Knowledge graph-based customer demand response method and system

Abstract

The application discloses a kind of customer demand response method and system based on knowledge graph.The method comprises the following steps: the relationship between each entity in carbon trading is constructed, and the carbon trading knowledge graph is obtained;Using time series analysis method obtains entity time correlation;Feature extraction is carried out on entity time correlation, and the extracted features are fused with entity attributes and relationship features in carbon trading knowledge graph to obtain comprehensive feature spectrum;The inference result and the prediction result for customer demand response are obtained by carrying out relationship reasoning and demand prediction on comprehensive feature spectrum;The influence of different carbon trading market conditions, carbon emission reduction policies and carbon emission reduction technologies on customer demand is simulated and analyzed, and the inference result and the prediction result are updated;Visual interactive display is carried out.The application realizes providing more comprehensive carbon trading information and more accurate demand prediction for customers, realizes fully integrating data and optimizing demand prediction model and service efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and particularly relates to a customer demand response method and system based on a knowledge graph. BACKGROUND

[0002] With the increasing global concern about climate change, the carbon trading market is becoming increasingly active. In carbon trading, accurately responding to customer demand is crucial for achieving carbon emission reduction targets and promoting the efficient operation of the market.

[0003] At present, customer demand response based on a knowledge graph has made some progress in carbon trading, but there are still the following technical problems:

[0004] First, the data integration is insufficient. Existing methods often limit to partial data sources when integrating carbon trading related data, such as only considering carbon trading market data and enterprise carbon emission data, while ignoring industry dynamic information, policy and regulation change records and scientific research data, etc., which leads to incomplete information of the knowledge graph and difficulty in accurately grasping the overall picture of customer demand;

[0005] Second, the demand prediction model is single. Current demand analysis and prediction mainly rely on traditional methods, lack of innovation, especially the combination of deep learning and knowledge graph is not close enough, and the advantages of both are not fully utilized. At the same time, the method of scenario simulation and prediction is relatively simple, and cannot fully consider various complex market, policy and technical changes;

[0006] Third, the customer interaction and service efficiency is low. The intelligent degree of intelligent customer service system needs to be improved, which cannot well understand the complex problems of customers and provide personalized solutions. Especially, the function of the visual interaction platform is limited, which is difficult to meet the customer's demand for in-depth understanding and convenient operation of carbon trading information. SUMMARY

[0007] In view of the deficiencies of the prior art, the present application provides a customer demand response method and system based on a knowledge graph, which provides customers with more comprehensive carbon trading information and more accurate demand prediction, fully integrates data and optimizes demand prediction model and service efficiency, and helps customers make more intelligent decisions in the complex carbon trading market.

[0008] To achieve the above purpose, the customer demand response method based on the knowledge graph designed by the present application has the following steps:

[0009] S1) Obtain multi-source data in carbon trading, use a knowledge graph construction tool to extract and model entities and relationships in various data, construct relationships between various entities in carbon trading, and obtain a carbon trading knowledge graph;

[0010] S2) adopting a time series analysis method to extract corresponding time features from the carbon trading knowledge graph, and combining entity relationships in the carbon trading knowledge graph to obtain entity time correlation relationships; the entity time correlation relationships include the change rule of data with time sequence in the multi-source data over time, and the change relationship of different entities in the time dimension;

[0011] S3) using a deep learning model to extract features of the entity time correlation relationships, and fusing the extracted features with entity attributes and relationship features in the carbon trading knowledge graph to obtain a comprehensive feature graph;

[0012] S4) performing relationship reasoning and demand prediction on the comprehensive feature graph to obtain reasoning results and prediction results for customer demand response; wherein the reasoning results include inferred unknown or potential relationships, and the prediction results include predicted customer demand;

[0013] S5) repeating steps S1) to S4), simulating and analyzing the influence of different carbon trading market conditions, carbon emission reduction policies and carbon emission reduction technologies on customer demand, and updating the reasoning results and prediction results of step S4) based on the analysis results.

[0014] Further, in S1), the multi-source data at least includes carbon trading data, enterprise production and operation data, industry dynamic information, policy and regulation change records, and scientific research achievement data.

[0015] Further, in S1), before extracting and modeling entities and relationships in various types of data, data cleaning and preprocessing techniques are used to remove noise and outliers.

[0016] Further, in S1), the entities include enterprises, policies, and technologies; the relationships include the association between enterprises and policies, the association between enterprises and technologies, and the business relationship between enterprises.

[0017] Further, in S2), the time features include trend features, seasonal features, and periodic features.

[0018] Further, in S3), the deep learning model includes a convolutional neural network, a recurrent neural network, or a long short-term memory network; the fusion method includes feature splicing and weighted summation.

[0019] Further, in S3), the entity attributes at least include enterprise size and industry type, and the relationship features at least include supply chain relationships between enterprises and the influence of policies on enterprises.

[0020] Further, in S5), simulating and analyzing the influence of different carbon trading market conditions on customer demand, specifically:

[0021] Set different carbon trading market scenarios, simulate the operation of the carbon trading market under different carbon trading market scenarios by using a mathematical model and a simulation tool, obtain the analysis result based on the behavior change and demand response of enterprise customers;

[0022] Simulate and analyze the influence of different carbon emission reduction policies on customer demand, specifically:

[0023] Obtain the historical records of different carbon emission reduction policies and the influence of the corresponding policies on enterprise carbon emission costs, production and operation decisions, and carbon trading demand, and perform similarity analysis combined with the carbon trading knowledge graph to obtain the analysis result;

[0024] Simulate and analyze the influence of different carbon emission reduction technologies on customer demand, specifically:

[0025] Evaluate the influence of carbon emission reduction technologies on enterprise carbon emission levels, production costs, and market competitiveness, and combine the carbon trading knowledge graph to obtain the analysis result based on the change in carbon trading demand after adopting the carbon emission reduction technology.

[0026] Further, it further comprises a step S6) of visualizing and interactively displaying the carbon trading knowledge graph, the entity-time correlation relationship, the comprehensive feature graph, the reasoning result, the prediction result, the updated reasoning result, and the updated prediction result.

[0027] The application also designs a customer demand response system based on a knowledge graph, which is applicable to the customer demand response method based on a knowledge graph described above, and the particularity of the system lies in that the system comprises a carbon trading knowledge graph acquisition module, an entity-time correlation relationship acquisition module, a comprehensive feature graph acquisition module, a reasoning and prediction module, a simulation and updating module, and a visualization module.

[0028] The carbon trading knowledge graph acquisition module is used to acquire multi-source data in carbon trading, extract and model entities and relationships in various data by using a knowledge graph construction tool, construct relationships between various entities in carbon trading, and obtain a carbon trading knowledge graph.

[0029] The entity-time correlation relationship acquisition module is used to extract corresponding time characteristics from the carbon trading knowledge graph by using a time series analysis method, and obtain an entity-time correlation relationship combined with the entity relationships in the carbon trading knowledge graph.

[0030] The comprehensive feature graph acquisition module is used to extract features of the entity-time correlation relationship by using a deep learning model, fuse the extracted features with entity attributes and relationship features in the carbon trading knowledge graph, and obtain a comprehensive feature graph.

[0031] The reasoning and prediction module is used for relationship reasoning and demand prediction on the comprehensive feature graph, to obtain reasoning results and prediction results for customer demand response; wherein, the reasoning results include inferred unknown or potential relationships, and the prediction results include predicted customer demands.

[0032] The simulation and update module is used for simulating and analyzing the influence of different carbon trading market conditions, carbon emission reduction policies and carbon emission reduction technologies on customer demands, and updating the reasoning results and prediction results of the reasoning and prediction module based on the analysis results.

[0033] The visualization module is used for visual interactive display of the carbon trading knowledge graph, entity time correlation relationship, comprehensive feature graph, reasoning results, prediction results, updated reasoning results and updated prediction results.

[0034] The present application has the following advantages:

[0035] The present application obtains a carbon trading knowledge graph comprehensively representing the relationships between entities in carbon trading by correlating and integrating multi-source data and knowledge graph technology, obtains a comprehensive feature graph by using time series analysis and deep learning fusion, and performs relationship reasoning and demand prediction, and then combines different scenario simulation and visual interactive display, to accurately respond to customer demands in carbon trading, provide customers with more comprehensive carbon trading information and more accurate demand prediction, fully integrate data and optimize demand prediction model and service efficiency, and help customers make more intelligent decisions in the complex carbon trading market, to promote efficient operation of the carbon trading market and realization of carbon emission reduction targets. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 The present application is a flowchart of a customer demand response method based on a knowledge graph.

[0037] Figure 2 The present application is a structural diagram of a customer demand response system based on a knowledge graph. DETAILED DESCRIPTION

[0038] The present application will be further described in detail below in combination with the drawings and specific embodiments.

[0039] As shown in the drawings, the present application is a customer demand response method based on a knowledge graph, comprising the following steps: Figure 1

[0040] S1) Obtain multi-source data in carbon trading, extract and model entities and relationships in various data using a knowledge graph construction tool, construct relationships between various entities in carbon trading, and obtain a carbon trading knowledge graph. The carbon trading knowledge graph is used to represent relationships between various entities in carbon trading.

[0041] ​Specifically, data from different channels are collected, and the multi-source data at least includes carbon trading data, enterprise production and operation data, industry dynamic information, policy and regulation change records, and scientific research data.

[0042] Specifically, the entities include enterprises, policies, technologies, etc., and the relationships include the association between enterprises and policies, the association between enterprises and technologies, business relationships between enterprises, etc.

[0043] Preferably, before extracting and modeling the entities and relationships in various types of data, data cleaning and preprocessing techniques are used to remove noise and outliers.

[0044] It should be noted that the present embodiment uses existing knowledge graph construction tools to obtain a carbon trading knowledge graph, and obtains a comprehensive and multi-dimensional knowledge graph that can reflect the complex relationships between various entities in the carbon trading field, thereby achieving the construction of an accurate and comprehensive carbon trading knowledge graph and providing a data basis for subsequent analysis and prediction.

[0045] S2) Adopting a time series analysis method to extract corresponding time features from the carbon trading knowledge graph, and combining the entity relationships in the carbon trading knowledge graph to obtain entity time correlation relationships; the entity time correlation relationships include the change rule of the data with time sequence in the multi-source data with time, and the change relationship of different entities in the time dimension.

[0046] Specifically, the time series analysis method includes moving average, exponential smoothing, ARIMA model, etc.

[0047] Specifically, the time features include trend features, seasonal features, and periodic features. The obtained entity time correlation relationships better grasp the change rule of the data with time and the correlation pattern of different entities in time.

[0048] In the present embodiment, the carbon trading price in the carbon trading data and the enterprise carbon emission data in the enterprise production and operation data are analyzed, and a time series analysis method such as moving average, exponential smoothing, ARIMA model, etc. is adopted to extract the trend, seasonality, and periodicity of the data; at the same time, the change relationship of different entities in the time dimension is analyzed in combination with the entity relationships in the knowledge graph, the change rule of the carbon trading data with time is found, and the correlation pattern of different entities in time is found, thereby providing a basis for predicting the time variation trend of customer demand.

[0049] S3) Using a deep learning model to extract features of the entity time correlation relationship, and fusing the extracted features with the entity attributes and relationship features in the carbon trading knowledge graph to obtain a comprehensive feature graph.

[0050] Preferably, the deep learning model comprises a convolutional neural network, a recurrent neural network, or a long short-term memory network; and the fusion method comprises feature concatenation or weighted summation.

[0051] Specifically, the entity attributes comprise at least enterprise size and industry type, and the relationship features comprise at least supply chain relationship between enterprises and influence relationship of policies on enterprises.

[0052] In this embodiment, a suitable deep learning model, such as a convolutional neural network, a recurrent neural network, or a long short-term memory network, is selected to extract features from the carbon trading data. The extracted features and the entity attributes and relationship features in the knowledge graph can be fused by using methods such as feature concatenation or weighted summation to obtain a comprehensive feature graph that integrates the deep learning features and the knowledge graph features, thereby more comprehensively describing the entities and relationships in carbon trading.

[0053] S4) performing relationship reasoning and demand prediction on the comprehensive feature graph to obtain reasoning results and prediction results for customer demand response; wherein the reasoning results comprise inferred unknown or potential relationships, and the prediction results comprise predicted customer demand.

[0054] The structural information of the carbon trading knowledge graph is utilized in combination with a deep learning model to reason about the entity relationships in carbon trading, analyze the relationships between known entities, and infer unknown or potential relationships to obtain reasoning results. The fused features are predicted by using a deep learning model to obtain prediction results. The reasoning results and the prediction results are used for customer demand response.

[0055] In this embodiment, the structural information of the knowledge graph is utilized in combination with a deep learning reasoning algorithm, such as a graph neural network-based reasoning method, to reason about the entity relationships in carbon trading by analyzing the relationships between known entities to infer unknown or potential relationships. Meanwhile, the fused features are predicted by using an existing deep learning model, such as to predict carbon trading demand of enterprises, carbon price trends, and the like. By reasoning about the entity relationships and predicting the fused features, reasoning results and prediction results are provided for customer demand response, thereby improving the accuracy of the response.

[0056] S5) repeating steps S1) to S4) to simulate and analyze the influence of different carbon trading market conditions, carbon emission reduction policies, and carbon emission reduction technologies on customer demand, and updating the reasoning results and the prediction results of step S4) based on the analysis results.

[0057] Specifically, the influence of different carbon trading market conditions on customer demand is simulated and analyzed, specifically: different carbon trading market scenarios are set, mathematical models and simulation tools are used to simulate the operation of the carbon trading market under different carbon trading market scenarios, and the analysis result is obtained based on the behavior change and demand response of enterprise customers.

[0058] The embodiment can use Monte Carlo simulation, system dynamics model, etc. to set different carbon trading market scenarios, such as carbon price fluctuation range, market supply and demand relationship change degree, etc., simulate the operation of the carbon trading market under these scenarios, analyze the behavior change and demand response of enterprise customers, and provide coping strategies for customers.

[0059] Specifically, the influence of different carbon reduction policies on customer demand is simulated and analyzed, specifically: different carbon reduction policies and the influence of the corresponding policies on enterprise carbon emission cost, production and operation decision, and carbon trading demand are obtained, and a similarity analysis is performed in combination with the carbon trading knowledge graph to obtain the analysis result.

[0060] In the embodiment, in combination with enterprise information and policy relationship in the knowledge graph, the influence of various carbon reduction policies, such as carbon emission restriction policy, carbon tax policy, and carbon subsidy policy, on enterprise carbon emission cost, production and operation decision, and carbon trading demand is researched by using existing similarity analysis technology, specific case analysis is performed, the corresponding similarity is obtained, and customer demand response is performed to help customers adapt to policy changes.

[0061] Specifically, the influence of different carbon reduction technologies on customer demand is simulated and analyzed, specifically: the influence of carbon reduction technologies on enterprise carbon emission level, production cost, and market competitiveness is evaluated, and the analysis result is obtained based on the change of carbon trading demand after the carbon reduction technology is adopted in combination with the carbon trading knowledge graph.

[0062] In the embodiment, by assuming the emergence of new carbon reduction technologies, which can also be energy technologies, the influence of the new carbon reduction technologies on enterprise carbon emission level, production cost, and market competitiveness is evaluated, and the change of carbon trading demand after the new technology is adopted is analyzed by using technical and economic analysis methods such as cost-benefit analysis and life cycle assessment in combination with technical information and enterprise relationship in the carbon trading knowledge graph, trend analysis and suggestions for technical innovation are provided for customers, and customers are helped to adopt new technologies in a timely manner, reduce carbon emission cost, and improve carbon trading efficiency, thereby realizing the analysis of the influence of different carbon reduction technologies on customer demand and providing technical selection suggestions for customers.

[0063] Based on the analysis result, the structural information of the carbon transaction knowledge graph is updated, the deep learning model is recombined, the entity relationship in the carbon transaction is reasoned, the relationship between the known entities is analyzed, the unknown or potential relationship is inferred, and the reasoning result is updated; the deep learning model is re-used to predict the demand based on the fused features, and the prediction result is updated; the updated reasoning result and prediction result are used for customer demand response, and the timeliness and accuracy of customer demand response are improved.

[0064] Preferably, the present application further comprises a step S6) of visualizing and interactively displaying the carbon transaction knowledge graph, the entity time correlation relationship, the comprehensive feature graph, the reasoning result, the prediction result, the updated reasoning result, and the updated prediction result.

[0065] The present application also designs a customer demand response system based on a knowledge graph, which is applicable to the above-mentioned customer demand response method based on a knowledge graph, and comprises a carbon transaction knowledge graph acquisition module 1, an entity time correlation relationship acquisition module 2, a comprehensive feature graph acquisition module 3, a reasoning and prediction module 4, a simulation and updating module 5, and a visualization module 6.

[0066] The carbon transaction knowledge graph acquisition module 1 is used to acquire multi-source data in carbon transactions, extract and model the entities and relationships in various types of data using a knowledge graph construction tool, construct the relationships between various entities in carbon transactions, and obtain a carbon transaction knowledge graph.

[0067] The entity time correlation relationship acquisition module 2 is used to extract corresponding time features from the carbon transaction knowledge graph using a time series analysis method, and obtain an entity time correlation relationship in combination with the entity relationships in the carbon transaction knowledge graph.

[0068] The comprehensive feature graph acquisition module 3 is used to extract features from the entity time correlation relationship using a deep learning model, fuse the extracted features with the entity attributes and relationship features in the carbon transaction knowledge graph, and obtain a comprehensive feature graph.

[0069] The reasoning and prediction module 4 is used to reason and predict the demand based on the comprehensive feature graph, and obtain reasoning results and prediction results for customer demand response; wherein the reasoning results include inferred unknown or potential relationships, and the prediction results include predicted customer demands.

[0070] The simulation and updating module 5 is used to simulate and analyze the influence of different carbon transaction market conditions, carbon emission reduction policies, and carbon emission reduction technologies on customer demand, and update the reasoning results and prediction results of the reasoning and prediction module 4 based on the analysis results.

[0071] The visualization module 6 is configured to visualize and interactively display the carbon transaction knowledge graph, the entity time correlation relationship, the comprehensive feature graph, the reasoning result, the prediction result, the updated reasoning result, and the updated prediction result.

[0072] In the embodiments provided by the present application, it should be understood that the embodiments described herein can be realized by hardware, software, firmware, middleware, codes or any proper combination thereof. For hardware implementation, the processor can be realized in one or more of the following components: an application specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a processor, a controller, a microcontroller, a microprocessor, other electronic units designed to implement the functions described herein, or a combination thereof. For software implementation, the part or all of the flow of the embodiments can be completed by computer program instructions. When implemented, the above program instructions can be stored in a computer readable storage medium or transmitted as one or more instructions or codes on a computer readable storage medium. The computer readable storage medium includes computer storage media and communication media, where the communication media includes any medium that facilitates the transfer of computer program from one place to another. The storage medium can be any available medium that can be accessed by a computer. The computer readable storage medium can include but not limited to RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage medium or other magnetic storage devices, or any other medium that can be used to carry or store desired program codes in the form of instructions or data structures and can be accessed by a computer.

[0073] The above embodiments are the preferred embodiments of the present application, but the embodiments of the present application are not limited to the above embodiments, and any changes, modifications, substitutions, combinations, simplifications made without departing from the spirit and principles of the present application shall be equivalent replacement modes, and all shall be included in the protection scope of the present application.

Claims

1. A knowledge graph-based customer demand response method, characterized in that, The method comprises the following steps: S1) obtaining multi-source data in carbon trading, using a knowledge graph construction tool to extract and model entities and relationships in various types of data, constructing relationships between various entities in carbon trading, and obtaining a carbon trading knowledge graph; S2) using a time series analysis method to extract corresponding time features from the carbon trading knowledge graph, and combining entity relationships in the carbon trading knowledge graph to obtain entity time correlation relationships; the entity time correlation relationships include the change rule of data with time sequence in the multi-source data over time, and the change relationship of different entities in the time dimension; S3) using a deep learning model to extract features of the entity time correlation relationship, fusing the extracted features with entity attributes and relationship features in the carbon trading knowledge graph, and obtaining a comprehensive feature graph; S4) performing relationship reasoning and demand prediction on the comprehensive feature graph to obtain reasoning results and prediction results for customer demand response; wherein the reasoning results include inferred unknown or potential relationships, and the prediction results include predicted customer demand; S5) repeating steps S1) to S4), simulating and analyzing the influence of different carbon trading market conditions, carbon emission reduction policies and carbon emission reduction technologies on customer demand, and updating the reasoning results and prediction results of step S4) based on the analysis results. 2.The knowledge graph based customer demand response method of claim 1, wherein: In S1), the multi-source data at least includes carbon trading data, enterprise production and operation data, industry dynamic information, policy and regulation change records, and scientific research achievement data. 3.The knowledge graph based customer demand response method of claim 2, wherein: In S1), before extracting and modeling entities and relationships in various types of data, data cleaning and preprocessing techniques are used to remove noise and outliers. 4.The knowledge graph based customer demand response method of claim 3, wherein: In S1), the entities include enterprises, policies, and technologies; the relationships include the association between enterprises and policies, the association between enterprises and technologies, and the business relationship between enterprises. 5.The knowledge graph based customer demand response method of claim 1, wherein: In S2), the time features include trend features, seasonal features, and periodic features. 6.The knowledge graph based customer demand response method of claim 1, wherein: In S3), the deep learning model includes a convolutional neural network, a recurrent neural network, or a long short-term memory network; the fusion method includes feature splicing and weighted summation. 7.The knowledge graph based customer demand response method of claim 6, wherein: In S3), the entity attributes at least include enterprise size and industry type, and the relationship features at least include supply chain relationships between enterprises and the influence of policies on enterprises. 8.The knowledge graph based customer demand response method of claim 1, wherein, In S5), the influence of different carbon trading market conditions on customer demand is simulated and analyzed, specifically: Different carbon trading market scenarios are set, the operation of the carbon trading market under different carbon trading market scenarios is simulated using mathematical models and simulation tools, and the analysis results are obtained based on the behavior changes and demand responses of enterprise customers; The influence of different carbon emission reduction policies on customer demand is simulated and analyzed, specifically: Different carbon emission reduction policies and the influence of the corresponding policies on enterprise carbon emission costs, production and operation decisions, and carbon trading demand are obtained, and similarity analysis is performed in combination with the carbon trading knowledge graph to obtain the analysis results; The influence of different carbon emission reduction technologies on customer demand is simulated and analyzed, specifically: The influence of the carbon emission reduction technology on the carbon emission level, production cost and market competitiveness of the enterprise is evaluated, and the analysis result is obtained based on the carbon trading knowledge graph and the change of carbon trading demand after the carbon emission reduction technology is adopted. 9.The knowledge graph based customer demand response method of claim 1, wherein: The carbon trading knowledge graph, the entity time correlation relationship, the comprehensive feature graph, the reasoning result, the prediction result, the updated reasoning result and the updated prediction result are visualized and interactively displayed.

10. A knowledge graph-based customer demand response system, suitable for the knowledge graph-based customer demand response method according to any one of claims 1 to 9, characterized in that, The system comprises a carbon trading knowledge graph acquisition module (1), an entity time correlation relationship acquisition module (2), a comprehensive feature graph acquisition module (3), a reasoning and prediction module (4), a simulation and updating module (5) and a visualization module (6). The carbon trading knowledge graph acquisition module (1) is used to acquire multi-source data in carbon trading, extract and model entities and relationships in various data by using a knowledge graph construction tool, construct relationships between various entities in carbon trading, and obtain a carbon trading knowledge graph. The entity time correlation relationship acquisition module (2) is used to extract corresponding time features from the carbon trading knowledge graph by using a time series analysis method, and obtain an entity time correlation relationship in combination with entity relationships in the carbon trading knowledge graph. The comprehensive feature graph acquisition module (3) is used to extract features of the entity time correlation relationship by using a deep learning model, fuse the extracted features with entity attributes and relationship features in the carbon trading knowledge graph, and obtain a comprehensive feature graph. The reasoning and prediction module (4) is used to perform relationship reasoning and demand prediction on the comprehensive feature graph, and obtain reasoning results and prediction results for customer demand response; wherein the reasoning results include inferred unknown or potential relationships, and the prediction results include predicted customer demand. The simulation and updating module (5) is used to simulate and analyze the influence of different carbon trading market conditions, carbon emission reduction policies and carbon emission reduction technologies on customer demand, and update the reasoning results and prediction results of the reasoning and prediction module (4) based on the analysis result. The visualization module (6) is used to visualize and interactively display the carbon trading knowledge graph, the entity time correlation relationship, the comprehensive feature graph, the reasoning result, the prediction result, the updated reasoning result and the updated prediction result.

Images