
CN113190683B (zh) * 2021-07-02 2021-09-17 平安科技（深圳）有限公司 基于聚类技术的企业esg指数确定方法及相关产品
CN113570281A (zh) * 2021-08-20 2021-10-29 瑞格人工智能科技有限公司 一种esg指数编制方法
CN113436024A (zh) * 2021-08-25 2021-09-24 江苏荣泽信息科技股份有限公司 基于区块链的esg可信指标综合服务平台
CN113988651A (zh) * 2021-10-29 2022-01-28 深圳壹账通智能科技有限公司 一种评分卡模型的评价方法、系统、设备及介质
CN114169686A (zh) * 2021-11-11 2022-03-11 上海华证指数信息服务有限公司 一种上市公司esg评价方法
CN114139539A (zh) * 2021-12-06 2022-03-04 城云科技（中国）有限公司 企业社会责任指标量化方法、系统及应用
CN118195325A (zh) * 2024-04-26 2024-06-14 深观(深圳)科技有限公司 项目评标方法、系统、设备及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104995650A (zh) * 2011-12-27 2015-10-21 汤姆森路透社全球资源公司 用于使用源于社交媒体的数据和情绪分析来生成复合索引的方法及系统
US20170004128A1 (en) * 2015-07-01 2017-01-05 Institute for Sustainable Development Device and method for analyzing reputation for objects by data mining
WO2019103183A1 (ko) * 2017-11-23 2019-05-31 지속가능발전소 주식회사 Esg 기반의 기업 평가 수행 장치 및 이의 작동 방법
CN110532357A (zh) * 2019-09-04 2019-12-03 深圳前海微众银行股份有限公司 Esg评分体系的生成方法、装置、设备及可读存储介质

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP6494687B2 (ja) * 2016-12-02 2019-04-03 本田技研工業株式会社 評価装置、評価方法、および評価プログラム
CN107909274B (zh) * 2017-11-17 2023-02-28 平安科技（深圳）有限公司 企业投资风险评估方法、装置及存储介质
CN108062306A (zh) * 2017-12-29 2018-05-22 国信优易数据有限公司 一种营商环境评价的指标体系构建系统及方法
CN109190986A (zh) * 2018-09-10 2019-01-11 张连祥 基于直接客观数据的营商环境分析评价系统及方法
CN109636150A (zh) * 2018-11-30 2019-04-16 南京市城市规划编制研究中心 一种智慧城市“多规合一”评价体系的建立方法及其系统
CN109657070B (zh) * 2018-12-11 2023-06-09 南京大学 一种终端辅助swot指标体系的构建方法
CN109684484B (zh) * 2018-12-11 2023-06-09 南京大学 一种swot指标体系构建系统

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104995650A (zh) * 2011-12-27 2015-10-21 汤姆森路透社全球资源公司 用于使用源于社交媒体的数据和情绪分析来生成复合索引的方法及系统
US20170004128A1 (en) * 2015-07-01 2017-01-05 Institute for Sustainable Development Device and method for analyzing reputation for objects by data mining
WO2019103183A1 (ko) * 2017-11-23 2019-05-31 지속가능발전소 주식회사 Esg 기반의 기업 평가 수행 장치 및 이의 작동 방법
CN110532357A (zh) * 2019-09-04 2019-12-03 深圳前海微众银行股份有限公司 Esg评分体系的生成方法、装置、设备及可读存储介质

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113837527A (zh) * 2021-08-02 2021-12-24 深圳前海微众银行股份有限公司 企业评级方法、装置、设备以及存储介质
CN113610645A (zh) * 2021-08-11 2021-11-05 平安信托有限责任公司 项目数据的风险评估方法、装置、设备及存储介质
WO2023050652A1 (zh) * 2021-09-28 2023-04-06 平安科技（深圳）有限公司 基于文本识别的区域内esg指数确定方法及相关产品
US20230162146A1 (en) * 2021-11-25 2023-05-25 Hitachi, Ltd. Value chain providing apparatus, value chain providing system, and value chain provision method

Also Published As

Publication number Publication date
CN110532357B (zh) 2024-03-12
CN110532357A (zh) 2019-12-03

Similar Documents

Publication Publication Date Title
WO2021004550A1 (zh) Esg评分体系的生成方法、装置、设备及可读存储介质
CA3129745C (en) Neural network system for text classification
US20240346059A1 (en) System for determining and optimizing for relevance in match-making systems
US20220067665A1 (en) Three-party recruiting and matching process involving a candidate, referrer, and hiring entity
US11908028B2 (en) Method and system for curriculum management services
US11586656B2 (en) Opportunity network system for providing career insights by determining potential next positions and a degree of match to a potential next position
US20150142423A1 (en) Phrase-based data classification system
US20120041769A1 (en) Requests for proposals management systems and methods
CN109492097B (zh) 一种企业新闻数据风险分类方法
CN112579729B (zh) 文档质量评价模型的训练方法、装置、电子设备和介质
KR102465711B1 (ko) 소셜 데이터 분석을 이용한 법률안의 국회 본회의 통과가능성 동적 예측 서비스 제공 시스템
US20210224843A1 (en) Method and system for determining equity index for a brand
Field et al. Assessing the value of removing earthquake-hazard-related epistemic uncertainties, exemplified using average annual loss in California
Yang et al. Big data-based human resource performance evaluation model using Bayesian network of deep learning
Kang et al. PG-CODE: Latent dirichlet allocation embedded policy knowledge graph for government department coordination
Chen Quality evaluation of student education management work based on wireless network data mining
WO2021035955A1 (zh) 文本新闻的处理方法、设备和存储介质
Madhusanka et al. Social network analysis of building energy and carbon policy networks in developing countries
CN106485466A (zh) 一种项目管理系统
CN113836244B (zh) 样本获取方法、模型训练方法、关系预测方法及装置
CN115796600A (zh) 一种舆情风险预警方法、系统、介质及电子设备
Jiang et al. [Retracted] Employment Recommendation for Education Talents Based on Big Data Precision Technology
US20200387991A1 (en) Data management system for use with agreements and data detailng concepts, designs, and ideas
CN111798217B (zh) 数据分析系统及方法
Tseng et al. An intelligent disease query system based on RASA NLU

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20837454

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20837454

Country of ref document: EP

Kind code of ref document: A1