WO2021184825A1 - 数据处理方法、装置、电子设备及存储介质 - Google Patents

数据处理方法、装置、电子设备及存储介质 Download PDF

Info

Publication number
WO2021184825A1
WO2021184825A1 PCT/CN2020/132765 CN2020132765W WO2021184825A1 WO 2021184825 A1 WO2021184825 A1 WO 2021184825A1 CN 2020132765 W CN2020132765 W CN 2020132765W WO 2021184825 A1 WO2021184825 A1 WO 2021184825A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
index
saving amount
carbon saving
amount corresponding
Prior art date
Application number
PCT/CN2020/132765
Other languages
English (en)
French (fr)
Inventor
嵇方方
汲小溪
王维强
Original Assignee
支付宝(杭州)信息技术有限公司
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 支付宝(杭州)信息技术有限公司 filed Critical 支付宝(杭州)信息技术有限公司
Publication of WO2021184825A1 publication Critical patent/WO2021184825A1/zh

Links

Images

Classifications

    • 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
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/26—Government or public services
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/80—Management or planning
    • Y02P90/84—Greenhouse gas [GHG] management systems

Definitions

  • One or more embodiments of this specification relate to the field of computer technology, and in particular to a data processing method, device, electronic device, and computer-readable storage medium.
  • the greenhouse effect caused by greenhouse gases is causing the global climate to become increasingly warmer, which has and will continue to bring disasters to the earth and mankind.
  • the greenhouse effect In order to protect the earth's ecological environment, avoid further deterioration of the greenhouse effect, save energy, reduce carbon emissions, and achieve green management and green life, it has become one of the important goals of social and economic development and production and operation activities in many countries around the world.
  • how to encourage business entities or individuals to take the initiative to save energy and reduce emissions is the direction that all civilization needs to work hard on.
  • One of the most critical aspects is to enable business entities to clearly understand whether their business activities and daily behaviors are green business behaviors for energy conservation and emission reduction, so as to further guide business entities to consciously and proactively conduct green operations and contribute to environmental protection.
  • one or more embodiments of this specification propose a data processing method, which can determine the amount of carbon savings corresponding to the business entity based on the relevant information of the business entity, and compare the specific data of the business entity based on the corresponding carbon savings of the business entity. To deal with it, so as to guide the business entity to consciously and proactively carry out energy-saving, emission-reduction, green management, and contribute to environmental protection.
  • the data processing method proposed in this manual includes: obtaining the index data of the business entity corresponding to each evaluation index according to at least two evaluation indexes; determining the carbon saving amount corresponding to each evaluation index according to the index data of each evaluation index;
  • the trained ranking fusion model fuses the carbon saving amount corresponding to each evaluation index to obtain the carbon saving amount corresponding to the business entity; wherein the ranking fusion model uses the index data of each evaluation index and the corresponding carbon saving amount
  • the amount is training data obtained through training; the specific data of the business entity is processed according to the carbon saving amount corresponding to the business entity; wherein the specific data is related to the carbon saving amount.
  • determining the carbon saving amount corresponding to each evaluation index according to the index data of each evaluation index includes: for each evaluation index, respectively determining the carbon saving amount corresponding to each index data of the evaluation index; The carbon saving amount corresponding to the index data is normalized; and the carbon saving amount corresponding to the normalized index data is fused to obtain the carbon saving amount corresponding to the evaluation index.
  • fusing the carbon saving amount corresponding to each index data includes: setting a weight value for each index data respectively; and performing a weighted summation of the carbon saving amount corresponding to each index data according to the set weight value , To obtain the carbon saving amount corresponding to the evaluation index.
  • setting weight values for the various index data respectively includes: using at least one of analytic hierarchy, principal component analysis, and anomaly detection methods to determine the importance of the various index data; and according to the various index data
  • the importance degrees of are respectively assigned corresponding weight values for the various index data; among them, the greater the importance degree, the greater the weight value corresponding to the index data.
  • the sorting fusion model is realized by a cubic Bezier curve; wherein, the input vector of the cubic Bezier curve is the normalized value of the carbon saving amount corresponding to the index data of each evaluation index And the output of the cubic Bezier curve is the carbon saving amount corresponding to the business entity.
  • the process of training the sorting fusion model includes: initializing the endpoints and control points of the cubic Bezier curve; performing the following steps for each evaluation index: A, according to the corresponding section of each index data of the evaluation index The normalized value of carbon content determines the input vector of the above cubic Bezier curve, and uses the carbon saving amount corresponding to the evaluation index as the known output of the cubic Bezier curve; B, determine the The projection of the input vector on the cubic Bezier curve; C, the error of this training is determined according to the projection and the known output; D, in response to the situation that the error is greater than a preset error threshold , Adjust the position of the control point in the cubic Bézier curve, and return to B; E, in response to the error being less than or equal to the preset error threshold, perform all evaluation indicators of the business entity When the training process is completed, the determined coefficient vectors of the cubic Bezier curve are output; otherwise, A is returned.
  • determining the input vector of the cubic Bezier curve according to the normalized value of each index data of the evaluation index includes: combining elements in the input vector that correspond to each index data of the evaluation index It is set to the normalized value of each index data of the evaluation index, and other elements of the input vector are set to 0.
  • the adjusting the position of the control point in the cubic Bezier curve includes: using a steepest gradient descent method or a gradient descent method to adjust the position of the control point in the cubic Bezier curve.
  • processing the specific data of the business entity according to the carbon saving amount corresponding to the business entity includes: allocating virtual data matching the carbon saving amount to the business entity according to the carbon saving amount corresponding to the business entity. thing.
  • the above method may further include: processing the business data of the business entity according to the carbon saving amount corresponding to each of the evaluation indicators.
  • the evaluation index includes at least two of the green business evaluation index, the green operator evaluation index, the green block evaluation index, the green map evaluation index, and the green user evaluation index.
  • One or more embodiments of this specification also disclose a data processing device, including: a data acquisition module for acquiring index data corresponding to each evaluation index of an operating entity according to at least two evaluation indexes; determining the amount of carbon saving The module is used to determine the carbon saving amount corresponding to each evaluation index according to the index data of each evaluation index; the ranking fusion module is used to fuse the carbon saving amount corresponding to each evaluation index using the trained ranking fusion model to obtain the The carbon saving amount corresponding to the business entity; wherein the ranking fusion model uses the index data of the evaluation indicators and the corresponding carbon saving amount as training data to be obtained through training; and the business processing module is used for operating according to the The carbon saving amount corresponding to the entity processes the specific data of the business entity; wherein, the specific data is related to the carbon saving amount.
  • the carbon saving amount determining module includes: a carbon saving amount determining unit, which is used to determine the carbon saving amount corresponding to each index data of an evaluation index; a normalization unit, which is used to separately compare each index data The corresponding carbon saving amount is normalized; the fusion unit is used to fuse the carbon saving amount corresponding to the various index data to obtain the carbon saving amount corresponding to the evaluation index.
  • the fusion unit includes: an importance determination sub-module for determining the importance of each index data by using at least one of hierarchical analysis, principal component analysis, and anomaly detection; and a weight value setting sub-module for According to the importance of each index data, the corresponding weight value is assigned to each index data; among them, the greater the importance of the index data, the greater the weight value; the summation sub-module is used to calculate the weight value according to the set weight value
  • the carbon saving amount corresponding to each index data of the evaluation index is weighted and summed to obtain the carbon saving amount corresponding to the evaluation index.
  • the sorting fusion model is realized by a cubic Bezier curve; wherein, the input vector of the cubic Bezier curve is the normalized value of the carbon saving amount corresponding to the index data of each evaluation index And the output of the cubic Bezier curve is the carbon saving amount corresponding to the business entity.
  • the above-mentioned data processing device may further include: a ranking fusion model training module for determining each coefficient vector of the cubic Bezier curve through training; wherein, the ranking fusion model training module includes: an initialization unit for initializing three times The endpoints and control points of the square Bezier curve; the training unit is used to perform the following for each evaluation index: A, the above-mentioned cubic square bee is determined according to the normalized value of the carbon saving amount corresponding to the index data of the evaluation index The input vector of the Zell curve, and the carbon saving amount corresponding to the evaluation index is taken as the known output of the cubic Bezier curve; B, it is determined that the input vector is on the cubic Bezier curve C, determine the error of this training according to the projection and the known output; D, in response to the error being greater than the preset error threshold, adjust the control point of the cubic Bezier curve Position, and return to B; E, in response to the error being less than or equal to the preset error threshold, when the training process has been performed on all the evaluation indicators of the business
  • the business processing module allocates virtual items matching the carbon saving amount to the business entity according to the carbon saving amount corresponding to the business entity.
  • the business processing module further processes the business data of the business entity according to the carbon saving amount corresponding to each evaluation index.
  • One or more embodiments of this specification also propose an electronic device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor, and the processor implements the above data processing method when the program is executed. .
  • One or more embodiments of this specification also propose a non-transitory computer-readable storage medium that stores computer instructions, and the computer instructions are used to make the computer execute the above-mentioned data processing method .
  • the data processing methods, devices, electronic equipment, and storage media provided by one or more embodiments of this specification can summarize various types of fragmented data related to business entities, and based on the summary
  • the various types of data used to determine the amount of carbon savings corresponding to the business activities and daily behaviors of the business entities are determined, so as to guide the business entities to take the initiative to save energy and reduce emissions and protect the environment.
  • This method will enable business entities to more intuitively know what environmental protection level their business activities belong to, without the need to inquire and calculate by themselves, which is more convenient for business entities.
  • the corresponding business provider can perform data and business processing methods such as accumulating points, upgrading the environmental protection level of the business entity, and providing corresponding rights and interests to the business entity based on the determined carbon savings, and compare the corresponding business with the carbon savings of the business entity. Make associations to guide and drive more business entities to pay attention to low-carbon operations, join energy conservation and environmental protection actions, promote each other, enhance stickiness, and create a green e-commerce and financial platform.
  • the ranking fusion model used in one or more embodiments of this specification is an unsupervised ranking model, and is not limited to linear ranking fusion, but learning the corresponding linear and non-linear ranking from the structure of the data itself.
  • the method is a method that supports nonlinear fusion, and the sorting and scoring results are more reasonable and objective through machine learning. Therefore, the carbon savings corresponding to the operating entities obtained through this sorting and fusion model is also very objective and accurate.
  • FIG. 1 is a schematic flowchart of a data processing method according to one or more embodiments of this specification
  • FIG. 2 is a schematic flow chart of the method for determining the carbon saving amount corresponding to an evaluation index according to one or more embodiments of this specification;
  • FIG. 3 is a schematic flowchart of the training method of the ranking fusion model according to one or more embodiments of this specification;
  • FIG. 4 is a schematic diagram of an application scenario for implementing the foregoing data processing method according to one or more embodiments of this specification;
  • Figure 5 is a schematic diagram of the green components of the business entity determined according to the carbon saving corresponding to the five evaluation indicators of green business, green business operator, green block, green map, and green user according to an embodiment of the specification. ;as well as
  • FIG. 6 is a schematic diagram of the internal structure of the data processing device according to one or more embodiments of this specification.
  • one or more embodiments of this specification propose a data processing method that can determine the amount of carbon savings corresponding to the business entity based on the relevant information of the business entity, and can also determine the amount of carbon savings corresponding to the business entity based on the corresponding information of the business entity.
  • the amount of carbon saving deals with the business of the operating entity, thereby guiding the operating entity to consciously and proactively conduct green operations and contribute to environmental protection.
  • the above-mentioned carbon saving amount can refer to the amount of reducing carbon emissions.
  • the amount of carbon saving can specifically refer to the attributes or business behaviors of the business entity corresponding to the one or more data obtained after quantifying one or more data of the business entity.
  • the resulting reduction in carbon emissions corresponds to a numerical value.
  • FIG. 1 is a schematic flowchart of a data processing method according to one or more embodiments of this specification. This method can be executed by an application server. As shown in FIG. 1, the data processing method includes steps 102 to 108.
  • step 102 index data corresponding to each evaluation index of the business entity is obtained.
  • the above-mentioned evaluation index is a set of one or more index data set in advance used to characterize a certain aspect of a business entity.
  • green management level evaluation for business entities the following five evaluation indicators can be set: green management, green operators, green blocks, green maps, and green users. These five evaluation indicators respectively represent the collection of indicator data of the operating entities in the five dimensions of green operations, green operators, green blocks, green maps, and green users. These indicator data may be data related to the energy saving or carbon emission reduction behavior of the operating entity among all the relevant data of the above-mentioned operating entity.
  • the green business evaluation indicators can specifically include data related to the business entity itself, that is, the business entity’s own registration data and business data, etc., and can include the following indicator data: the business entity’s business scale, business qualifications, business stability, and line Office and energy consumption, etc.
  • Green business operator evaluation indicators can specifically include data related to the business entity’s business operator, that is, the attribute data and behavioral data of the business entity business operator, etc., and can include the following indicator data: the business’s green travel data, the electronic coupons used Data, data on electronic payment of living expenses, data on electronic payments, data on the provision or use of environmentally friendly tableware, etc.
  • the above-mentioned behavioral data of the operator may be data generated when the operator uses the Internet service.
  • the data may also include Internet service identification information, thereby marking the source of the data.
  • Green block evaluation indicators can specifically include data related to the industry or region of the business entity, that is, industry-related data related to the business scope of the business entity and geographical related data related to the location of the business entity, etc., and can include the following indicators Data: Green rating data for operating industries and green rating data for operating regions.
  • the green map evaluation index may specifically include data related to other business entities associated with the business entity, and may include the following indicator data: green rating data of the associated business entity.
  • Green user evaluation indicators can specifically include attribute data and behavioral data of users (for example, consumers) associated with the business entity, etc., and can include the following indicator data: user's green travel data, electronic coupon data used, and electronic means Data on living payment, data on the use of environmentally friendly tableware, etc.
  • the above-mentioned user behavior data may be data generated when the user uses the Internet service.
  • the data may also include Internet service identification information, thereby marking the source of the data.
  • the above indicator data relates to all aspects that can measure the contribution of an operating entity in terms of energy conservation and emission reduction.
  • the index data of the above-mentioned multiple evaluation indexes are fragmented, and it is difficult to integrate them together through simple methods to comprehensively measure the contribution of business entities to environmental protection. Therefore, in the embodiments of this specification, the evaluation index used to measure the characteristics of the business entity in a certain aspect is first set, and then the index data included in each evaluation index is set.
  • the purpose of this setting is mainly to integrate the relevant index data together to measure the characteristics of the business entity in one aspect; then merge the characteristics of the business entity in all aspects to comprehensively measure the characteristics of the business entity.
  • the fusion of data is more reasonable, and the sorting results obtained are more objective.
  • each evaluation index listed in the above-mentioned embodiment and the index data contained in each evaluation index are merely examples, and the technical solution of this specification is not limited to the above-listed evaluation indexes and index data.
  • the addition, change or deletion of the above evaluation indicators and indicator data will not exceed the scope of protection of the embodiments of this specification.
  • the above-mentioned data can all be identified using the business entity identification (ID) that indicates the identity of the business entity, so as to illustrate the relevant data of which business entity.
  • ID business entity identification
  • the application server that executes the data processing method can obtain the above-mentioned index data related to the business entity from the server that manages the business entity, and the aforementioned application server can also obtain the data from the business entity's application client.
  • the application client, the operator of the business entity, third-party applications that implement various applications, and the database of the third-party application server collect the relevant data of the aforementioned business entities.
  • the identification of the user and operating entity associated with the above-mentioned operating entity can be obtained through the relationship data of the above-mentioned operating entity.
  • step 104 the carbon saving amount corresponding to each evaluation index is determined according to the index data of each evaluation index.
  • Fig. 2 is a schematic flow chart of a method for determining a carbon saving amount corresponding to an evaluation index according to one or more embodiments of this specification. As shown in FIG. 2, the above method may include step 202 to step 206.
  • step 202 the carbon saving amount corresponding to each index data of the above-mentioned evaluation index is determined respectively.
  • the quantification algorithm of the carbon saving amount corresponding to each index data should be preset, so that the carbon saving amount corresponding to each index data can be determined according to the preset carbon saving quantification algorithm.
  • the adopted carbon saving quantification algorithm may be the same or different.
  • the indicator data of the user’s green travel data in the green user evaluation index and the indicator data of the green travel data in the green operator’s evaluation index can be based on the user’s
  • the number of steps or distance walked determines the amount of carbon saved corresponding to the indicator data of the green travel data.
  • the carbon saving amount corresponding to the indicator data of green travel data can also be determined according to the number of times and/or distance the user takes public transportation (bus or subway or shared bicycle). Among them, the greater the number of walking steps or the number of times of taking public transportation, the greater the corresponding carbon saving; or the longer the walking distance or the distance of taking public transportation, the greater the corresponding carbon saving.
  • the walking distance and the carbon generated by ordinary vehicles within a unit distance can be determined.
  • the product of emissions is used as the carbon saving corresponding to this indicator data. This is because users can reduce their own driving trips by walking and taking public transportation, so the carbon savings of users are the carbon emissions brought by driving trips.
  • the carbon emissions produced by ordinary vehicles per unit distance can be determined based on the average fuel consumption per unit distance of ordinary vehicles and the carbon emission coefficients of energy such as gasoline and diesel. For example, take a car as an ordinary transportation bus, and assume that a car will consume an average of 0.1 liters of gasoline per kilometer. Since the gasoline carbon emission coefficient of gasoline is 2.361kg CO2/L, the carbon emission per kilometer of a car is 0.2361kg. In this way, it can be determined that each user walks 1 kilometer and the carbon saved is 0.2361 kilograms.
  • the corresponding data of the two index data can be determined according to the amount of paper products saved by the user. Carbon saving. Among them, the more electronic coupons used and the more electronic payment for living expenses, the greater the corresponding carbon saving.
  • the amount of carbon saving corresponding to the two indicator data can also be determined according to the amount of paper products saved by the user. Among them, the more online office work (such as online leave, online reimbursement, etc.), the greater the corresponding carbon saving.
  • the product of the number of paper products saved by the user and the carbon emission required to produce a unit number of paper products can be used as these The amount of carbon saved corresponding to the indicator data. This is because users can avoid printing paper products (paper receipts, paper receipts, paper payment voucher receipts, consumer receipts, paper coupons, etc.) by using online office, electronic payment and electronic coupons, etc. Therefore, the amount of carbon saved by users is the amount of carbon emissions brought about by the production of these paper products.
  • the carbon emissions required to produce a unit number of paper products are different in different regions, so the carbon emissions required to produce a unit number of paper products can be determined according to the emission intensity of bill paper production in a certain region; also The carbon emissions required to produce a unit number of paper products can be determined based on the average emission intensity required to produce a unit number of paper products.
  • these data indicators can also be based on the user’s location when performing business online.
  • the distance between the location and the nearest business location to determine the amount of carbon savings corresponding to these data indicators can be used as the carbon saving amount corresponding to these index data.
  • green travel indicator data For specific methods, please refer to the above description of green travel indicator data.
  • the carbon saving amount corresponding to this indicator data can be determined according to the reduction in the amount of white waste generated or the amount of paper products saved. Specifically, the product of the amount of white garbage generated and the carbon emission generated by incineration or processing of a unit amount of white garbage can be used as the carbon saving amount corresponding to this indicator.
  • the carbon saving corresponding to this indicator data can be determined according to the average energy consumption of the business entity.
  • the larger the average energy consumption the smaller the carbon saving corresponding to this indicator data.
  • the average energy consumption of business entities of the same industry and the same scale can be counted in advance, and then the difference between the above average energy consumption and the energy consumption of the business entity can be calculated, and the difference between the above difference and the corresponding energy
  • the product of the carbon emission coefficient is used as the carbon saving amount corresponding to the above indicator data.
  • the above difference AB can be calculated based on the average power consumption B per unit time of other business entities of the same size in the same industry, and then based on each kWh
  • the carbon emission coefficient corresponding to electricity is the product of the above difference and the above carbon emission coefficient as the carbon saving amount corresponding to the indicator data of the electricity consumption of the business entity.
  • a similar method can be used to determine the water consumption per unit time of an operating entity.
  • this indicator can be determined according to the green level of the operating industry or the green level of the operating area corresponding to the operating entity The amount of carbon savings corresponding to the data. Among them, the higher the green rating of the operating industry and the higher the green rating of the operating area, the greater the amount of carbon savings corresponding to these two indicator data.
  • the carbon saving amount corresponding to this indicator data can be determined according to the green rating of the business entity associated with the business entity.
  • the higher the green rating of the business entities associated with the green business industry, and the greater the number of related business entities that reach a certain green level the greater the amount of carbon savings corresponding to this indicator data.
  • a weighted sum or weighted average of the green ratings of other business entities managed by one business entity can be performed, and the carbon saving amount corresponding to this indicator data can be determined according to the calculation result.
  • step 204 the carbon saving amount corresponding to the above-mentioned index data is normalized.
  • the embodiments of this specification do not limit the normalization method of each index data.
  • the embodiments of the present specification can also directly normalize the carbon saving amount corresponding to the above-mentioned index data according to a commonly used normalization function, such as the Sigmoid function.
  • step 206 the carbon saving amount corresponding to the above-mentioned normalized index data is merged to obtain the carbon saving amount corresponding to the above-mentioned evaluation index.
  • the carbon saving amount corresponding to each index data of an evaluation index can be directly summed or averaged, and the calculation result is used as the carbon saving amount corresponding to the evaluation index.
  • a weight value may be set for each index data of the evaluation index in advance, and the carbon saving amount corresponding to each index data of the above evaluation index may be weighted according to the set weight value. Sum/weighted average, etc., to obtain the carbon saving corresponding to the above-mentioned evaluation index.
  • multiple methods can be used to set the weight value of each index data of an evaluation index. For example, you can set a fixed weight value in advance; you can also use one of many methods such as analytic hierarchy process (AHP), principal component analysis (PCA), and anomaly detection (specifically, isolated forest algorithm) to determine the importance of each index data. Then, according to the importance of each index data, a corresponding weight value is assigned to each index data. Among them, the greater the importance of the index data, the larger the corresponding weight value.
  • AHP analytic hierarchy process
  • PCA principal component analysis
  • anomaly detection specifically, isolated forest algorithm
  • step 106 the trained ranking fusion model is used to fuse the carbon saving amount corresponding to each evaluation index to obtain the carbon saving amount corresponding to the business entity.
  • the above-mentioned trained ranking fusion model is obtained through training using the index data of each of the above-mentioned evaluation indexes and the corresponding carbon saving amount as training data.
  • the above-mentioned training may be training data based on the data of multiple evaluation indicators of the business entity itself and the corresponding carbon saving amount.
  • the above-mentioned training can also be based on the data of multiple evaluation indicators of multiple different business entities and their corresponding carbon savings as training data, and the trained ranking fusion model can be applied to different business entities. The amount of carbon.
  • the above-mentioned sorting fusion model can be implemented by a cubic Bezier curve. It can be seen that the cubic Bezier curve can be expressed by the following calculation expression (1).
  • t is the input vector, which can be specifically a vector composed of the normalized values of the index data included in the evaluation indexes of the above-mentioned business entity. That is, the dimension of the input vector may be the number of items from the index data.
  • the above training process is to use the vector determined according to the normalized value of the index data contained in each evaluation index of the business entity as the input vector, and the carbon saving corresponding to each evaluation index as the known output.
  • the coefficient vectors P 0 , P 1 , P 2 and P 3 of the above cubic Bezier curve are determined. The specific training method will be described in detail later, so I will skip it here.
  • the vector consisting of the normalized values of the index data contained in the evaluation indicators of the business entity can be used as the input vector t, and it is determined that the input vector t is
  • the projection on the cubic Bezier curve, that is, the calculation result B(t) of the cubic Bezier curve shown in the calculation expression (1), is used as the carbon saving amount corresponding to the business entity.
  • the above cubic Bezier curve can be regarded as an unsupervised ranking fusion model RPC (Ranking Principal Curve).
  • RPC model conforms to five meta-rules for evaluating unsupervised ranking results and guiding the design of ranking functions.
  • These meta-rules include: (1) Translatability: for different data dimensions, ranking points The value does not change; (2) Strict monotonicity: Assuming that for the green management sub-module, the higher the sub-module score, the higher the score of the final comprehensive sorting, which requires strict monotonicity. (3) Linear and non-linear compatibility: Assuming that the green management module has a linear relationship with the final comprehensive ranking score, then other modules and the final ranking score are also linear; assuming a non-linear relationship, other modules are also non-linear Relationship, otherwise there will be model deviations.
  • the above-mentioned RPC sorting method proposed in the embodiment of this specification is not limited to linear sorting fusion, but learning the corresponding linear and non-linear sorting methods from the structure of the data itself, which belongs to a way to support non-linear fusion. Moreover, the sorting results are more reasonable and objective through machine learning.
  • step 108 the specific data of the business entity is processed according to the carbon saving amount corresponding to the business entity.
  • the above-mentioned specific data is data related to the amount of carbon saving.
  • the above-mentioned processing may include statistics and analysis of the carbon saving amount of the business entity, converting the carbon saving amount into the form of points, or may also convert the carbon saving amount corresponding to the business entity into a value of 0-100. According to experience, it is divided into multiple stars. The higher the star rating, the greener and more environmentally friendly the business entity is.
  • the business entity can also provide corresponding rights and interests for the business entity based on the green star rating or points corresponding to the business entity.
  • the foregoing processing may include: allocating virtual items matching the foregoing carbon saving amount to the business entity according to the carbon saving amount corresponding to the business entity.
  • the business provider can perform processing such as accumulation of points, upgrade of the level of the business entity, and provision of corresponding rights and interests on the data of the business entity based on the carbon savings corresponding to the determined business entity.
  • processing such as accumulation of points, upgrade of the level of the business entity, and provision of corresponding rights and interests on the data of the business entity based on the carbon savings corresponding to the determined business entity.
  • the corresponding business is associated with the carbon savings of the operating entity.
  • the business provider can further process specific data of the business entity based on the carbon savings corresponding to each evaluation index of the business entity, such as the accumulation of points, the upgrade of the business entity level, and the provision of corresponding rights and interests, etc. Wait.
  • the green component of the operating entity can be determined according to the carbon saving amount corresponding to the above-mentioned at least two evaluation indicators, and the green component diagram can be generated and displayed to the operating entity, so that the operating entity can clearly understand its own environmental protection evaluation dimensions.
  • the amount of carbon savings can also provide the operating entity with corresponding rights and interests based on the green component of the operating entity. For example, it can allocate virtual items that match the above-mentioned carbon savings to the operating entity based on the carbon savings corresponding to the above various evaluation indicators of the operating entity. .
  • the data processing method can aggregate various types of data fragmented by business entities, and determine the business entity based on the aggregated various types of data The amount of carbon saved, so as to guide the operating entities to take the initiative to save energy and reduce emissions.
  • This method will enable business entities to more intuitively know what green level their actions belong to, without the need to inquire and calculate by themselves, which is more convenient for business entities.
  • the corresponding business provider can further perform data processing methods such as accumulation of points, upgrade of the business entity level, and provision of corresponding rights and interests on the operating entity based on the determined carbon savings, and associate the corresponding business with the carbon savings of the operating entity. This will guide and drive more business entities to pay attention to low-carbon operations, join energy conservation and environmental protection actions, promote each other, enhance stickiness, and create a green e-commerce and financial platform.
  • the ranking fusion model used in one or more embodiments of this specification is an unsupervised ranking model, and is not limited to linear ranking fusion, but learning the corresponding linear and non-linear ranking from the structure of the data itself.
  • the method is a method that supports nonlinear fusion, and the sorting and scoring results are more reasonable and objective through machine learning. Therefore, the carbon savings corresponding to the operating entities obtained through this sorting and fusion model is also very objective and accurate.
  • Fig. 3 shows the training method of the ranking fusion model described in one or more embodiments of this specification. As shown in FIG. 3, the training method includes steps 302 to 316.
  • step 302 the endpoints and control points of the cubic Bezier curve are initialized.
  • the end points P 0 and P 3 and the control points P 0 and P 3 of the above-mentioned Bezier curve can be initialized based on experience.
  • the endpoints P 0 and P 3 of the cubic Bezier curve can be initialized by the following calculation expressions (2) and (3).
  • the above ⁇ is a preset endpoint control parameter.
  • the above ⁇ can be set based on experience.
  • the above-mentioned control point can also be initialized by the above-mentioned similar method based on experience.
  • step 304 the input vector of the cubic Bezier curve is determined according to the normalized value of the carbon saving corresponding to each index data of the above evaluation index, and the carbon saving corresponding to the above evaluation index is taken as the above three The known output of the square Bezier curve.
  • an input vector of the cubic Bezier curve can be determined according to the normalized value of each index data of an evaluation index.
  • the elements corresponding to the various index data of the evaluation index in the above input vector are the normalized values of the various index data of the evaluation index, and other elements can be set to 0.
  • step 306 the projection of the input vector on the cubic Bezier curve is determined.
  • the above-mentioned projection refers to the above-mentioned input vector as the input vector t in the above-mentioned calculation expression, and the calculation result B(t) of the cubic Bezier curve shown in the above-mentioned calculation expression (1) It is the projection of the above-mentioned input vector on the above-mentioned cubic Bezier curve.
  • step 308 the error of this training is determined according to the above projection and the above known output.
  • the above-mentioned error may be the distance between the above-mentioned projection and the corresponding known output.
  • step 310 the above-mentioned error is compared with a preset error threshold. If the above-mentioned error is greater than the preset error threshold, step 312 is continued; otherwise, the current round of training is ended, and step 314 is continued.
  • step 312 adjust the position of the control point in the cubic Bezier curve, and then return to step 306.
  • the method for adjusting the position of the control point in the cubic Bezier curve may include: using the steepest gradient descent method to adjust the position of the control point in the cubic Bezier curve or using a gradient descent method Adjust the position of the control points in the cubic Bezier curve.
  • the moving unit step size used in the steepest gradient descent method may be greater than the moving unit step size used in the gradient descent method. Therefore, when the above error is large, the fastest gradient descent method can be used to adjust the position of the control point in the cubic Bezier curve to achieve the purpose of rapid training convergence; and when the above error is small, the gradient descent method can be used to adjust The position of the control point in the cubic Bezier curve mentioned above to avoid the shock effect during training.
  • step 314 it is determined whether the above-mentioned training process has been performed for all the evaluation indicators of the above-mentioned business entity, if so, step 316 is performed; otherwise, the above-mentioned step 304 is returned.
  • step 316 the determined coefficient vectors P 0 , P 1 , P 2 and P 3 of the aforementioned cubic Bezier curve are output.
  • the above-mentioned scoring rules for each evaluation index can be integrated by fitting the cubic Bezier curve according to the carbon saving amount corresponding to each index data of the above evaluation index and the corresponding carbon saving amount. , Get a scoring result that integrates the index data of all the evaluation indexes.
  • Figure 4 shows an application scenario for implementing the above-mentioned data processing method.
  • the application server that executes the above-mentioned data processing method is connected to the application client of the business entity.
  • the above-mentioned application server may obtain various index data of the above-mentioned multiple evaluation indexes from an operator, a third-party application, and a third-party application server through the application client.
  • the above-mentioned application server may also obtain various index data of the above-mentioned multiple evaluation indexes from other servers (not shown) in its own system, for example, from a business entity management server. After acquiring the index data of the multiple evaluation indexes, the application server may execute the data processing method to obtain the carbon saving amount corresponding to the business entity.
  • multiple evaluation indicators that can be used to evaluate the green management level of the business entity can be predefined, for example, the above five green business, green business operator, green block, green map, and green user. Evaluation index. In addition, it is also necessary to predefine the index data of these five evaluation indicators.
  • the evaluation index of green operation includes: the energy consumption of the business entity itself, the scale of operation and other index data;
  • the evaluation index of the green business operator includes: Indicator data such as green travel data, electronic payment or electronic payment data;
  • the evaluation indicator of Green Atlas includes: indicator data such as the green management level of other operating entities closely related to the operating entity;
  • the evaluation indicator of green block includes : Indicator data such as industry green rating data related to business entities, and regional green rating data related to business entities;
  • the evaluation indicators of green users include: green travel data of users associated with business entities, and electronic Index data such as coupon data, data on electronic payment of living expenses, data on the use of environmentally friendly tableware, and so on.
  • the isolated forest algorithm of anomaly detection can be used to segment the index data of the evaluation index, and the importance of each index data can be determined according to the number of times the index data is segmented (corresponding to the depth of the node) The index data with less number of times of segmentation can be considered to have higher importance.
  • an input vector can be generated according to the normalized value of the carbon saving corresponding to all the above index data, and input the trained cubic Bezier curve to obtain the above input vector on the cubic Bezier curve. And use the value corresponding to the projection as the carbon saving amount corresponding to the business entity.
  • the corresponding carbon savings can be measured.
  • the specific data of the business entity has been processed.
  • the business entity can be rated for green operation according to the carbon saving amount of the business entity, and the green star rating of the business entity can be obtained. And it can also provide the business entity with corresponding rights based on its green star rating, such as virtual items that match the carbon saving or green star rating.
  • the green component of the business entity can be determined based on the carbon savings of the business entity on the five evaluation indicators of green business, green business operator, green block, green map, and green user, and can be visualized Generate a schematic diagram of the green components of the business entity as shown in Figure 5 and provide it to the user to guide the business entity's business behavior.
  • the business entity through the schematic diagram of the green components of a business entity as shown in Figure 5, it can be informed that the business entity’s carbon savings in the four aspects of green operations, green operators, green blocks, and green users are relatively high or relatively balanced.
  • the amount of carbon saved in the green map is relatively low, which can guide the daily behavior of business entities in more detail.
  • after determining the green star rating and green components of the operating entity it can also provide the operating entity with corresponding rights and interests, and further encourage the operating entity to save energy and reduce emissions, so as to actively contribute to environmental protection.
  • a data processing device includes: a data acquisition module 602, a carbon saving amount determination module 604, a ranking fusion module 606, and a business processing module 608.
  • the data acquisition module 602 is configured to acquire index data corresponding to each evaluation index of the business entity according to at least two evaluation indexes.
  • the carbon saving amount determining module 604 is used to determine the carbon saving amount corresponding to each evaluation index according to the index data of each evaluation index.
  • the ranking fusion module 606 is configured to use the trained ranking fusion model to fuse the carbon saving amount corresponding to each evaluation index to obtain the carbon saving amount corresponding to the business entity; wherein the ranking fusion model uses the respective evaluation index
  • the index data and the corresponding carbon saving amount are obtained through training as training data.
  • the business processing module 608 is configured to process the specific data of the business entity according to the carbon saving amount corresponding to the business entity; wherein, the above specific data is related to the carbon saving amount.
  • the above-mentioned carbon saving amount determining module 604 includes: a carbon saving amount determining unit for determining the carbon saving amount corresponding to each index data of an evaluation index; a normalization unit, It is used to normalize the carbon saving amount corresponding to each index data; the fusion unit is used to fuse the carbon saving amount corresponding to each index data of the evaluation index to obtain the carbon saving corresponding to the evaluation index quantity.
  • the aforementioned fusion unit includes: an importance determination sub-module for determining the importance of the aforementioned index data by using at least one of hierarchical analysis, principal component analysis, and anomaly detection;
  • the weight value setting sub-module is used to assign corresponding weight values to each index data according to the importance of each index data; among them, the greater the importance of the index data, the greater the corresponding weight value;
  • the sum sub-module uses The carbon saving amount corresponding to each index data of the evaluation index is weighted and summed according to the set weight value to obtain the carbon saving amount corresponding to the evaluation index.
  • the above-mentioned sorting fusion model is implemented by a cubic Bezier curve; wherein, the input vector of the cubic Bezier curve is normalized by the carbon saving amount corresponding to the index data And the output of the cubic Bezier curve is the carbon saving amount corresponding to the business entity.
  • the above-mentioned data processing device may further include: a ranking fusion model training module, configured to determine each coefficient vector of the cubic Bezier curve through training; wherein, the above-mentioned ranking fusion model
  • the training module includes: an initialization unit, which is used to initialize the endpoints and control points of the cubic Bezier curve; a training unit, which is used to perform respectively for each evaluation index: A, according to the corresponding index data of the evaluation index The normalized value of the carbon saving amount determines the input vector of the cubic Bezier curve, and the carbon saving amount corresponding to the evaluation index is used as the known output of the cubic Bezier curve; B, confirm The projection of the input vector on the cubic Bezier curve; C, the error of this training is determined according to the projection and the known output; D, in response to the error being greater than a preset error threshold Situation, adjust the position of the control point in the cubic Bézier curve and return to B; E, in response to the error being less than or equal to the preset error threshold
  • the above-mentioned data processing apparatus may be regarded as an electronic device. Therefore, the internal structure of the above-mentioned data processing apparatus may be as shown in FIG. 6, including: a processor 610, a memory 620, an input/output Interface 630, communication interface 640, and bus 650. Among them, the processor 610, the memory 620, the input/output interface 630, and the communication interface 640 implement communication connections between each other in the device through the bus 650.
  • the memory 620 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc.
  • the memory 620 can store an operating system and other application programs, and can also store various modules of the above-mentioned data processing apparatus provided in the embodiments of this specification.
  • the relevant program codes are stored In the memory 620, the processor 610 calls for execution.
  • the above-mentioned processor 610 may be implemented by a general CPU (Central Processing Unit, central processing unit), microprocessor, application specific integrated circuit (Application Specific Integrated Circuit, ASIC), or one or more integrated circuits for execution. Relevant procedures to realize the technical solutions provided in the embodiments of this specification.
  • the input/output interface 630 can be used to connect an input/output module to realize information input and output.
  • the input/output/module can be configured in the device as a component (not shown in the figure), or it can be connected to the device to provide corresponding functions.
  • the input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and an output device may include a display, a speaker, a vibrator, an indicator light, and the like.
  • the communication interface 640 is used to connect a communication module (not shown in the figure) to realize the communication interaction between the device and other devices.
  • the communication module can realize communication through wired means (such as USB, network cable, etc.), or through wireless means (such as mobile network, WIFI, Bluetooth, etc.).
  • the bus 650 includes a path to transmit information between various components of the device (for example, a processor, a memory, an input/output interface, and a communication interface).
  • the device may also include other components necessary for normal operation.
  • the above-mentioned device may also include only the components necessary to implement the solutions of the embodiments of the present specification, and not necessarily include all the components shown in the figures.
  • the technical carriers involved in payment in the embodiments of this specification may include, for example, Near Field Communication (NFC), WIFI, 3G/4G/5G, POS machine swiping technology, QR code scanning technology, and barcode scanning technology.
  • NFC Near Field Communication
  • WIFI Wireless Fidelity
  • 3G/4G/5G 3G/4G/5G
  • POS machine swiping technology 3G/4G/5G
  • QR code scanning technology QR code scanning technology
  • barcode scanning technology e.g., Bluetooth, infrared, Short Message Service (SMS), Multimedia Message Service (MMS), etc.
  • SMS Short Message Service
  • MMS Multimedia Message Service
  • the biological characteristics involved in the biometric identification in the embodiments of this specification may include, for example, eye characteristics, voice prints, fingerprints, palm prints, heartbeats, pulses, chromosomes, DNA, human tooth bite marks, and the like.
  • the eye patterns can include biological features such as iris and sclera.
  • the methods in one or more embodiments of this specification can be executed by a single device, such as a computer or a server.
  • the method in this embodiment can also be applied in a distributed scenario, and multiple devices cooperate with each other to complete.
  • one of the multiple devices can only perform one or more steps in the method of one or more embodiments of this specification, and the multiple devices will perform each other. Interact to complete the described method.
  • the apparatus of the foregoing embodiment is used to implement the corresponding method in the foregoing embodiment, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.
  • the computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be realized by any method or technology.
  • the information can be computer-readable instructions, data structures, program modules, or other data.
  • Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, CD-ROM, digital versatile disc (DVD) or other optical storage, Magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media can be used to store information that can be accessed by computing devices.
  • PRAM phase change memory
  • SRAM static random access memory
  • DRAM dynamic random access memory
  • RAM random access memory
  • ROM read-only memory
  • EEPROM electrically erasable programmable read-only memory
  • flash memory or other memory technology
  • the accompanying drawings may or may not be shown in relation to integrated circuit (IC) chips and other components.
  • IC integrated circuit
  • Well-known power/ground connection IC
  • the device may be shown in the form of a block diagram in order to avoid making one or more embodiments of this specification difficult to understand, and this also takes into account the fact that the details about the implementation of these block diagram devices are highly dependent on the implementation of the present invention. Description of the platform of one or more embodiments (that is, these details should be fully within the understanding of those skilled in the art).
  • DRAM dynamic RAM

Abstract

一种数据处理方法，以及实现所述方法的数据处理装置、电子设备及存储介质，该方法包括：按照至少两个评价指标，获取经营实体对应于每个评价指标的指标数据（102）；根据每个评价指标的指标数据确定每个评价指标对应的节碳量（104）；利用经过训练的排序融合模型对各评价指标对应的节碳量进行融合，得到所述经营实体对应的节碳量（106）；其中，所述排序融合模型以所述各评价指标的指标数据及其对应的节碳量为训练数据通过训练得到；根据所述经营实体对应的节碳量对所述经营实体的特定数据进行处理（108）；其中特定数据与节碳量相关。

Description

数据处理方法、装置、电子设备及存储介质 技术领域
本说明书一个或多个实施例涉及计算机技术领域，尤其涉及一种数据处理方法、装置、电子设备及计算机可读存储介质。
背景技术
目前，由于温室气体造成的温室效应正导致全球气候日益变暖，已经并将继续为地球和人类带来灾难。为了保护地球生态环境，避免温室效应进一步恶化，节约能源，减少碳排放，实现绿色经营、绿色生活已经成为全球众多国家社会经济发展和生产经营活动的重要目标之一。换言之，如何鼓励经营实体或个人主动进行节能减排是全人类需要努力的方向。而其中很关键的一环就是让经营实体能够清楚了解自己的经营活动以及日常行为是否属于节能减排的绿色经营行为，从而可以进一步指导经营实体自觉主动地进行绿色经营，为环境保护做贡献。
发明内容
有鉴于此，本说明书一个或多个实施例提出一种数据处理方法，可以根据经营实体的相关信息确定经营实体对应的节碳量，并根据经营实体对应的节碳量对经营实体的特定数据进行处理，从而指导经营实体自觉主动地进行节能减排、绿色经营，为环境保护做贡献。
本说明书提出的数据处理方法包括：按照至少两个评价指标，获取经营实体对应于每个评价指标的指标数据；根据每个评价指标的指标数据确定每个评价指标对应的节碳量；利用经过训练的排序融合模型对各评价指标对应的节碳量进行融合，得到所述经营实体对应的节碳量；其中，所述排序融合模型以所述各评价指标的指标数据及其对应的节碳量为训练数据通过训练得到；根据所述经营实体对应的节碳量对所述经营实体的特定数据进行处理；其中，所述特定数据与节碳量有关。
其中，根据每个评价指标的指标数据确定每个评价指标对应的节碳量包括：针对每个评价指标，分别确定所述评价指标的各项指标数据对应的节碳量；分别对所述各项指标数据对应的节碳量进行归一化；以及将归一化后的所述各项指标数据对应的节碳量进行融合，得到所述评价指标对应的节碳量。
其中，将所述各个指标数据对应的节碳量进行融合包括：对所述各项指标数据分别设置权重值；根据设置的权重值将所述各项指标数据对应的节碳量进行加权求和，得到所述评价指标对应的节碳量。
其中，对所述各项指标数据分别设置权重值包括：采用层次分析、主成分分析以 及异常检测方法中的至少一种确定所述各项指标数据的重要度；以及根据所述各项指标数据的重要度分别为所述各项指标数据分配对应的权重值；其中，重要度越大的指标数据所对应的权重值越大。
其中，所述排序融合模型通过三次方贝塞尔曲线实现；其中，所述三次方贝塞尔曲线的输入向量为由所述各个评价指标的指标数据对应的节碳量归一化后的值组成的向量；以及所述三次方贝塞尔曲线的输出为所述经营实体对应的节碳量。
其中，训练所述排序融合模型的过程包括：初始化三次方贝塞尔曲线的端点和控制点；针对每个评价指标分别执行如下步骤：A，根据所述评价指标的各项指标数据对应的节碳量归一化后的值确定上述三次方贝塞尔曲线的输入向量，并将所述评价指标所对应的节碳量作为所述三次方贝塞尔曲线的已知输出；B，确定所述输入向量在所述三次方贝塞尔曲线上的投影；C，根据所述投影以及所述已知输出确定本次训练的误差；D，响应于所述误差大于预先设置的误差阈值的情况，调整上述三次方贝塞尔曲线中控制点的位置，并返回B；E，响应于所述误差小于或等于预先设置的误差阈值的情况，在已对所述经营实体的所有评价指标均执行了所述训练过程时，输出确定的所述三次方贝塞尔曲线的各个系数向量；否则，返回A。
其中，根据所述评价指标的各项指标数据归一化后的值确定上述三次方贝塞尔曲线的输入向量包括：将所述输入向量中与所述评价指标的各项指标数据对应的元素设置为所述评价指标的各项指标数据归一化后的值，并将所述输入向量的其他元素设置为0。
其中，所述调整所述三次方贝塞尔曲线中控制点的位置包括：使用最速梯度下降法或者梯度下降法调整所述三次方贝塞尔曲线中控制点的位置。
其中，根据所述经营实体对应的节碳量对所述经营实体的特定数据进行处理包括：根据所述经营实体对应的节碳量为所述经营实体分配与所述节碳量相匹配的虚拟物品。
上述方法可以进一步包括：根据所述每个评价指标对应的节碳量对所述经营实体的业务数据进行处理。
其中，所述评价指标包括：绿色经营评价指标、绿色经营者评价指标、绿色区块评价指标、绿色图谱评价指标以及绿色用户评价指标中的至少两个。
本说明书的一个或多个实施例还公开了一种数据处理装置，包括：数据获取模块，用于按照至少两个评价指标，获取经营实体对应于每个评价指标的指标数据；节碳量确定模块，用于根据每个评价指标的指标数据确定每个评价指标对应的节碳量；排序融合模块，用于利用经过训练的排序融合模型对各评价指标对应的节碳量进行融合，得到所述经营实体对应的节碳量；其中，所述排序融合模型以所述各评价指标的指标数据及其 对应的节碳量为训练数据通过训练得到；以及业务处理模块，用于根据所述经营实体对应的节碳量对所述经营实体的特定数据进行处理；其中，所述特定数据与节碳量有关。
其中，所述节碳量确定模块包括：节碳量确定单元，用于分别确定一个评价指标的各项指标数据对应的节碳量；归一化单元，用于分别对所述各项指标数据对应的节碳量进行归一化；融合单元，用于将所述各项指标数据对应的节碳量进行融合，得到所述评价指标对应的节碳量。
其中，所述融合单元包括：重要度确定子模块，用于采用层次分析、主成分分析以及异常检测中的至少一种确定所述各项指标数据的重要度；权重值设置子模块，用于根据所述各项指标数据的重要度分别为各项指标数据分配对应的权重值；其中，重要度越大的指标数据对应的权重值越大；求和子模块，用于根据设置的权重值将所述评价指标的各项指标数据对应的节碳量进行加权求和，得到所述评价指标对应的节碳量。
其中，所述排序融合模型通过三次方贝塞尔曲线实现；其中，所述三次方贝塞尔曲线的输入向量为由所述各个评价指标的指标数据对应的节碳量归一化后的值组成的向量；以及所述三次方贝塞尔曲线的输出为所述经营实体对应的节碳量。
上述数据处理装置可以进一步包括：排序融合模型训练模块，用于通过训练确定所述三次方贝塞尔曲线的各个系数向量；其中，所述排序融合模型训练模块包括：初始化单元，用于初始化三次方贝塞尔曲线的端点和控制点；训练单元，用于针对每个评价指标分别执行：A，根据所述评价指标的指标数据对应的节碳量归一化后的值确定上述三次方贝塞尔曲线的输入向量，并将所述评价指标所对应的节碳量作为所述三次方贝塞尔曲线的已知输出；B，确定所述输入向量在所述三次方贝塞尔曲线上的投影；C，根据所述投影以及所述已知输出确定本次训练的误差；D，响应于所述误差大于预先设置的误差阈值的情况，调整上述三次方贝塞尔曲线中控制点的位置，并返回B；E，响应于所述误差小于或等于预先设置的误差阈值的情况，在已对上述经营实体的所有评价指标均执行了上述训练过程时，输出确定的所述三次方贝塞尔曲线的各个系数向量；否则，返回A。
其中，所述业务处理模块根据所述经营实体对应的节碳量为所述经营实体分配与所述节碳量相匹配的虚拟物品。
其中，所述业务处理模块进一步根据所述每个评价指标对应的节碳量对所述经营实体的业务数据进行处理。
本说明书一个或多个实施例还提出一种电子设备，包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时实现上述数据 处理方法。
本说明书一个或多个实施例还提出一种非暂态计算机可读存储介质，所述非暂态计算机可读存储介质存储计算机指令，所述计算机指令用于使所述计算机执行上述数据处理方法。
从上述技术方案可以看出，本说明书一个或多个实施例提供的数据处理方法、装置、电子设备以及存储介质，可将与经营实体相关的碎片化的各类数据进行汇总，并基于汇总后的各类数据，确定出用于评估经营实体的经营活动以及日常行为对应的节碳量，从而指导经营实体主动进行节能减排，保护环境。这样的方式将使经营实体可以更加直观地获知自身的经营行为属于怎样的环保级别，而无需自行查询、计算，对于经营实体而言较为便捷。此外，相应的业务提供方可基于确定的节碳量对经营实体进行诸如积分累计、经营实体环保等级提升以及提供相应权益等数据和业务处理方式，将相应的业务与经营实体对应的节碳量进行关联，从而引导带动更多经营实体关注低碳经营，加入节能环保行动，互相促进、增强黏性、打造绿色电商与金融平台。
更进一步，在本说明书一个或多个实施例中采用的排序融合模型是无监督的排序模型，且不拘泥于线性的排序融合，而是从数据本身结构中学习对应的线性和非线性的排序方式，属于一种支持非线性融合的方式，而且通过机器学习的方式使排序及打分结果更为合理和客观。因此，通过这种排序融合模型得到的经营实体对应的节碳量也是非常客观和准确的。
附图说明
为了更清楚地说明本说明书一个或多个实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本说明书一个或多个实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为本说明书一个或多个实施例所述的数据处理方法的流程示意图；
图2为本说明书一个或多个实施例所述的确定一个评价指标对应节碳量的方法流程示意图；
图3为本说明书一个或多个实施例所述的排序融合模型的训练方法流程示意图；
图4为本说明书一个或多个实施例所述的实现上述数据处理方法的应用场景示意图；
图5为本说明书一个实施例所述的根据经营实体在绿色经营、绿色经营者、绿色区块、绿色图谱以及绿色用户这五个评价指标对应的节碳量确定的该经营实体的绿色成 分示意图；以及
图6为本说明书一个或多个实施例所述的数据处理装置的内部结构示意图。
具体实施方式
为使本公开的目的、技术方案和优点更加清楚明白，以下结合具体实施例，并参照附图，对本公开进一步详细说明。
需要说明的是，除非另外定义，本说明书一个或多个实施例使用的技术术语或者科学术语应当为本公开所属领域内具有一般技能的人士所理解的通常意义。本说明书一个或多个实施例中使用的“第一”、“第二”以及类似的词语并不表示任何顺序、数量或者重要性，而只是用来区分不同的组成部分。“包括”或者“包含”等类似的词语意指出现该词前面的元件或者物件涵盖出现在该词后面列举的元件或者物件及其等同，而不排除其他元件或者物件。“连接”或者“相连”等类似的词语并非限定于物理的或者机械的连接，而是可以包括电性的连接，不管是直接的还是间接的。“上”、“下”、“左”、“右”等仅用于表示相对位置关系，当被描述对象的绝对位置改变后，则该相对位置关系也可能相应地改变。
如前所述，由于温室气体造成的温室效应正导致全球气候日益变暖，已经并将继续为地球和人类带来灾难。而碳排放是关于温室气体排放的一个总称或简称。人类的活动都有可能造成碳排放，比如汽车尾气造成碳排放，火电站造成碳排放等等。而节约能源、减少碳排放则应当是每个个人和经营实体努力的方向。如此，为了鼓励经营实体主动进行节能减排，本说明书一个或多个实施例提出一种数据处理方法，可以根据经营实体的相关信息确定经营实体对应的节碳量，并可以根据经营实体对应的节碳量对经营实体的业务进行处理，从而指导经营实体自觉主动地进行绿色经营，为环境保护做贡献。上述节碳量可以是指减少碳排放的量。在本说明书的实施例中，节碳量具体可以是指对经营实体的一项或多项数据进行量化处理后得到的与该一项或多项数据对应的由该经营实体的属性或经营行为导致的减少碳排放的量所对应一个数值。
图1为本说明书一个或多个实施例所述的数据处理方法的流程示意图。该方法可以由一个应用服务器执行。如图1所示，该数据处理方法包括步骤102～步骤108。
在步骤102，按照至少两个评价指标，获取经营实体对应于每个评价指标的指标数据。
在本说明书的实施例中，上述评价指标即为预先设置的用于刻画业务实体在某一方面特性的一项或多项指标数据组成的集合。
例如，在对经营实体进行绿色经营等级评估的应用中，可以设置以下5个评价指 标：绿色经营、绿色经营者、绿色区块、绿色图谱以及绿色用户。这5个评价指标分别代表经营实体在绿色经营、绿色经营者、绿色区块、绿色图谱以及绿色用户五个维度方面的指标数据所组成的集合。这些指标数据可以是上述经营实体的所有相关数据中与经营实体节约能源或减少碳排放等行为相关的数据。
其中，绿色经营评价指标具体可以包括与经营实体自身相关的数据，也即经营实体自身的注册数据和经营数据等，可以包括如下指标数据：经营实体的经营规模、经营资质、经营稳定性、线上办公以及能源消耗量等等。
绿色经营者评价指标具体可以包括与经营实体的经营者相关的数据，也即经营实体经营者的属性数据和行为数据等，可以包括如下指标数据：经营者的绿色出行数据、所使用的电子券数据、采用电子方式进行生活缴费的数据、电子收款的数据、提供或使用环保餐具的数据等等。上述经营者的行为数据可以是经营者在使用互联网服务时产生的数据，这些数据中除了标识了经营者的身份之外还可以包括互联网服务标识信息，从而标注数据的来源。
绿色区块评价指标具体可以包括与经营实体所属行业或地区相关的数据，也即与经营实体经营范围相关联的行业相关数据以及与经营实体所在地域相关联的地域相关数据等，可以包括如下指标数据：经营行业绿色评级数据以及经营地域绿色评级数据。
绿色图谱评价指标具体可以包括与经营实体相关联的其他经营实体相关的数据，可以包括如下指标数据：关联经营实体的绿色评级数据。
绿色用户评价指标具体可以包括与经营实体相关联的用户(例如，消费者)的属性数据和行为数据等，可以包括如下指标数据：用户的绿色出行数据、所使用的电子券数据、采用电子方式进行生活缴费的数据、使用环保餐具的数据等等。上述用户的行为数据可以是用户在使用互联网服务时产生的数据，这些数据中除了标识了用户的身份之外还可以包括互联网服务标识信息，从而标注数据的来源。
可以看出，上述指标数据涉及到可以衡量一个经营实体在节能减排方面所做贡献的方方面面。而且上述多项评价指标的指标数据是碎片化的，很难通过简单的方法融合在一起，以综合衡量经营实体为环保所作的贡献。因此，在本说明书的实施例中，首先设置用于衡量经营实体在某一方面的特性的评价指标，再设置各个评价指标所包含的指标数据。这样设置的目的主要在于，先将相关度大的指标数据融合在一起，衡量经营实体在一个方面的特性；然后再将该经营实体在各方面的特性融合在一起，综合衡量经营实体的特性，从而使数据的融合更为合理，得到的排序结果更为客观。
此外，上述实施例所列的各个评价指标以及各个评价指标所包含的指标数据仅是 一种举例，本说明书的技术方案并不限于上述列举出的各个评价指标以及指标数据。对上述评价指标以及指标数据的增加、变更或删减并不会超出本说明书实施例所要保护的范围。
需要说明的是，上述数据均可以使用表明上述经营实体身份的经营实体标识(ID)来进行标识，以此说明是哪个经营实体的相关数据。
在本说明书的实施例中，执行该数据处理方法的应用服务器可以从对经营实体进行管理的服务器处获取上述与经营实体各项指标数据，上述应用服务器还可以通过经营实体的应用客户端从该应用客户端、该经营实体的经营者以及实现各种应用的第三方应用以及第三方应用服务器的数据库中收集上述经营实体的相关数据。具体地，对于与上述经营实体相关联的用户的数据以及与上述经营实体相关联的经营实体的数据可以先通过上述经营实体的关系数据，得到与上述经营实体相关联的用户和经营实体的标识信息，进而根据上述标识信息从相应的服务器处获取与这些标识信息相对应的属性信息和/或行为信息等。
在步骤104，根据每个评价指标的指标数据确定每个评价指标对应的节碳量。
下面将结合附图对本说明书的实施例中确定每个评价指标对应节碳量的方法进行详细说明。
图2为本说明书一个或多个实施例所述的确定一个评价指标对应节碳量的方法流程示意图。如图2所示，上述方法可以包括步骤202～步骤206。
在步骤202，分别确定上述评价指标的各项指标数据对应的节碳量。
在本说明书的实施例中，应当预先设定各项指标数据对应的节碳量的量化算法，从而可以根据预先设定的节碳量量化算法分别确定各项指标数据对应的节碳量。其中，对于不同的指标数据，所采用的节碳量量化算法可以是相同的也可以是不同的。
例如，在本说明书的实施例中，对于绿色用户评价指标中用户的绿色出行数据这一指标数据以及绿色经营者评价指标中绿色出行数据这一指标数据来讲，可以根据用户在预定时间周期内步行的步数或距离来确定绿色出行数据这一指标数据所对应的节碳量。此外，还可以根据用户乘坐公共交通工具(公共汽车或地铁或共享自行车)的次数和/或距离来确定绿色出行数据这一指标数据所对应的节碳量。其中，步行的步数或乘坐公共交通工具的次数越多，对应的节碳量越大；或者步行的距离或乘坐公共交通工具的距离越长，对应的节碳量越大。具体地，在根据用户在预定时间周期内步行的步数或距离来确定绿色出行数据这一指标数据所对应的节碳量时，可以将步行的距离以及普通交通工具在单位距离内产生的碳排放量的乘积作为这一指标数据所对应的节碳量。这是由 于用户采用步行、乘坐公共交通工具的方式可以减少单独的开车出行，故用户的节碳量即为开车出行所带来的碳排放量。其中，普通交通工具在单位距离内产生的碳排放量可以根据普通交通工具在单位距离的平均耗油量以及汽油、柴油等能源的碳排放系数确定。例如，以小汽车作为普通交通公交为例，并假设小汽车每公里平均将消耗0.1升汽油。由于汽油的汽油碳排放系数为2.361kg CO2/L，则小汽车每行驶1公里的碳排放量为0.2361公斤。如此，即可确定每用户步行1公里则节碳量为0.2361公斤。
又例如，对于绿色用户评价指标中用户使用的电子券数据、采用电子方式进行生活缴费的数据这两个指标数据来讲，可以根据用户节约纸制品的数量来确定这两个指标数据所对应的节碳量。其中，使用的电子券次数越多、采用电子方式进行生活缴费的次数越多，则对应的节碳量越大。类似地，对于绿色经营评价指标中线上办公这一指标数据来讲，也可以根据用户节约纸制品的数量来确定这两个指标数据所对应的节碳量。其中，线上办公(例如线上请假、线上报销等)次数越多，则对应的节碳量越大。具体地，在根据用户节约纸制品的数量来确定指标数据所对应的节碳量时，可以将用户所节约的纸制品的数量与生产制造单位数量纸制品所需的碳排放量的乘积作为这些指标数据所对应的节碳量。这是由于用户采用线上办公、电子方式进行缴费以及使用电子券等方式可以避免打印纸制品(纸质票据、纸质单据，纸质支付凭证票据，消费小票以及纸质券等等)，故用户的节碳量即为生产这些纸制品所带来的碳排放量。其中，生产制造单位数量纸制品所需的碳排放量在不同的地区是不同的，因此可以根据某地区票据纸张生产上的排放强度来确定生产制造单位数量纸制品所需的碳排放量；也可以根据生产制造单位数量纸制品所需的平均排放强度来确定生产制造单位数量纸制品所需的碳排放量。
另一方面，对应上述指标数据，由于用户采用了线上办公、线上支付等方式，从而避免了不必要的出行，因此，对于这些数据指标还可以根据用户以线上方式执行业务时所处的位置与最近的业务地点之间的距离来确定这些数据指标所对应的节碳量。具体地，可以将上述距离以及普通交通工具在单位距离内产生的碳排放量的乘积作为这些指标数据所对应的节碳量。具体方法可以参考上述对绿色出行指标数据的说明。
而对于使用环保餐具的数据这一指标数据，可以根据减少产生白色垃圾的量或者节约纸制品的数量来确定这一指标数据所对应的节碳量。具体地，可以将产生的白色垃圾的数量与焚烧或处理单位数量白色垃圾所产生的碳排放量的乘积作为这一指标所对应的节碳量。
还例如，对于绿色经营评价指标中经营实体的能源消耗量这一指标数据来讲，可以根据经营实体的平均能源消耗量来确定这一指标数据所对应的节碳量。其中，平均能 源消耗量越大，则这一指标数据所对应的节碳量就越小。具体地，可以预先统计同行业以及同等规模的经营实体的平均能源消耗量，再计算上述平均能源消耗量与经营实体的能源消耗量之间的差值，并将上述差值与各能源对应的碳排放系数的乘积作为上述指标数据对应的节碳量。例如，对于经营实体的在单位时间内的耗电量A，可以根据其他同行业、同等规模的经营实体在单位时间内的平均耗电量B，计算上述差值A-B，然后，再根据每度电对应的碳排放系数，将上述差值与上述碳排放系数的乘积，作为经营实体用电消耗这一指标数据对应的节碳量。对于经营实体的在单位时间内的耗水量也可以使用类似的方法确定。
再例如，对于绿色区块评价指标中经营行业绿色评级数据以及经营地域绿色评级数据这两项指标数据来讲，可以根据该经营实体对应的经营行业绿色等级或者经营区域绿色等级来确定这一指标数据所对应的节碳量。其中，经营行业绿色评级越高以及经营地域绿色评级越高，则这两项指标数据所对应的节碳量就越大。
类似地，对于绿色图谱评价指标中关联经营实体的绿色评级数据这一指标数据来讲，可以根据该经营实体所关联的经营实体的绿色等级来确定这一指标数据所对应的节碳量。其中，经营行业绿色所关联的经营实体的绿色等级评级越高、达到一定绿色等级的关联经营实体数量越多，则这一指标数据所对应的节碳量就越大。具体地，可以将一个经营实体所管理的其他经营实体的绿色等级进行加权求和或加权求平均，并根据计算结果确定这一指标数据所对应的节碳量。
由于各项指标数据的节碳量量化算法都是预先设定的，本说明书对具体的量化算法不进行限定，因此，在此不再一一举例说明了。
在步骤204，对上述各项指标数据对应的节碳量进行归一化。
需要说明的是，本说明书的实施例并不对各项指标数据的归一化方法进行限定。例如，针对每项指标数据，可以根据预先对一定数量用户或经营实体对应这项指标的节碳量进行统计(例如，统计最大值、最小值、平均值或分布等)，并根据统计结果来对该项指标数据对应的节碳量进行归一化。此外，本说明书的实施例也可以直接根据常用的归一化函数，例如Sigmoid函数对上述各项指标数据对应的节碳量进行归一化。
在步骤206，将上述归一化后的各项指标数据对应的节碳量进行融合，得到上述评价指标所对应的节碳量。
在本说明书的实施例中，可以采用多种方法将一个评价指标的各项指标数据对应的节碳量进行融合。
例如，在本说明书的一些实施例中，可以直接对一个评价指标的各项指标数据对 应的节碳量进行求和或者求平均等，并将计算结果作为该评价指标对应的节碳量。
又例如，在本说明书的另一些实施例中，可以预先对评价指标的各项指标数据设置一个权重值，并根据设置的权重值将上述评价指标的各项指标数据对应的节碳量进行加权求和/加权求平均等，从而得到上述评价指标所对应的节碳量。
在本说明书的实施例中，可以采用多种方法设置一个评价指标的各项指标数据的权重值。例如，可以预先设置固定的权重值；还可以采用层次分析(AHP)，主成分分析(PCA)以及异常检测(具体例如，孤立森林算法)等多种方法之一先确定各项指标数据的重要度；然后，再根据各项指标数据的重要度为各项指标数据分配一个对应的权重值，其中，重要度越大的指标数据所对应的权重值可以越大。
在步骤106，利用经过训练的排序融合模型对各评价指标对应的节碳量进行融合，得到所述经营实体对应的节碳量。
在本说明书的实施例中，上述经过训练的排序融合模型是以上述各评价指标的指标数据及其对应的节碳量为训练数据通过训练得到的。在本说明书的实施例中，上述训练可以根据该经营实体自身的多个评价指标的数据及其对应的节碳量为训练数据。此外，作为替代方案，上述训练还可以根据多个不同经营实体的多个评价指标的数据及其对应的节碳量为训练数据，且训练好的排序融合模型可以应用于针对不同经营实体的节碳量中。
在本说明书的一些实施例中，上述排序融合模型可以通过三次方贝塞尔曲线实现。可知三次方贝塞尔曲线可以通过如下计算表达式(1)来表示。
B(t)＝P 0(1-t) 3+3P 1t(1-t) 2+3P 2t 2(1-t)+P 3t 3      (1)
其中，t为输入向量，具体可为由上述经营实体的各个评价指标包含的各项指标数据归一化后的值组成的向量。也即，上述输入向量的维数可以为从上述指标数据的项数。
上述训练过程即是分别以根据经营实体的每个评价指标所包含各项指标数据归一化后的值确定的向量作为输入向量，以及分别以各个评价指标对应的节碳量为已知输出，通过调整误差，确定上述三次方贝塞尔曲线的各个系数向量P 0、P 1、P 2和P 3。具体训练方法将在后文中详细描述，在此暂且略过。
在上述三次方贝塞尔曲线的各个系数向量确定之后，即可以将上述经营实体各个评价指标所包含各项指标数据归一化后的值组成的向量作为输入向量t，确定上述输入向量t在上述三次方贝塞尔曲线上的投影，也即将上述计算表达式(1)所示的三次方贝塞尔曲线的计算结果B(t)作为该经营实体对应的节碳量。可以看出，上述三次方贝塞尔曲线可以视为一种无监督的排序融合模型RPC(Ranking Principal Curve)。
本领域的技术人员可以理解，上述RPC模型符合用以评估无监督排名结果以及指导排名函数设计的五个元规则，这些元规则包括：(1)可平移性：针对不同的数据维度，排序分值是没有变化的；(2)严格单调性：假设针对绿色经营子模块，子模块分值越高，最后的综合排序的分值也是越高的，需要严格单调性。(3)线性和非线性相容性：假设绿色经营模块与最后综合排序分值是线性的关系，那么其他模块和最后的排序分值也是线性关系；假设是非线性关系，则其他模块也是非线性关系，否则会出现模型偏差。(4)平滑性：假设经营子模块发生很小的变化，那么最后的综合排序分值也要发生微小的变化，不能够跳跃到一个很大差异的值。(5)参数固定性：最后子模块的参数和权重要保持不变，这样综合排序分值是公平的。因此，除了三次方贝塞尔曲线之外，还可以使用其他满足上述五个元规则的RPC模型来替代上述三次方贝塞尔曲线作为上述排序融合模型。
可以理解，本说明书实施例提出的上述RPC排序方法是不拘泥于线性的排序融合，而是从数据本身结构中学习对应的线性和非线性的排序方式，属于一种支持非线性融合的方式，而且通过机器学习的方式使排序结果更为合理和客观。
在步骤108，根据所述经营实体对应的节碳量对所述经营实体的特定数据进行处理。
在本说明书的实施例中，上述特定数据是与节碳量有关的数据。
在本说明书的实施例中，上述处理可以包括对经营实体的节碳量进行统计、分析、将节碳量转换为积分的形式或者还可以将经营实体对应的节碳量转化成0-100之间的分值，并根据经验分为多个星级，其中，星级越高则表示经营实体的经营越绿色越环保。并且，还可以根据经营实体对应的绿色星级或积分为经营实体提供相应的权益。具体地，在本说明书的一些实施例中，上述处理可以包括：根据经营实体对应的节碳量为经营实体分配与上述节碳量相匹配的虚拟物品。
如上所述，业务提供方可基于确定的经营实体对应的节碳量对经营实体的数据进行诸如积分累计、经营实体等级提升以及提供相应权益等处理。从而将相应的业务与经营实体的节碳量进行关联。
在本说明书的实施例中，业务提供方还可以进一步基于上述经营实体在各个评价指标上对应的节碳量对经营实体的特定数据进行处理，例如积分累计、经营实体等级提升以及提供相应权益等等。具体地，可以根据经营实体在上述至少两个评价指标上对应的节碳量确定该经营实体的绿色成分，生成绿色成分图显示给经营实体，从而可以让经营实体一目了然了解自身在各个环保评估维度的节碳量。并且还可以根据经营实体的绿色成分为经营实体提供相应的权益等等，例如可以根据经营实体在上述各个评价指标上 对应的节碳量为经营实体分配与上述节碳量相匹配的虚拟物品等。
从上面所述技术方案可以看出，本说明书一个或多个实施例提供的数据处理方法，可将经营实体碎片化的各类数据进行汇总，并基于汇总后的各类数据，确定出经营实体的节碳量，从而指导经营实体主动进行节能减排。这样的方式将使经营实体可更加直观地获知自身的行为属于怎样的绿色等级，而无需自行查询、计算，对于经营实体而言较为便捷。此外，相应的业务提供方可进一步基于确定的节碳量对经营实体进行诸如积分累计、经营实体等级提升以及提供相应权益等数据处理方式，将相应的业务与经营实体的节碳量进行关联，从而引导带动更多经营实体关注低碳经营，加入节能环保行动，互相促进、增强黏性、打造绿色电商与金融平台。
更进一步，在本说明书一个或多个实施例中采用的排序融合模型是无监督的排序模型，且不拘泥于线性的排序融合，而是从数据本身结构中学习对应的线性和非线性的排序方式，属于一种支持非线性融合的方式，而且通过机器学习的方式使排序及打分结果更为合理和客观。因此，通过这种排序融合模型得到的经营实体对应的节碳量也是非常客观和准确的。
下面将结合附图以及具体的示例详细说明本说明书实施例所述的排序融合模型的训练方法。
图3显示了本说明书一个或多个实施例所述的排序融合模型的训练方法。如图3所示，该训练方法包括步骤302～步骤316。
在步骤302，初始化三次方贝塞尔曲线的端点和控制点。
在本说明书的实施例中，可以根据经验初始化上述贝塞尔曲线的端点P 0和P 3以及控制点P 0和P 3。例如，可以通过如下计算表达式(2)和(3)初始化三次方贝塞尔曲线的端点P 0和P 3。
P 0＝0.5(1-α)      (2)
P 3＝0.5(1+α)      (3)
其中，上述α为预先设置的端点控制参数。上述α可以根据经验设置。上述控制点也可以根据经验通过上述类似的方法初始化。
接下来，针对上述经营实体的各个评价指标分别执行如下步骤304-314。
在步骤304，根据上述评价指标的各项指标数据所对应节碳量归一化后的值确定上述三次方贝塞尔曲线的输入向量，并将上述评价指标所对应的节碳量作为上述三次方贝塞尔曲线的已知输出。
在本说明书的实施例中，可以根据一个评价指标的各项指标数据归一化后的值确 定一个上述三次方贝塞尔曲线的输入向量。上述输入向量中该评价指标的各项指标数据对应的元素为该评价指标的各项指标数据归一化后的值，而其他元素可以设置为0。
在步骤306，确定上述输入向量在上述三次方贝塞尔曲线上的投影。
在本说明书的实施例中，上述投影是指将上述输入向量作为上述计算表达式中的输入向量t，上述计算表达式(1)所示的三次方贝塞尔曲线的计算结果B(t)即为上述输入向量在上述三次方贝塞尔曲线上的投影。
在步骤308，根据上述投影以及上述已知输出确定本次训练的误差。
在本说明书的实施例中，上述误差可以为上述投影与对应已知输出之间的距离。
在步骤310，将上述误差与预先设置的误差阈值进行比较，如果上述误差大于预先设置的误差阈值，则继续执行步骤312；否则，结束本轮训练，并继续执行步骤314。
在步骤312，调整上述三次方贝塞尔曲线中控制点的位置，然后返回上述步骤306。
在本说明书的实施例中，上述调整上述三次方贝塞尔曲线中控制点的位置的方法可以包括：使用最速梯度下降法调整上述三次方贝塞尔曲线中控制点的位置或者使用梯度下降法调整上述三次方贝塞尔曲线中控制点的位置。具体地，最速梯度下降法所采用的移动单位步长可以大于梯度下降法所采用的移动单位步长。因此，当上述误差较大时，可以采用最速梯度下降法调整上述三次方贝塞尔曲线中控制点的位置以实现使训练快速收敛的目的；而当上述误差较小时，可以采用梯度下降法调整上述三次方贝塞尔曲线中控制点的位置以避免训练过程中的震荡效应。
在步骤314，确定是否已对上述经营实体的所有评价指标均执行了上述训练过程，如果是，则执行步骤316；否则，返回上述步骤304。
在步骤316，输出确定的上述三次方贝塞尔曲线的各个系数向量P 0、P 1、P 2和P 3。
可以看出，通过根据上述评价指标的各项指标数据所对应的节碳量及其对应的节碳量拟合三次方贝塞尔曲线的方法可以将上述对各个评价指标的打分规则融合在一起，得到一个综合所有评价指标的指标数据的打分结果。
图4显示了实现上述数据处理方法的一个应用场景。在图4所示的应用上述数据处理方法的系统中，执行上述数据处理方法的应用服务器连接到经营实体的应用客户端。上述应用服务器可以通过应用客户端从经营者、第三方应用以及第三方应用服务器等处获取上述多个评价指标的各项指标数据。上述应用服务器还可以从自身系统中的其他服务器(未示出)，例如从经营实体管理服务器，获取上述多个评价指标的各项指标数据。在获取上述多个评价指标的各项指标数据后，上述应用服务器可以执行上述数据处理方法得到该经营实体对应的节碳量。
下面再结合具体的示例以对经营实体进行绿色经营等级评估为应用场景对本说明书实施例所述的数据处理方法进行说明。
如前所述，在上述应用中，可以预先定义多个可以用于评估经营实体绿色经营等级的评价指标，例如，上述绿色经营、绿色经营者、绿色区块、绿色图谱以及绿色用户这5个评价指标。此外，还需要预先定义这5个评价指标的指标数据，例如，绿色经营这一评价指标包括：经营实体本身的能源消耗量、经营规模等指标数据；绿色经营者这一评价指标包括：经营者的绿色出行数据、电子收款或电子缴费的数据等指标数据；绿色图谱这一评价指标包括：与经营实体密切相关的其他经营实体的绿色经营等级等指标数据；绿色区块这一评价指标包括：与经营实体相关的行业绿色评级数据、以及与经营实体相关的区域绿色评级数据等等指标数据；绿色用户这一评价指标包括：与经营实体相关联的用户的绿色出行数据、所使用的电子券数据、采用电子方式进行生活缴费的数据、使用环保餐具的数据等等指标数据。
在从各方服务器的数据库、用户以及应用汇总收集得到上述指标数据后，先根据各项指标数据对应的节碳量两个算法确定各项指标数据对应的节碳量，再分别对指标数据对应的节碳量进行归一化处理。
接下来，对于每个评价指标，可以利用异常检测的孤立森林算法对该评价指标的指标数据进行切分，并根据指标数据被切分的次数(对应节点的深度)确定各项指标数据的重要度，其中被切分的次数越少的指标数据则可被认为有越高的重要度。
进而根据各项指标数据的重要度确定各项指标数据对应的权重值，并利用确定的权重值对各项指标数据归一化后的值进行加权求和，将得到的和作为该评价指标对应的节碳量。
如此，即可得到绿色经营数据、绿色经营者数据、绿色区块数据、绿色图谱数据以及绿色用户数据这5个评价指标对应的节碳量。
接下来，分别利用上述5个评价指标中每个评价指标的各项指标数据所对应的节碳量以及该评价指标对应的节碳量训练三次方贝塞尔曲线，拟合得到一个符合上述5个评价指标评价标准的三次方贝塞尔曲线。
再接下来，可以根据上述全部指标数据所对应节碳量归一化后的值生成一个输入向量，输入经过训练的三次方贝塞尔曲线，获得上述输入向量在上述三次方贝塞尔曲线上的投影，并将该投影所对应的值作为该经营实体对应的节碳量。
最终，在获得了该经营实体对应的节碳量以及上述经营实体在绿色经营、绿色经营者、绿色区块、绿色图谱以及绿色用户这5个评价指标方面对应的节碳量后，即可对 该经营实体的特定数据进行处理了。例如，可以根据上述经营实体的节碳量对经营实体进行绿色经营评级，得到该经营实体的绿色星级。并且还可以根据经营实体的绿色星级为经营实体提供相应的权益，例如与节碳量或绿色星级相匹配的虚拟物品。还例如，可以根据所述经营实体在绿色经营、绿色经营者、绿色区块、绿色图谱以及绿色用户这五个评价指标上的节碳量确定该经营实体的绿色成分，并可以通过可视化的方式生成如图5所示的经营实体绿色成分示意图提供给用户，指导经营实体的经营行为。例如，通过如图5所示的经营实体绿色成分示意图，可以告知该经营实体在绿色经营、绿色经营者、绿色区块以及绿色用户这四个方面的节碳量都相对比较高或比较均衡，但是在绿色图谱这方面的节碳量相对比较低，从而可以更为细致地指导经营实体的日常行为。并且，在确定了经营实体的绿色星级以及绿色成分后，还可以为经营实体提供相应的权益，进一步鼓励经营实体进行节能减排，从而主动地为环境保护做出贡献。
基于上述数据处理方法，本说明书的一个或多个实施例提出了一种数据处理装置。图6显示了本说明书的一个或多个实施例提出的数据处理装置的内部结构。如图6所示，该数据处理装置包括：数据获取模块602、节碳量确定模块604、排序融合模块606、业务处理模块608。
数据获取模块602，用于按照至少两个评价指标，获取经营实体对应于每个评价指标的指标数据。
节碳量确定模块604，用于根据每个评价指标的指标数据确定每个评价指标对应的节碳量。
排序融合模块606，用于利用经过训练的排序融合模型对各评价指标对应的节碳量进行融合，得到所述经营实体对应的节碳量；其中，所述排序融合模型以所述各评价指标的指标数据及其对应的节碳量为训练数据通过训练得到。
业务处理模块608，用于根据所述经营实体对应的节碳量对所述经营实体的特定数据进行处理；其中，上述特定数据与节碳量相关。
在本说明书的一个或多个实施例中，上述节碳量确定模块604包括：节碳量确定单元，用于分别确定一个评价指标的各项指标数据对应的节碳量；归一化单元，用于对各项指标数据对应的节碳量分别进行归一化；融合单元，用于将所述评价指标的各项指标数据对应的节碳量进行融合，得到所述评价指标对应的节碳量。
在本说明书的一个或多个实施例中，上述融合单元包括：重要度确定子模块，用于采用层次分析、主成分分析以及异常检测中的至少一种确定上述各项指标数据的重要度；权重值设置子模块，用于根据各项指标数据的重要度分别为各项指标数据分配对应 的权重值；其中，重要度越大的指标数据所对应的权重值越大；求和子模块，用于根据设置的权重值将所述评价指标的各项指标数据对应的节碳量进行加权求和，得到所述评价指标对应的节碳量。
在本说明书的一个或多个实施例中，上述排序融合模型通过三次方贝塞尔曲线实现；其中，所述三次方贝塞尔曲线的输入向量为由指标数据所对应节碳量归一化后的值组成的向量；以及所述三次方贝塞尔曲线的输出为所述经营实体对应的节碳量。
在本说明书的一个或多个实施例中，上述数据处理装置可以进一步包括：排序融合模型训练模块，用于通过训练确定所述三次方贝塞尔曲线的各个系数向量；其中，上述排序融合模型训练模块包括：初始化单元，用于初始化三次方贝塞尔曲线的端点和控制点；训练单元，用于针对所述每个评价指标分别执行：A，根据所述评价指标各项指标数据所对应节碳量归一化后的值确定上述三次方贝塞尔曲线的输入向量，并将所述评价指标所对应的节碳量作为所述三次方贝塞尔曲线的已知输出；B，确定所述输入向量在所述三次方贝塞尔曲线上的投影；C，根据所述投影以及所述已知输出确定本次训练的误差；D，响应于所述误差大于预先设置的误差阈值的情况，调整上述三次方贝塞尔曲线中控制点的位置，并返回B；E，响应于所述误差小于或等于预先设置的误差阈值的情况，在已对上述经营实体的所有评价指标均执行了上述训练过程时，输出确定的所述三次方贝塞尔曲线的各个系数向量；否则，返回A。
此外，在本说明书的实施例中，上述数据处理装置可以被视为一个电子设备，因此，上述数据处理装置的内部结构可如图6所示，包括：处理器610、存储器620、输入/输出接口630、通信接口640和总线650。其中，处理器610、存储器620、输入/输出接口630和通信接口640通过总线650实现彼此之间在设备内部的通信连接。
存储器620可以采用ROM(Read Only Memory，只读存储器)、RAM(Random Access Memory，随机存取存储器)、静态存储设备，动态存储设备等形式实现。存储器620可以存储操作系统和其他应用程序，还可以存储本说明书实施例提供的上述数据处理装置的各个模块，通过软件或者固件来实现本说明书实施例所提供的技术方案时，相关的程序代码保存在存储器620中，并由处理器610来调用执行。
上述处理器610可以采用通用的CPU(Central Processing Unit，中央处理器)、微处理器、应用专用集成电路(Application Specific Integrated Circuit，ASIC)、或者一个或多个集成电路等方式实现，用于执行相关程序，以实现本说明书实施例所提供的技术方案。
输入/输出接口630可以用于连接输入/输出模块，实现信息输入及输出。输入输出 /模块可以作为组件配置在设备中(图中未示出)，也可以外接于设备以提供相应功能。其中输入设备可以包括键盘、鼠标、触摸屏、麦克风、各类传感器等，输出设备可以包括显示器、扬声器、振动器、指示灯等。
通信接口640用于连接通信模块(图中未示出)，以实现本设备与其他设备的通信交互。其中，通信模块可以通过有线方式(例如USB、网线等)实现通信，也可以通过无线方式(例如移动网络、WIFI、蓝牙等)实现通信。
总线650包括一通路，在设备的各个组件(例如处理器、存储器、输入/输出接口和通信接口)之间传输信息。
需要说明的是，尽管上述设备仅示出了处理器、存储器、输入/输出接口、通信接口以及总线，但是在具体实施过程中，该设备还可以包括实现正常运行所必需的其他组件。此外，本领域的技术人员可以理解的是，上述设备中也可以仅包含实现本说明书实施例方案所必需的组件，而不必包含图中所示的全部组件。
本说明书实施例中所述支付涉及的技术载体，例如可以包括近场通信(Near Field Communication，NFC)、WIFI、3G/4G/5G、POS机刷卡技术、二维码扫码技术、条形码扫码技术、蓝牙、红外、短消息(Short Message Service，SMS)、多媒体消息(Multimedia Message Service，MMS)等。
本说明书实施例中所述生物识别所涉及的生物特征，例如可以包括眼部特征、声纹、指纹、掌纹、心跳、脉搏、染色体、DNA、人牙咬痕等。其中眼纹可以包括虹膜、巩膜等生物特征。
需要说明的是，本说明书一个或多个实施例的方法可以由单个设备执行，例如一台计算机或服务器等。本实施例的方法也可以应用于分布式场景下，由多台设备相互配合来完成。在这种分布式场景的情况下，这多台设备中的一台设备可以只执行本说明书一个或多个实施例的方法中的某一个或多个步骤，这多台设备相互之间会进行交互以完成所述的方法。
上述对本说明书特定实施例进行了描述。其它实施例在所附权利要求书的范围内。在一些情况下，在权利要求书中记载的动作或步骤可以按照不同于实施例中的顺序来执行并且仍然可以实现期望的结果。另外，在附图中描绘的过程不一定要求示出的特定顺序或者连续顺序才能实现期望的结果。在某些实施方式中，多任务处理和并行处理也是可以的或者可能是有利的。
为了描述的方便，描述以上装置时以功能分为各种模块分别描述。当然，在实施本说明书一个或多个实施例时可以把各模块的功能在同一个或多个软件和/或硬件中实现。
上述实施例的装置用于实现前述实施例中相应的方法，并且具有相应的方法实施例的有益效果，在此不再赘述。
本实施例的计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带磁磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。
所属领域的普通技术人员应当理解：以上任何实施例的讨论仅为示例性的，并非旨在暗示本公开的范围(包括权利要求)被限于这些例子；在本公开的思路下，以上实施例或者不同实施例中的技术特征之间也可以进行组合，步骤可以以任意顺序实现，并存在如上所述的本说明书一个或多个实施例的不同方面的许多其它变化，为了简明它们没有在细节中提供。
另外，为简化说明和讨论，并且为了不会使本说明书一个或多个实施例难以理解，在所提供的附图中可以示出或可以不示出与集成电路(IC)芯片和其它部件的公知的电源/接地连接。此外，可以以框图的形式示出装置，以便避免使本说明书一个或多个实施例难以理解，并且这也考虑了以下事实，即关于这些框图装置的实施方式的细节是高度取决于将要实施本说明书一个或多个实施例的平台的(即，这些细节应当完全处于本领域技术人员的理解范围内)。在阐述了具体细节(例如，电路)以描述本公开的示例性实施例的情况下，对本领域技术人员来说显而易见的是，可以在没有这些具体细节的情况下或者这些具体细节有变化的情况下实施本说明书一个或多个实施例。因此，这些描述应被认为是说明性的而不是限制性的。
尽管已经结合了本公开的具体实施例对本公开进行了描述，但是根据前面的描述，这些实施例的很多替换、修改和变型对本领域普通技术人员来说将是显而易见的。例如，其它存储器架构(例如，动态RAM(DRAM))可以使用所讨论的实施例。
本说明书一个或多个实施例旨在涵盖落入所附权利要求的宽泛范围之内的所有这样的替换、修改和变型。因此，凡在本说明书一个或多个实施例的精神和原则之内，所做的任何省略、修改、等同替换、改进等，均应包含在本公开的保护范围之内。

Claims (20)

  1. 一种数据处理方法，该方法包括：
    按照至少两个评价指标，获取经营实体对应于每个评价指标的指标数据；
    根据每个评价指标的指标数据确定每个评价指标对应的节碳量；
    利用经过训练的排序融合模型对各评价指标对应的节碳量进行融合，得到所述经营实体对应的节碳量；其中，所述排序融合模型以所述各评价指标的指标数据及其对应的节碳量为训练数据通过训练得到；
    根据所述经营实体对应的节碳量对所述经营实体的特定数据进行处理；其中，所述特定数据与节碳量有关。
  2. 根据权利要求1所述的数据处理方法，其中，根据每个评价指标的指标数据确定每个评价指标对应的节碳量包括：
    针对每个评价指标，分别确定所述评价指标的各项指标数据对应的节碳量；分别将所述各项指标数据对应的节碳量进行归一化；以及将归一化后的所述各项指标数据对应的节碳量进行融合，得到所述评价指标对应的节碳量。
  3. 根据权利要求2所述的数据处理方法，其中，将归一化后的所述各个指标数据对应的节碳量进行融合包括：
    对所述各项指标数据分别设置权重值；
    根据设置的权重值将归一化后的所述各项指标数据对应的节碳量进行加权求和，得到所述评价指标对应的节碳量。
  4. 根据权利要求3所述的数据处理方法，其中，对所述各项指标数据分别设置权重值包括：
    采用层次分析、主成分分析以及异常检测方法中的至少一种确定所述各项指标数据的重要度；以及
    根据所述各项指标数据的重要度分别为所述各项指标数据分配对应的权重值；其中，重要度越大的指标数据所对应的权重值越大。
  5. 根据权利要求2所述的数据处理方法，其中，所述排序融合模型通过三次方贝塞尔曲线实现；其中，所述三次方贝塞尔曲线的输入向量为由所述各个评价指标中各项指标数据对应的节碳量归一化后的值组成的向量；以及所述三次方贝塞尔曲线的输出为所述经营实体对应的节碳量。
  6. 根据权利要求5所述的数据处理方法，其中，训练所述排序融合模型的过程包括：
    初始化三次方贝塞尔曲线的端点和控制点；
    针对每个评价指标分别执行如下步骤：
    A，根据所述评价指标的各项指标数据对应的节碳量归一化后的值确定上述三次方贝塞尔曲线的输入向量，并将所述评价指标所对应的节碳量作为所述三次方贝塞尔曲线的已知输出；
    B，确定所述输入向量在所述三次方贝塞尔曲线上的投影；
    C，根据所述投影以及所述已知输出确定本次训练的误差；
    D，响应于所述误差大于预先设置的误差阈值的情况，调整上述三次方贝塞尔曲线中控制点的位置，并返回B；
    E，响应于所述误差小于或等于预先设置的误差阈值的情况，在已对所述经营实体的所有评价指标均执行了所述训练过程时，输出确定的所述三次方贝塞尔曲线的各个系数向量；否则，返回A。
  7. 根据权利要求6所述的数据处理方法，其中，根据所述评价指标的各项指标数据对应的节碳量归一化后的值确定上述三次方贝塞尔曲线的输入向量包括：将所述输入向量中与所述评价指标的各项指标数据对应的元素设置为所述评价指标的各项指标数据对应的节碳量归一化后的值，并将所述输入向量的其他元素设置为0。
  8. 根据权利要求6所述的数据处理方法，其中，所述调整所述三次方贝塞尔曲线中控制点的位置包括：使用最速梯度下降法或者梯度下降法调整所述三次方贝塞尔曲线中控制点的位置。
  9. 根据权利要求1所述的数据处理方法，其中，根据所述经营实体对应的节碳量对所述经营实体的特定数据进行处理包括：根据所述经营实体对应的节碳量为所述经营实体分配与所述节碳量相匹配的虚拟物品。
  10. 根据权利要求1所述的数据处理方法，进一步包括：根据所述每个评价指标对应的节碳量对所述经营实体的业务数据进行处理。
  11. 根据权利要求1所述的数据处理方法，其中，所述评价指标包括：绿色经营评价指标、绿色经营者评价指标、绿色区块评价指标、绿色图谱评价指标以及绿色用户评价指标中的至少两个。
  12. 一种数据处理装置，包括：
    数据获取模块，用于按照至少两个评价指标，获取经营实体对应于每个评价指标的指标数据；
    节碳量确定模块，用于根据每个评价指标的指标数据确定每个评价指标对应的节碳 量；
    排序融合模块，用于利用经过训练的排序融合模型对各评价指标对应的节碳量进行融合，得到所述经营实体对应的节碳量；其中，所述排序融合模型以所述各评价指标的指标数据及其对应的节碳量为训练数据通过训练得到；以及
    业务处理模块，用于根据所述经营实体对应的节碳量对所述经营实体的特定数据进行处理；其中，所述特定数据与节碳量有关。
  13. 根据权利要求12所述的数据处理装置，其中，所述节碳量确定模块包括：
    节碳量确定单元，用于分别确定一个评价指标的各项指标数据对应的节碳量；
    归一化单元，用于分别对所述各项指标数据对应的节碳量进行归一化；
    融合单元，用于将归一化后的所述各项指标数据对应的节碳量进行融合，得到所述评价指标对应的节碳量。
  14. 根据权利要求13所述的数据处理装置，其中，所述融合单元包括：
    重要度确定子模块，用于采用层次分析、主成分分析以及异常检测中的至少一种确定所述各项指标数据的重要度；
    权重值设置子模块，用于根据所述各项指标数据的重要度分别为各项指标数据分配对应的权重值；其中，重要度越大的指标数据对应的权重值越大；
    求和子模块，用于根据设置的权重值将归一化后的所述评价指标的各项指标数据对应的节碳量进行加权求和，得到所述评价指标对应的节碳量。
  15. 根据权利要求12所述的数据处理装置，其中，所述排序融合模型通过三次方贝塞尔曲线实现；其中，所述三次方贝塞尔曲线的输入向量为由所述各个评价指标的指标数据对应的节碳量归一化后的值组成的向量；以及所述三次方贝塞尔曲线的输出为所述经营实体对应的节碳量。
  16. 根据权利要求15所述的数据处理装置，进一步包括：排序融合模型训练模块，用于通过训练确定所述三次方贝塞尔曲线的各个系数向量；其中，
    所述排序融合模型训练模块包括：
    初始化单元，用于初始化三次方贝塞尔曲线的端点和控制点；
    训练单元，用于针对每个评价指标分别执行：
    A，根据所述评价指标的指标数据对应的节碳量归一化后的值确定上述三次方贝塞尔曲线的输入向量，并将所述评价指标所对应的节碳量作为所述三次方贝塞尔曲线的已知输出；
    B，确定所述输入向量在所述三次方贝塞尔曲线上的投影；
    C，根据所述投影以及所述已知输出确定本次训练的误差；
    D，响应于所述误差大于预先设置的误差阈值的情况，调整上述三次方贝塞尔曲线中控制点的位置，并返回B；
    E，响应于所述误差小于或等于预先设置的误差阈值的情况，在已对上述经营实体的所有评价指标均执行了上述训练过程时，输出确定的所述三次方贝塞尔曲线的各个系数向量；否则，返回A。
  17. 根据权利要求12所述的数据处理装置，其中，根据业务处理模块根据所述经营实体对应的节碳量为所述经营实体分配与所述节碳量相匹配的虚拟物品。
  18. 根据权利要求17所述的数据处理装置，其中，所述业务处理模块进一步根据所述每个评价指标对应的节碳量对所述经营实体的业务进行处理。
  19. 一种电子设备，包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时实现如权利要求1至11中任意一项所述的数据处理方法。
  20. 一种非暂态计算机可读存储介质，所述非暂态计算机可读存储介质存储计算机指令，所述计算机指令用于使所述计算机执行如权利要求1至11中任意一项所述的数据处理方法。
PCT/CN2020/132765 2020-03-20 2020-11-30 数据处理方法、装置、电子设备及存储介质 WO2021184825A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010203025.7 2020-03-20
CN202010203025.7A CN111401759B (zh) 2020-03-20 2020-03-20 数据处理方法、装置、电子设备及存储介质

Publications (1)

Publication Number Publication Date
WO2021184825A1 true WO2021184825A1 (zh) 2021-09-23

Family

ID=71431113

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/132765 WO2021184825A1 (zh) 2020-03-20 2020-11-30 数据处理方法、装置、电子设备及存储介质

Country Status (2)

Country Link
CN (1) CN111401759B (zh)
WO (1) WO2021184825A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2024055546A1 (zh) * 2022-09-16 2024-03-21 广东邦普循环科技有限公司 动力电池回收量预测方法、装置及介质
CN117808497A (zh) * 2024-03-01 2024-04-02 清华四川能源互联网研究院 基于距离和方向特征的电力碳排放异常检测模块和方法

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111401759B (zh) * 2020-03-20 2022-08-23 支付宝(杭州)信息技术有限公司 数据处理方法、装置、电子设备及存储介质
CN112231891A (zh) * 2020-09-09 2021-01-15 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） 电子电器产品绿色设计方法以及系统
CN112488538A (zh) * 2020-12-04 2021-03-12 国泰新点软件股份有限公司 评价指标的上报处理方法、装置及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102034023A (zh) * 2010-12-07 2011-04-27 上海交通大学 基于证据理论的多源信息融合风险分析方法
CN108090208A (zh) * 2017-12-29 2018-05-29 广东欧珀移动通信有限公司 融合数据处理方法及装置
CN108520434A (zh) * 2016-08-24 2018-09-11 阿里巴巴集团控股有限公司 一种数据处理方法及装置
CN109460825A (zh) * 2018-10-24 2019-03-12 阿里巴巴集团控股有限公司 用于构建机器学习模型的特征选取方法、装置以及设备
CN111401759A (zh) * 2020-03-20 2020-07-10 支付宝(杭州)信息技术有限公司 数据处理方法、装置、电子设备及存储介质

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8165891B2 (en) * 2007-12-31 2012-04-24 Roberts Charles E S Green rating system and associated marketing methods
CN108229337B (zh) * 2017-12-14 2021-03-30 创新先进技术有限公司 一种数据处理的方法、装置及设备
CN110222762A (zh) * 2019-06-04 2019-09-10 恒安嘉新（北京）科技股份公司 对象预测方法、装置、设备、及介质
CN110245757B (zh) * 2019-06-14 2022-04-01 上海商汤智能科技有限公司 一种图像样本的处理方法及装置、电子设备和存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102034023A (zh) * 2010-12-07 2011-04-27 上海交通大学 基于证据理论的多源信息融合风险分析方法
CN108520434A (zh) * 2016-08-24 2018-09-11 阿里巴巴集团控股有限公司 一种数据处理方法及装置
CN108090208A (zh) * 2017-12-29 2018-05-29 广东欧珀移动通信有限公司 融合数据处理方法及装置
CN109460825A (zh) * 2018-10-24 2019-03-12 阿里巴巴集团控股有限公司 用于构建机器学习模型的特征选取方法、装置以及设备
CN111401759A (zh) * 2020-03-20 2020-07-10 支付宝(杭州)信息技术有限公司 数据处理方法、装置、电子设备及存储介质

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2024055546A1 (zh) * 2022-09-16 2024-03-21 广东邦普循环科技有限公司 动力电池回收量预测方法、装置及介质
CN117808497A (zh) * 2024-03-01 2024-04-02 清华四川能源互联网研究院 基于距离和方向特征的电力碳排放异常检测模块和方法

Also Published As

Publication number Publication date
CN111401759A (zh) 2020-07-10
CN111401759B (zh) 2022-08-23

Similar Documents

Publication Publication Date Title
WO2021184825A1 (zh) 数据处理方法、装置、电子设备及存储介质
US20210342721A1 (en) Method, apparatus, and computer program product for determining closing metrics
TWI709867B (zh) 資料處理方法及裝置
US11120345B2 (en) Method, apparatus, and computer program product for determining closing metrics
CN108154401A (zh) 用户画像刻画方法、装置、介质和计算设备
KR102286415B1 (ko) 제품수명주기에 따른 수명주기별 온라인과 오프라인 정보분석 서비스 시스템
Zhang et al. Supply chain coordination based on a buyback contract under fuzzy random variable demand
JP7271529B2 (ja) 自動化されたアトリビューションモデリングおよび測定
CN111340246A (zh) 用于企业智能决策分析的处理方法、装置和计算机设备
Agarski et al. An approach to multi-criteria environmental evaluation with multiple weight assignment
US20150095111A1 (en) Method and system for using social media for predictive analytics in available-to-promise systems
CN112561598A (zh) 基于客户画像的客户流失预测及挽回方法和系统
Wu et al. An online niche-market tour identification system for the travel and tourism industry
CN106372964A (zh) 一种行为忠诚度的识别及管理方法、系统和终端
US20090276290A1 (en) System and method of optimizing commercial real estate transactions
Yuping et al. New methods of customer segmentation and individual credit evaluation based on machine learning
Ekasari et al. Service Quality and After-Sales Service on IoT-Based Car User Satisfaction and Repeat Purchases Services in Indonesia
CN108961037B (zh) 一种基于对车辆使用情况评估算法的车辆贷款风控方法及装置
Zhang et al. Customer service robot model based on e-commerce dual-channel channel supply coordination and compensation strategy in the perspective of big data
US11694258B2 (en) Systems, methods, and devices for generating and trading environmental credits
CN104408558B (zh) 一种财务运营仿真模拟模型体系
Liu et al. Social-Transportation Analytic Toolbox (STAT) for Transit Networks
US11836769B2 (en) Methods and systems for providing estimated transactional data
KR102615356B1 (ko) 탄소 감축 플랫폼 제공 시스템, 탄소 배출 정보 제공 방법 및 이를 위한 컴퓨터 프로그램
Nie et al. From data acquisition to validation: a complete workflow for predicting individual customer lifetime value

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: 20925720

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: 20925720

Country of ref document: EP

Kind code of ref document: A1