Digital service design apparatus, digital service design method, digital service design system, and digital service system

Abstract

The challenge is to accurately profile customers' lifestyles by time of day and provide digital services that enhance the value of the customer experience. [Solution] The digital service design device 100 collects smart meter information 2 and questionnaire information 4 as customer characteristic information indicating the attributes of energy consumers and their equipment from an external communication network 108, identifies service information similar to the collected customer characteristic information from service information 3 which shows the characteristics of existing services provided to consumers in qualitative language descriptions, compares the customer characteristic information and the identified service information to calculate the difference, and generates new service information indicating a new service that complements the difference.

Description

Technical Field

[0001] The present invention relates to a technology that utilizes customer information. In particular, it relates to a technology for designing digital services that provide a convenient user experience for customers.

Background Art

[0002] In recent years, it has been required to appropriately grasp and analyze characteristic data that can be obtained from customers with various lifestyles, and for service providers to provide various user experiences to customers. Based on this, it is also required for various retail service providers such as those in the power industry to further enhance the added value of customers. In particular, with the development of IoT (Internet of Things) technology and digital transformation (DX), the business environment has changed significantly. Therefore, it has come to be assumed that customer information collected from CIS (Customer Information System) / CRM (Customer Relation Management) and MDMS (Meter Data Management System), etc., is effectively utilized for the purpose of linking digital services and customer acceptance. Under such circumstances, the importance of establishing digital service design has been increasing. Here, digital service design is a comprehensive approach for deeply understanding customer needs and providing new value by utilizing IoT devices and digital technologies.

[0003] In particular, the creation and utilization of user profiles form the basis of effective digital service design. User profiles provide important information in the digital ecosystem and enable the understanding of customer needs and preferences.

[0004] Patent Document 1 discloses a system that sets the user's initial consumer behavior profile as the perceived consumer behavior profile, identifies the agreement or disagreement with the consumer profile obtained from actual behavior, and continuously updates the record of responses to the user's consumer behavior to generate a true consumer behavior profile.

[0005] Furthermore, Patent Document 2 discloses a questionnaire classification system that can classify questionnaires using cluster analysis without requiring expert knowledge, in order to support product planning and advertising activities tailored to lifestyles.

[0006] Furthermore, Patent Document 3 discloses a survey implementation system that can extract survey subjects from a customer database and obtain more appropriate survey results for consumers who receive services from lifeline sources such as electricity, gas, and water. [Prior art documents] [Patent Documents]

[0007] [Patent Document 1] Special Publication No. 2003-524236 [Patent Document 2] Japanese Patent Publication No. 2005-258774 [Patent Document 3] Japanese Patent Publication No. 2007-199878 [Overview of the Initiative] [Problems that the invention aims to solve]

[0008] Patent Document 1 describes a method for understanding user characteristics using data such as online activity and purchase history on various websites. However, the data that can be obtained in Patent Document 1, such as online activity and purchase history, is limited to specific target behaviors. Therefore, there are limitations in comprehensively profiling a user's lifestyle, where their behavior and preferences can change depending on the time of day and their areas of interest.

[0009] Furthermore, Patent Document 2 describes a system that assists experts in creating classification rules for each lifestyle trend by performing principal component analysis based on survey results. However, in Patent Document 2, the classification rules created in advance are limited to the range of survey items, and when used for entirely new product planning or advertising, the scope is limited according to the range of survey items.

[0010] Furthermore, in Patent Document 3, survey participants who fit the survey content are selected from customers, and these survey participants and survey content are received via a mobile device. However, in Patent Document 3, by limiting the survey participants to customers who fit the survey content, customers who have potential needs for new services, etc., are excluded. As a result, the possibilities for newly designed digital services that can appeal to users with potential needs are limited.

[0011] Therefore, the present invention aims to provide more detailed services by analyzing the circumstances of customers, such as their lifestyles, with greater accuracy. [Means for solving the problem]

[0012] To achieve the above objectives, the present invention identifies service information similar to customer characteristic information, which indicates the attributes of energy consumers and their equipment, from service information that qualitatively describes the characteristics of existing services provided to consumers, calculates the difference by comparing the customer characteristic information and the identified service information, and generates new service information that indicates a new service that complements the difference.

[0013] More specifically, the digital service design device generates service information indicating services for multiple energy users, and includes a storage unit that stores user information indicating the attributes of the multiple users and equipment, demand information indicating the amount of energy the users demand, and service information indicating the characteristics of the services that a service provider provides to the multiple users in qualitative language descriptions; a user characteristic analysis unit that uses the user information and / or demand information to perform cluster analysis on the quantitative characteristics of each of the multiple users according to weight parameters set in advance to associate the service information with each of the multiple users, classifies the multiple users into multiple user groups, identifies user characteristic information indicating the characteristics of each user group in qualitative language descriptions based on the characteristic values ​​of each classified user group and the language descriptions of the user information; and a digital service generation unit that identifies service information similar to the user characteristic information, compares the user characteristic information and the identified service information to calculate the difference, and generates new service information indicating new services that complement the difference.

[0014] Furthermore, the present invention also includes a digital service design system including a digital service design device, and a digital service design method executed by the digital service design device. In addition, the present invention also includes a digital service design program that enables the digital service design device to function as a computer, and a storage medium for storing the same. [Effects of the Invention]

[0015] According to the present invention, it is possible to analyze the circumstances of customers, such as their lifestyle, with greater accuracy and provide more detailed services. [Brief explanation of the drawing]

[0016] [Figure 1A] A block diagram showing the overall configuration of a digital service design system 10 according to one embodiment of the present invention. [Figure 1B] Hardware configuration diagram of the digital service design device 100 according to an embodiment of the present invention. [Figure 2] A diagram for explaining a questionnaire information cycle showing the cycle of questionnaire information 4 in an embodiment of the present invention. [Figure 3] A diagram showing an example of a questionnaire survey as an information contact point in an embodiment of the present invention. [Figure 4] A diagram showing a digital service platform in an embodiment of the present invention. [Figure 5] A diagram showing an example of profiling customers from multiple questionnaire survey results in an embodiment of the present invention. [Figure 6] A diagram for explaining a method of classifying customer characteristics in an embodiment of the present invention. [Figure 7] A diagram showing the overall processing flow in customer clustering using the X-means method in an embodiment of the present invention. [Figure 8] A diagram showing the overall processing flow in customer profiling in an embodiment of the present invention. [Figure 9] A diagram for explaining the mapping process for generating profiling results in an embodiment of the present invention. [Figure 10] A diagram showing the correspondence between customer profiling results and digital service design plans in an embodiment of the present invention. [Figure 11] A diagram showing smart meter information 2 in an embodiment of the present invention. [Figure 12] A diagram showing a digital service design plan having a service use case 1200 and its service content 1205 in an embodiment of the present invention.

Embodiments for Carrying Out the Invention

[0017] One embodiment of the present invention will be described below. In this embodiment, survey information and smart meter information from users are collected periodically or irregularly on a continuous basis, and a design proposal for a digital service is generated based on this information. To this end, in this embodiment, the user's lifestyle, energy usage patterns, and preferences are quantitatively understood through principal component analysis and cluster analysis, and user profiling is precisely performed by a generation AI based on the quantitative characteristics of each user or user group. Then, from the user profiling results, a design proposal is created in which the features of the digital service to be provided are expressed in linguistic descriptions for each user or user group. Embodiments of the present invention will be described below with reference to the drawings.

[0018] Figure 1A is a block diagram showing the overall configuration of the digital service design system 10 according to this embodiment. The digital service design system 10 has a digital service design device 100 that performs the main processing in this embodiment. Based on smart meter information 2, questionnaire information 4, and existing service information 3, the digital service design device 100 generates a design proposal for a digital service described in a language description, which is an example of new service information. Here, service information refers to information about apps and software used in digital services that provide a convenient user experience to consumers. Examples of apps and software include energy management apps, food review sharing management apps, drive planning management apps, and education management apps. However, apps and software are not limited to specific service business fields. Furthermore, information about apps and software also includes descriptions of the features of apps and software expressed multimodally using natural language, charts, images, videos, audio, or at least two combinations thereof, as well as content used in apps and software, service menus, design documents, specifications, and other documents that form the basis of apps, and program code for apps and software. Furthermore, the digital service design system 10 of this embodiment is connected to customers 109 and service provider systems 116 via an external communication network 108 such as the Internet. The external communication network 108 may also be implemented using multiple networks. Before describing the digital service design system 10 that performs the main processing of this embodiment, customers 109 and service provider systems 116 will be described below.

[0019] First, the external communication network 108 is connected to a customer 109, which includes its information terminal 111 and smart meter 112. The term "customer 109" encompasses a broad range of organizations, including individual customers 110, families, and businesses. Customer 109 receives and utilizes electricity, an example of energy. While this embodiment uses electricity as the energy source, the present invention can also be applied to infrastructure that allows for monitoring customer usage, such as gas, water (water and sewage), and communications (including landline and mobile phones). Figure 1A shows an example where customer 109 is equipped with power-related facilities such as a solar power generation system 114 and a storage battery 115. Customer 109 is not required to have these facilities, and may have other facilities. Although Figure 1A shows only one customer 109, multiple customers 109s can exist. Figure 1A omits the depiction of power systems such as transmission and distribution networks and power plants.

[0020] Furthermore, the information terminal 111 is a terminal device used by the customer 110, and can be a smartphone, tablet, or PC. The information terminal 111 is used by the customer 110 to access the services designed in this embodiment.

[0021] Furthermore, the smart meter 112 is connected to the power transmission and distribution network, and the amount of electricity demanded (used) at customer 109 is measured. Although not shown in Figure 1A, customer 109 uses electrical appliances that consume electricity.

[0022] Furthermore, the electricity generated by the solar power generation system 114 is either used by the consumer 109 or sold to the power company. The battery 115 stores electricity supplied by the power company and surplus electricity generated by the solar power generation system 114, and is used by the consumer 109 when needed, separate from when it is supplied or generated.

[0023] Furthermore, the service provider system 116 is a computer system used by a service provider that provides the services designed in this embodiment. There may be multiple service provider systems 116, and a service provider may offer multiple services. Therefore, it is desirable for the service provider system 116 to store service information 3. This service information 3 will be explained later.

[0024] Furthermore, the service provider system 116 can be implemented as a digital service device that provides service information 3 to multiple energy users in order to provide services. More preferably, the service provider system 116 transmits the service information 3 to the information terminals 111 used by those users. To this end, the service provider system 116, being a digital service device, periodically or irregularly distributes the service information 3 generated by the digital service design device 100 to user groups identified by user characteristic information in users in a predetermined area via an external communication network 108 in that predetermined area. More preferably, the service provider system 116 distributes the service information 3 to the information terminals 111 of the identified user groups.

[0025] In this case, the receiving and distributing unit 101 will receive smart meter information 2, which is customer information and / or demand amount information, with individual customers and / or distribute service information 3. The receiving and distributing unit 101 may be part of the digital service design device 100, as described later, or it may be configured as a separate device from the service provider system 116 (digital service device) and the digital service design device 100.

[0026] As a result of the above, the digital service system is configured by the digital service design device 100, the service provider system 116 which is a digital service device, and the receiving / distribution unit 101.

[0027] Next, we will describe the digital service design system 10. The digital service design system 10 is an example of a service design system that generates new service information based on demand information, customer information, and existing service information.

[0028] Therefore, the digital service design system 10 includes a digital service design device 100, an MDMS 105, a CIS / CRM 106, and a user terminal device 113.

[0029] First, the digital service design device 100 designs a digital service, that is, generates service information 3. For this purpose, the digital service design device 100 includes a receiving and distribution unit 101, a customer characteristics analysis unit 102, a digital service generation unit 103, and a demand forecasting unit 104.

[0030] First, the receiving and distributing unit 101 connects to the customer 109 (information terminal 111, smart meter 112) and the service provider system 116 via the external communication network 108. The receiving and distributing unit 101 also connects to the MDMS 105, CIS / CRM 106, and user terminal device 113 via the internal communication network 107. The receiving and distributing unit 101 then communicates with each of these devices, sending and receiving information. For example, the receiving and distributing unit 101 obtains survey information 4 to be distributed to the customer from the CIS / CRM 106. Then, the receiving and distributing unit 101 distributes the survey information to the information terminal 111 periodically or irregularly via the external communication network 108.

[0031] Furthermore, the receiving and distributing unit 101 periodically or irregularly collects survey information answered by the customer 109 or customer 110 from the information terminal 111 via the external communication network 108. This survey information is stored in the CIS / CRM 106 as survey information 4.

[0032] Furthermore, the receiving and distributing unit 101 periodically collects smart meter information from the smart meter 112. This smart meter information is then stored in the MDMS 105 as smart meter information 2.

[0033] Furthermore, the receiving and distributing unit 101 periodically or irregularly distributes the generated service information 3 to the information terminal 111 via the external communication network 108. The receiving and distributing unit 101 also acquires service information transmitted periodically or irregularly by the information terminal 111. This service information is then stored in the CIS / CRM 106 as service information 3.

[0034] Next, the customer characteristics analysis unit 102 classifies multiple customers into customer groups based on at least one of the questionnaire information 4 obtained from the CIS / CRM 106 and the smart meter information 2 obtained from the MDMS 105. Then, the customer characteristics analysis unit 102 identifies customer characteristics information that is expressed in qualitative language descriptions of the characteristics of each customer group. To this end, the customer characteristics analysis unit 102 uses at least one of the questionnaire information 4 and the smart meter information 2 to perform cluster analysis on the quantitative characteristics of multiple customers according to weight parameters that have been set in advance to associate service information 3 with each of the multiple customers, thereby classifying the multiple customers into multiple customer groups. Note that customer characteristics information such as questionnaire information 4 and smart meter information 2 is collected from the information terminal 111 and smart meter 112 via the external communication network 108. The weight parameters can be set in advance, or they may be made interactively change via the user terminal device 113 when a service designer designs a desired service.

[0035] The customer characteristic analysis unit 102 then identifies customer characteristic information, which is expressed in a customer language description, a qualitative language description of the characteristics of each of the classified customer groups, based on the characteristic values ​​of each customer group and the language description of the customer information. It is desirable to store this customer characteristic information description in a storage unit (not shown).

[0036] Furthermore, the digital service generation unit 103 compares the customer characteristic information identified by the customer characteristic analysis unit 102 with the service information 3 to calculate the difference and generates new service information indicating a new service that complements that difference. In this case, it is desirable to use the actual service data of the service indicated by the service information 3.

[0037] More specifically, the digital service generation unit 103 associates customer characteristic information with service information 3 (more preferably its actual data) based on the similarity of the qualitative language descriptions, and generates new service information by supplementing the difference between the service information shown in the qualitative language description and the actual data. This new service information is stored in the CIS / CRM 106 as service information 3.

[0038] Furthermore, the demand forecasting unit 104 acquires smart meter information 2 and forecasts the amount of electricity demand for each customer 109. In this process, environmental data such as temperature and humidity for each customer 109 may also be used. The demand forecasting unit 104 also forecasts the amount of electricity demand for each customer (for example, each household) for short periods such as 30 minutes, medium periods such as one hour to half a day, and long periods such as one day to several days. Note that the short-term, medium-term, and long-term time periods may be time periods of various lengths other than those mentioned above. The receiving and distributing unit 101 to the demand forecasting unit 104 of the digital service design device 100 are connected to each other via a communication channel such as a bus.

[0039] Next, MDMS105 is a Meter Data Management System that stores smart meter information 2. Therefore, MDMS105 can be implemented as a database system and will be used by the power company to calculate electricity charges. Thus, smart meter information 2 is smart meter information that includes time-series electricity meter values ​​collected from smart meters 112 of customers 109 via an external communication network 108. In this way, smart meter information 2 is an example of demand information that shows the amount of energy demanded by each customer 109.

[0040] Furthermore, the MDMS 105 is connected to the receiving and distributing unit 101 via the internal communication network 107. As a result, the digital service design device 100, such as the customer characteristics analysis unit 102, can perform processing using the smart meter information 2.

[0041] Furthermore, CIS / CRM106 is a Customer Relation Management system that stores service information 3 and survey information 4. Service information 3 is information indicating the services provided to customers. More preferably, service information 3 is designed by the service provider. Service information 3 is expressed in a linguistic description. Service information 3 may also be recorded in the service provider system 116.

[0042] Furthermore, Questionnaire Information 4 is information showing a questionnaire and responses from customer 109 regarding the attributes of customer 109 and the equipment they possess. Specifically, Questionnaire Information 4 includes questions and answer choices regarding the customer's residential area, family structure, hobbies and preferences, load, power generation, energy storage, and building equipment, as well as quantitative characteristic values ​​shown in the response results. This Questionnaire Information 4 may be new questionnaire information created by the service provider for customers.

[0043] Thus, questionnaire information 4 is an example of customer information showing the attributes of the customer 109 and the equipment they possess. This customer information may be obtained through means other than questionnaires, such as through integration with other systems or from contract information with power companies. Furthermore, CIS / CRM 106 may be implemented as a Customer Information System (CIS). In addition, information from at least one of MDMS 105 and CIS / CRM 106 may be configured to be stored in the storage unit of the digital service design device 100.

[0044] Next, the user terminal device 113 is a terminal device used by users of the digital service design device 100, and it instructs the design of digital services, such as generating new service information, and displays them. For this reason, the user terminal device 113 can be implemented as a computer having an input unit and a display unit. The user terminal device 113 may also be configured to connect to the digital service design device 100 via an external communication network 108. For this reason, the user terminal device 113 may be provided in the service provider system 116. Furthermore, the functions of the user terminal device 113 may be provided in the digital service design device 100.

[0045] Next, an example of an implementation of the digital service design device 100 in this embodiment will be described. The digital service design device 100 can be implemented as a computer such as a server that executes processing according to a program. Figure 1B is a hardware configuration diagram of the digital service design device 100 according to this embodiment. In this embodiment, the digital service design device 100 is implemented as a server, which is an example of a computer, and in particular as a cloud. As shown in Figure 1A, the digital service design device 100 has a processing unit 11, a communication device 12, a memory 13, and a sub-storage device 14, which are connected to each other via a communication path.

[0046] First, the processing unit 11 can be implemented with a processor such as a CPU and executes processing according to the digital service design program 15 stored in the sub-memory 14, which will be described later. The digital service design program 15 will be described later. The communication device 12 corresponds to the receiving and distributing unit 101 in Figure 1A and connects to the external communication network 108 and to the information terminal 111, etc. In the example in Figure 1A, the internal communication network 107 is omitted because the digital service design device 100 stores smart meter information 2, service information 3, and questionnaire information 4, but the communication device 12 may also be connected to the internal communication network 107. The digital service design device 100 may also be provided with an input unit that receives input from the user of the digital service design device 100 and an output unit that outputs information. In this case, these can be implemented with input devices such as a keyboard and display devices such as a display screen.

[0047] Furthermore, the memory 13 and the sub-storage device 14 correspond to the storage unit described above. The memory 13 is where the digital service design program 15 and the information used for processing in the processing unit 11, which are stored in the sub-storage device 14, are loaded. The sub-storage device 14 can be implemented as so-called storage and stores the digital service design program 15, smart meter information 2, service information 3, and questionnaire information 4.

[0048] Furthermore, the secondary storage device 14 may be implemented using various storage media such as an external HDD (Hard Disk Drive), SSD (Solid State Drive), or memory card, or it may be implemented as a separate device from the digital service design device 100, such as the MDMS 105 or CIS / CRM 106, as described above.

[0049] Here, the digital service design program 15 consists of a customer characteristics analysis module 16, a digital service generation module 17, and a demand forecasting module 18. Note that each of these modules may be implemented as an individual program or in some combinations thereof.

[0050] Furthermore, the configuration shown in Figure 1A, which performs the same function as each module, is as follows: Customer Characteristics Analysis Module 16: Customer Characteristics Analysis Unit 102 Digital service generation module 17: Digital service generation unit 103 Demand forecasting module 18: Demand forecasting unit 104 Therefore, the processing unit 11 executes the processing of the customer characteristics analysis unit 102, the digital service generation unit 103, and the demand forecasting unit 104 in accordance with the digital service design program 15. This concludes the explanation of Figure 1B.

[0051] Next, the customer information cycle, which shows the cycle of using customer information in this embodiment, will be explained using questionnaire information 4 as an example. Figure 2 is a diagram illustrating the cycle of questionnaire information 4 in this embodiment. Figure 2 shows the relationship between the service provider and the customer experience via questionnaire information 4. The questionnaire information cycle repeats the collection / registration phase 201, the analysis / classification phase 202, and the utilization phase 203, thereby increasing the added value of the user experience of the customer group 205. The customer group 205 is a concept that represents various customers 109, but includes individual customers 110 and their families, etc.

[0052] First, in the collection and registration phase 201, the user experience needs of diverse customer groups 205 with various lifestyles are collected through information touchpoints 204, such as regular or irregular surveys. Information touchpoints 204 may include physical paper documents, websites, email, social media, and other information distribution apps between the service provider and the user. In this case, user attributes can be collected via the app. Thus, information touchpoints 204 are not limited to surveys; customer information includes information other than survey data 4.

[0053] Furthermore, information touchpoints 204 such as questionnaires will use question and answer formats categorized by purpose, with the purpose of collection clearly defined, to facilitate the extraction of customer characteristics. The questionnaire information 4 collected in this way will be categorized by purpose, such as user hobbies and preferences, values, requests, goals, skills, and experiences, and registered in a database, such as CIS / CRM 106.

[0054] Next, in the analysis and classification phase 202, the collected and registered questionnaire information 4 is obtained from the database, and representative customer characteristic components are analyzed as manifest customer characteristics using analytical methods such as principal component analysis (207). At this time, depending on the characteristics of the service to be provided to the customer group 205, customer characteristic components other than the principal components may also be weighted to make them manifest, thereby representing them as latent customer characteristics. Furthermore, representative customer characteristic components of customer characteristics, or latent customer characteristic components weighted to be representative, are matched with the characteristics of the service to be provided to the customers and the customer characteristics (208). In this way, the customer group 205 is provided with offer and recommendation information 209, point information 210 tailored to the customer's experience, and safe and secure information management 211 for the customer information of the questionnaire information.

[0055] In the utilization phase 203, the service provider retrieves the customer characteristics analyzed by the service provider from the database and provides products, services, and events 212 to the customer group 205 as part of a personalized and optimal user experience. As a result, the customer group 205 will experience user experiences such as self-actualization, enrichment of leisure time, and efficiency of life through the products, services, and events 212. This user experience feedback information 213 is collected through information touchpoints 204, including physical paper documents, websites, email, social media, and other information distribution apps, in order to be used in the design of new services. The user experience feedback information 213 is stored as service information 3 in a database such as CIS / CRM 106. As described above, the cycle of collecting, registering, analyzing, classifying, and utilizing customer information exemplified in survey information 4 enables a positive cycle in which more appropriate services can be provided to the customer group 205.

[0056] Next, we will describe the questionnaire used to collect the questionnaire information 4 in this embodiment. Figure 3 is a diagram showing an example of a questionnaire survey as an information touchpoint in this embodiment. First, questionnaire survey A301 designed for service A consists of a set 302 of questions asking about the basic characteristics of customers and answer choices for each question, and questions asking about the characteristics of customers for service A and answer choices for each question. Similarly, questionnaire survey B303 designed for service B consists of a set 304 of questions asking about the basic characteristics of customers and answer choices for each question, and questions asking about the characteristics of customers for service B and answer choices for each question.

[0057] Here, we will design a new service for the X service business, which will include the A service business and the B service business. In this case, the X service business questionnaire X305 will consolidate the questions and answer choices from questionnaires A301 and B303, and expand the questions and answer choices for each question to inquire about the characteristics of the new service business. As a result, new questionnaire information 4 can be obtained by reusing the questionnaire information 4 for which responses have already been received. However, the collection of questionnaire information 4 is not limited to this method, and the collected questionnaires A301 and B302 may also be used as they are.

[0058] Furthermore, it is desirable that the aggregation of the X Service Business Questionnaire X305 be carried out by the Questionnaire Information Collection Unit, which is an example of a customer information collection unit installed in the digital service design device 100 or the service provider system 116. In addition, it is desirable that the Questionnaire Information Collection Unit transmits the X Service Business Questionnaire X305, which has been aggregated on the information terminal 111, via the communication unit.

[0059] Next, the framework of the digital services targeted by this embodiment will be described. Figure 4 shows an example of a digital service platform in this embodiment. The digital service platform of this embodiment links various service businesses such as railway passenger services, automobiles, food and beverage services, travel services, education services, and finance services. In other words, it targets service providers 401-A to G of this embodiment. These service providers 401-A to G conduct surveys with expanded question and answer options to understand the characteristics of consumers regarding the anticipated digital services through various information touchpoints. Information touchpoints include physical paper documents, websites, email, social networking services (SNS), and other information distribution applications 402-A to G.

[0060] Furthermore, this system aims to synergistically enhance the added value of the user experience for customers 109 by mutually linking the information touchpoints that each service provider 401-A to G had previously provided independently. These information touchpoints include physical paper documents, websites, email, social media, and other information receiving and distribution applications 402-A to G.

[0061] Alternatively, service provider X400 may compile the results of the digital service design obtained through the analysis and classification 207 of the customer characteristic information analysis and classification phase 202 in Figure 2 and the matching 208 between the service and customers 109 (customer group 205 in Figure 2). In this case, service provider X400 may become a digital service platform provider whose service business is to provide this information to other service providers 401-A to G. As a result, the creation (design) and provision of application X is performed as a new service in this embodiment.

[0062] Next, the analysis of customers in this embodiment will be described. First, Figure 5 shows an example of profiling customers from the results of multiple questionnaires in this embodiment. Figure 5 shows an example of profiling the characteristics of the customer group in questionnaire A to a deeper level by using the question items of questionnaire B508, which are related to the profiling result 501 from questionnaire A, as keys. Questionnaire A501 profiles customer types based on the ownership status of utility-related products among a customer group whose residence is a detached house, has 3 to 4 people in the household, and owns an electric vehicle.

[0063] In this case, if the utility-related products are the solar power generation system 114, storage battery 115, electric water heater, and gas water heater shown in the diagram, they can be profiled as follows. Customer type a502, owning solar power generation and electric water heaters. Customer type b503 owns a solar power generation system 114, a storage battery 115, and an electric water heater. Customer type c504 owns solar power generation, battery storage, and gas water heater. Customer type d505 who owns a battery storage system and an electric water heater. • Customer type e506 who owns only an electric water heater. Customer type f507, who owns only a gas water heater.

[0064] On the other hand, in Questionnaire B508, which asks about the ownership of other utility-related products not asked about in Questionnaire A, profiling is performed as follows. First, from the same questions as Questionnaire A501, customer types a, b, d, and e are identified among the 3-4 person households living in detached houses who own electric vehicles and electric water heaters. Then, these customer types a, b, d, and e are further profiled into customer types (509-511) that own induction cooktops and 1-3 air conditioners, do not own any heating equipment other than air conditioners, and whose monthly electricity bill ranges. As a result, it becomes possible to understand the trends of customer types down to a deeper level. Note that although Questionnaire Information 4 was used as an example in Figure 5, at least some of this information can be analyzed using Smart Meter Information 2, for example, monthly electricity bills and usage times. For this reason, in this embodiment, profiling can be performed based on Smart Meter Information 2 and / or Questionnaire Information 4.

[0065] Next, Figure 6 is a diagram illustrating the method for classifying customer characteristics in this embodiment. In other words, Figure 6 shows a determination method in clustering (customer clustering), which is an example of a classification method. Figure 6 shows an example in which the X-means method is used as the clustering method. In the X-means method, the Bayesian Information Criterion (BIC), represented by (Equation 1) below, is used as the cluster partitioning stop criterion to determine the optimal number of clusters.

[0066]

number

[0067] In (Equation 1), the optimal solution is the number of clusters that minimizes the sum of the first term, which represents the goodness of fit of the distribution model by decreasing the error variance estimator between the cluster centroid and each data point in the characteristic value distribution model, and the second term, which represents the complexity of the distribution model by increasing the number of clusters.

[0068] In this embodiment, all clustering result information other than the optimal solution for the number of clusters that minimizes the BIC is also recorded. Then, during the digital service design process, the clustering result in which customer characteristic values ​​that match the service characteristics are extracted as principal components is adopted as the service-appropriate solution. Furthermore, if no principal components of customer characteristic values ​​that match the service characteristics are extracted during the digital service design process, new service characteristics are added interactively during the digital service design process. In this way, the clustering result in which principal components of customer characteristic values ​​that match the service characteristics are extracted is adopted as the service-appropriate solution. In addition to BIC, AIC criteria (Akaike Information Criterion) and GIC criteria (Generalized Information Criterion) may also be used, and clustering may be performed using a combination of weighted sums of BIC, AIC, and GIC values ​​to minimize the result. These weighted sums are set in advance by the operator. The ability to set these information criteria allows for the interactive evaluation of the validity of matching various customer clustering results with service characteristics, which is effective in generating new services. In the following sections, the BIC method will be explained primarily, but it may also be interpreted as an example using a combination of weighted sums of BIC, AIC, and GIC values. The classification of customer characteristics described above will be performed by the customer characteristics analysis unit 102.

[0069] Next, Figure 7 shows the overall processing flow in customer clustering using the X-means method in this embodiment. In Figure 7, first, in step S701, the customer characteristics analysis unit 102 prepares a dataset of the distribution model to be clustered. The dataset consists, for example, of the questionnaire question numbers and the answer choice numbers answered by the customers in the questionnaire information 4. For this purpose, the customer characteristics analysis unit 102 reads the questionnaire information 4, which will be the dataset, from the CIS / CRM 106. In addition to the questionnaire information 4, or instead, the customer characteristics analysis unit 102 may also use smart meter information 2.

[0070] In step S702, the customer characteristics analysis unit 102 sets the initial number of clusters k=1. In step S703, the customer characteristics analysis unit 102 generates k clusters using a specified K-means method.

[0071] In step S704, the customer characteristics analysis unit 102 assigns the dataset read in step S701 to the target cluster. In step S705, the customer characteristics analysis unit 102 calculates the error variance estimate for the distance between the cluster centroid and each data point from the N-dimensional element components for the assigned target cluster. In step S706, the customer characteristics analysis unit 102 calculates the BIC shown in (Equation 1).

[0072] Furthermore, in step S707, the customer characteristics analysis unit 102 evaluates the number of clusters based on the calculated BIC value. For this purpose, the customer characteristics analysis unit 102 retains the current BIC value and also calculates the BIC when k is increased. Furthermore, the customer characteristics analysis unit 102 evaluates the number of clusters based on the BIC value. If the BIC decreases when k is increased (BIC decrease), the process returns to increasing k and performing clustering (step S703). On the other hand, if the BIC increases when the value of k is increased, the clustering result at the previous k is considered a provisional solution, and the process proceeds to step S708.

[0073] Furthermore, in step S708, the customer characteristics analysis unit 102 reads the service characteristics in the digital service design process. For this purpose, the customer characteristics analysis unit 102 reads, for example, the service characteristics contained in the service information 3 of the CIS / CRM 106.

[0074] In step S709, the customer characteristics analysis unit 102 matches customer characteristics with service characteristics in the provisional clustering result. If, as a result, no principal components of customer characteristics values ​​that match the service characteristics are extracted during the digital service design process, the customer characteristics analysis unit 102 determines that the match is valid. In this case, the customer characteristics analysis unit 102 interactively adds new service characteristics during the digital service design process, and the clustering result in which principal components of customer characteristics values ​​that match the service characteristics are extracted is designated as the service-fit solution. If no principal components are extracted (failure), the process returns to step S703. Then, in step S710, the customer characteristics analysis unit 102 outputs the service-fit solution from step S709. However, this step may be omitted.

[0075] Next, Figure 8 shows the overall processing flow of customer profiling in this embodiment. Here, we will also discuss service design based on the profiling results. In Figure 8, first, in step S801, the customer characteristics analysis unit 102 performs processing that applies the X-means method, which corresponds to the processing flow in Figure 7. As a result, the customer characteristics analysis unit 102 generates a suitable clustering solution that matches the service characteristics. In this case, the above-mentioned (Equation 1) is used.

[0076] In step S802, the customer characteristics analysis unit 102 extracts the Nth-order component quantity Gn(A1, A2, A3,···,An) of the cluster centroid in the fitted solution. In step S803, the customer characteristics analysis unit 102 weights the extracted Nth-order component of the cluster centroid as weighting coefficients for the corresponding N question-answer choices to create an input prompt. • Residential area keyword × A1 • Residence type keyword × A2 • Household composition keywords × A3 ... • Hobby-related keywords × An For example, the above input prompt will be created.

[0077] Furthermore, in step S804, the created input prompt is used as input to the generating AI to iteratively generate customer characteristics as a profile in natural language. As a result, quantitative cluster characteristics are expressed in language as a profiling result.

[0078] Furthermore, in step S805, the customer characteristics analysis unit 102 outputs profiling results generated as natural language descriptions by the generating AI. An example of these profiling results (G4) is "a family of 3-4 people, consisting of a couple and young children in a suburban detached house," "middle class, frugal," and "enthusiastic about children's education and extracurricular activities," which are then output. These profiling results are an example of customer characteristics information.

[0079] Then, in step S806, the digital service generation unit 103 generates a digital service design proposal. To do this, the digital service generation unit 103 uses the generated profiling results as input prompts for the generating AI to derive service use cases to be included in the digital service design proposal that are suitable for the customer characteristics.

[0080] As an example of this service use case, the digital service generation unit 103 generates an educational concierge application. In this process, the digital service generation unit 103 compares the profiling results with the functions of the existing service information 3, which are applications 402-A to G, and calculates the difference. The difference is expressed as a language description of a function that can realize a service suitable for the profiling results that cannot be realized or is difficult to realize with the functions of applications 402-A to G. In the example in Figure 8, the educational function is calculated as the difference.

[0081] The digital service generation unit 103 then generates a service use case, which is an example of new service information that complements the differences. For example, the digital service generation unit 103 will generate an educational concierge app as new service information. In this case, the functions of apps 402-A to G that are compared may be limited to apps 402 used by the target customer group. Furthermore, it is more desirable for the digital service generation unit 103 to compare the profiling results with the usage data for each function of apps 402-A to G, which are existing service information 3. If the comparison target is the usage data for apps 402 used by the customer group and each function of apps 402-A to G, the digital service generation unit 103 may also generate unused existing functions and apps 402. Furthermore, the digital service generation unit 103 may also generate service information as a result of integrating these apps.

[0082] Next, we will explain the details of generating the profiling results. Figure 9 is a diagram illustrating the mapping process for generating profiling results in this embodiment. Here, the mapping process is the process of associating the customer clustering results from the survey results with the customer profiling results. Therefore, Figure 9 shows an example of mapping customer clusters based on the clustering results with the characteristics that become the main components of the customer profiling results from the customer characteristics in each cluster.

[0083] Figure 9 shows the N-th order space for the N questions Qn in a consumer survey. The questions Qn in Figure 9 are as follows. First, Q1 to Q6 are questions that describe the basic characteristics of the household. • Q1: Question 901 asking about top residential area characteristics such as prefectures, • Q2: Question 902 asking about top-level residential area characteristics such as prefectures, • Q3: Question 903 asks about the type of residence, such as detached house, apartment building, owned home, or personal residence. • Q4: Question 904 asks about household composition characteristics such as single person, couple only, with children, or two-family household. • Q5: Question 905 asking about the number of people in a household, such as 1 to n people. • Q6: Question 906 asks about household age characteristics, such as the youngest and oldest ages in the household.

[0084] Furthermore, questions Q1 through Qn are questions that inquire about the customer's lifestyle. • Ql: Question 907 asks about household income characteristics, such as the main sources of household income, such as salaries and pensions, and their approximate amounts. • Qm: Question 908, which asks about the characteristics of household products owned by the household, such as the types of household products and the number of units they possess. • Qn: Question 909, which inquires about the tastes and preferences of consumers.

[0085] Figure 9 shows an N-th order space where i customers Mi(A1,A2,···,An)910 are distributed, with An being the normalized answer choice number for each of these questions. In this N-th order space where the customer elements from the questionnaire survey results for customers 109 are distributed, the customer characteristic analysis unit 102 performs the clustering shown in Figure 7, forming clusters such as the following. For example, clusters 911 with G1 as the centroid, cluster 912 with G2 as the centroid, cluster 913 with G3 as the centroid, and cluster G914 with G4 as the centroid are formed as groups of customer elements. The customer characteristic analysis unit 102 extracts the principal component characteristics for each of these clusters and performs the language generation step S804 of customer profiling by the generating AI shown in Figure 8. As a result, the customer characteristic analysis unit 102 generates the following profiling results.

[0086] For example, for cluster 911 of customer group G1, a relatively middle-income household 915 is generated consisting of a single person on a business trip living in company housing in an urban area. Similarly, for cluster 912 of customer group G2, a relatively low-income household 916 is generated consisting of a single university student with an interest in entertainment living in an urban apartment. Furthermore, for cluster 913 of customer group G3, a relatively high-income household 917 is generated consisting of an elderly couple with an interest in arts living in a suburban apartment complex. Finally, for cluster 914 of customer group G4, a relatively middle-income household 918 is generated consisting of a young parent and child in elementary or middle school who are frugal and living in a suburban detached house.

[0087] Next, Figure 10 shows the correspondence between the customer profiling results and the proposed digital service design in this embodiment. Customer group classification result 1000 corresponds to the customer group that is the result of the customer profiling shown in Figure 9. The customer group classification result 1000 is as follows. • Customer Group G1: Energy-Saving Focused Group 1001 • Customer Group G2: Gourmet-oriented household group 1002, • Customer group G3: Automobile-oriented household group 1003 • Customer Group G4: Education-Oriented Household Group 1004 • Customer Group Gn: A characteristic-oriented household group 1005 in which customer characteristics other than those mentioned above were extracted as the main component in the results of customer profiling.

[0088] Digital service design proposal 1006 corresponds to each of these customer groups as follows. Note that digital service design proposal 1006 is the digital service design proposal generated in step S806 of Figure 8. • Digital service design proposal S1: Energy management service 1007, generated from the profiling results of energy-saving-oriented household groups, corresponds to customer group G1. • Digital service design proposal S2: The meal reservation and food review service 1008, generated from the profiling results of a gourmet-oriented household group, will be handled by customer group G2. • Digital service design proposal S3: Car-sharing drive plan service 1009, generated from the profiling results of car-mobility-oriented household groups, corresponds to customer group G3. • Digital service design proposal S4: Educational concierge service 1010, generated from the profiling results of education-oriented household groups, corresponds to customer group G4. • Digital service design proposal Sn: Customer characteristic-oriented service 1011, generated from the profiling results of characteristic-oriented household groups in which customer characteristics other than those mentioned above are extracted as the main components, corresponds to customer group Gn.

[0089] As described above, customer characteristic-oriented household groups and customer characteristic-oriented services are matched. In this way, the digital service generation unit 103 generates service information indicated by the digital service design proposal from the customer characteristic information indicated by the customer characteristic-oriented household groups. This concludes the explanation of the processing flow of this embodiment, and next, the information used in this embodiment will be described.

[0090] Figure 11 shows the smart meter information in this embodiment. In this embodiment, smart meter information 2 is stored in the MDMS 105 managed by the power company. The smart meter information 2 is information about the amount of electricity demanded by the customer 109, collected from the smart meter 112. Therefore, the smart meter information 2 consists of the following items: meter ID 1101, customer ID 1102, contract type 1103, contract capacity 1104, measurement date and time 1105, and electricity amount 1106.

[0091] Meter ID 1101 is an ID that identifies the smart meter 112 individually. Customer ID 1102 is an ID that identifies the customer 109 who has a power contract. Contract type 1103 indicates the type of power contract.

[0092] Furthermore, the contracted capacity 1104 indicates the maximum contracted power capacity. The measurement date and time 1105 indicates the start and end time periods for measuring electricity usage. In addition, the electricity usage 1106 indicates the amount of electricity used during that time period.

[0093] Here, the power company calculates the electricity consumption of each customer 109 based on the collected smart meter information 2, calculates the monthly electricity charge based on the electricity contract, and bills each customer 109 for the electricity charge. Currently, electricity meters are being replaced from conventional mechanical and electronic meters to smart meters, which can measure electricity consumption every 30 minutes and even in finer time units. As a result, many digital information services have been proposed in recent years, partly due to the establishment of information banks.

[0094] The customer characteristics analysis unit 102 then uses smart meter information 2 (data on electricity consumption) from the smart meter 112 to create a vector of time-series data on a daily basis (e.g., 48 points / day) and performs classification using machine learning. For example, the customer characteristics analysis unit 102 selects a specific day from each of the four seasons (spring, summer, autumn, and winter) for a given customer. The customer characteristics analysis unit 102 then creates a vector with 48 × 4 = 192 components representing the electricity consumption of the four days, generates n vectors from multiple customers n, and performs clustering. As for the clustering method, for example, one of the following may be used: K-means, X-means, G-means, or support vector machine classification.

[0095] Alternatively, multiple clustering results can be used in combination, and the operator can visually confirm and select the most suitable clustering result. It should be noted that the demand of each customer changes due to environmental data such as temperature and humidity, as well as changes in contracts, etc. Therefore, it is necessary to re-cluster the data after making demand forecasts that take such changes into account.

[0096] Therefore, the demand forecasting unit 104 performs demand forecasting using weight coefficients with temperature and demand data as explanatory variables. For this purpose, the demand forecasting unit 104 can selectively use a method that uses multiple regression analysis on past temperature and demand data or a Kalman filter method, which the operator has set for the reference range of past data, in order to calculate the weight coefficients. In other words, the demand forecasting unit 104 predicts the amount of demand using the learning results from one of the following: multiple regression analysis, Kalman filter, or ensemble learning function, with environmental data and smart meter information as input.

[0097] On the other hand, regarding predictions in general, while expanding the scope of reference data from past measurements improves prediction accuracy, it also increases computation time. Furthermore, narrowing the scope of reference data from past measurements reduces computation time but decreases prediction accuracy. In other words, there is a trade-off between prediction accuracy and computation time.

[0098] Therefore, in this embodiment, the demand forecasting unit 104 performs forecasting offline in advance when the reference range of measurement data, such as smart meter information 2 and environmental data, is wide. Also, when the reference range of measurement data is narrow, the demand forecasting unit 104 performs demand forecasting online sequentially. In this way, the demand forecasting unit 104 of this embodiment is equipped with ensemble learning that combines offline forecasting and online forecasting, thereby reducing calculation time while suppressing forecasting errors. In other words, in this embodiment, forecasting accuracy can be maintained while reducing calculation time by combining reference patterns of past measurement data, and forecasting accuracy can be improved by appropriately correcting weight coefficients and forecasting error gains with the measurement data of the day.

[0099] Figure 12 shows a proposed digital service design having a service use case 1200 and its service content 1205 in this embodiment. In Figure 12, each record represents a proposed digital service design, and the service use case 1200 and its service content 1205 are associated with each as follows.

[0100] In digital service design proposal S1, service content 1206 is associated with a personalized energy management app 1201 for energy-saving-oriented households, which provides visualization of energy usage and advice on saving energy. In digital service design proposal S2, service content 1207 is associated with a restaurant reservation and food review sharing management app 1202 for gourmet-oriented households, which provides local restaurant reservations, food reviews, and suggestions for special dining experiences. In digital service design proposal S3, the following service content is associated with a car-sharing and drive planning management app 1203 for drive-oriented households: service content 1208 is associated with providing a comfortable driving and travel experience by suggesting driving routes, guiding users to gas stations and EV charging spots, and making reservations for parking, restaurants, and hotels at travel destinations. Furthermore, in digital service design proposal S4, service content 1209 is associated with an education concierge app 1204 for education-oriented households, which supports children's education. Services 1209 that support children's education include, for example, creating study plans, test preparation, timing for preparation and review, and introducing extracurricular activity classes.

[0101] The following provides a more detailed explanation of the service use cases 1200 and service content 1205 in service design proposals S1, S2, S3, and S4.

[0102] For S1 energy-saving households, the personalized energy management app will include the following features (a new app will be created or the following features will be added):

[0103] • Peak energy usage warning: This feature sends notifications to users encouraging them to conserve electricity during peak electricity demand hours. This helps to reduce consumption during times when electricity supply is tight.

[0104] • Energy use forecast: This feature uses AI to learn the user's energy usage patterns and predict future energy consumption. This allows users to plan their energy use more effectively.

[0105] Setting energy saving targets and displaying achievement levels: This feature allows users to set their own energy-saving goals and check their progress in real time. This helps users maintain their motivation to conserve energy.

[0106] Comparison display of energy consumption: This feature displays a comparison of energy usage with other users in the same area or with similar family structures. This allows users to verify whether their energy usage is appropriate.

[0107] Recommendations for energy-efficient products: This feature recommends energy-efficient products based on the user's energy usage. This makes it easy for users to find products that help improve their energy efficiency.

[0108] By adding these features, the app can go beyond simply visualizing energy usage and providing energy-saving advice; it can offer support to users in saving energy more effectively. These features provide users with a deeper understanding of their energy use and concrete means to practice energy conservation in their daily lives. This can help energy-conscious households optimize their energy use and improve energy efficiency, thereby enhancing the value of the user experience.

[0109] The S2 restaurant reservation and food review sharing and management app for gourmet-conscious households will include the following features (a new app will be created or the following features will be added).

[0110] Ingredient information: The service provides information about the ingredients used by restaurants. This allows users to learn about the freshness, origin, nutritional value, and other details of the ingredients.

[0111] Chef's profile: The website provides profiles of the chefs at each restaurant, allowing users to understand the chefs' backgrounds and culinary styles.

[0112] Cooking recipes: Some restaurants offer recipes for specific dishes, allowing users to recreate those dishes at home.

[0113] Calorie counting for meals: The system calculates the calories of the dishes served and provides this information to the user. This allows users to manage their health more effectively.

[0114] Restaurant event information: The service provides information on special events and campaigns hosted by restaurants, allowing users to enjoy unique dining experiences.

[0115] Exchanging dishes between users: This feature allows users to post photos and reviews of their dishes and share them with other users. This helps users gain new cooking ideas.

[0116] By adding these features, the app can go beyond simply making restaurant reservations and sharing food reviews, offering support to help users enjoy gourmet dining experiences more deeply. These features provide users with concrete means to understand food more deeply and pursue gourmet dining experiences in their daily lives. This can help gourmet-conscious households maximize their enjoyment of meals and improve the quality of their food, thereby enhancing the value of the user experience.

[0117] For S3 live-oriented households, the car-sharing and drive planning management app will include the following features (a new app will be created or the following features will be added):

[0118] Providing traffic information: We provide real-time traffic information (congestion, construction, accidents, etc.) to enable users to travel smoothly.

[0119] Weather information provided: It provides weather information along the driving route, allowing users to make appropriate preparations.

[0120] Recommended tourist spots: The system recommends tourist spots based on the user's interests and travel purpose.

[0121] Drive scoring: It analyzes users' driving behavior and provides feedback to encourage safe driving.

[0122] Eco-driving suggestions: We propose driving methods to optimize fuel efficiency.

[0123] Vehicle management: It provides information related to vehicle management, such as vehicle maintenance schedules and insurance renewal reminders.

[0124] By adding these features, we can provide support for users to have a more comfortable and safer driving experience, going beyond mere car sharing and drive planning. These features provide concrete means for users to enjoy driving more deeply. This helps drive-oriented households maximize the enjoyment of driving and improve the quality of their driving. These features can provide concrete means for users to enjoy driving more deeply and can enhance the value of the user experience.

[0125] In S4's education concierge app for education-oriented households, the customer group is interested in features that support children's education, such as creating study plans, test preparation, notifications for pre-study and review timing, and introductions to extracurricular activity classes. Based on these results, the education concierge app is effective in improving customer value as one of the service menus delivered to this customer group. In this embodiment, the lifestyles of customers can be analyzed accurately by time of day and location.

[0126] Furthermore, in this embodiment, by integrating and utilizing survey data and smart meter information, the user's lifestyle, energy usage patterns, and preferences are comprehensively understood. Then, by utilizing IoT and DX technologies, data collection and analysis are carried out continuously, whether regularly or irregularly, to create a more precise user profile. This makes it possible to provide personalized digital services tailored to individual needs. The present invention contributes to improving user satisfaction and business results.

[0127] Furthermore, based on information regarding each customer's residential area, family structure, hobbies and preferences, load, power generation, energy storage, and building facilities, it is possible to accurately profile each customer's lifestyle by household, by household member, and by time of day. In addition, according to the present invention, it is possible to understand each customer's lifestyle by time of day, so that digital services can be accurately provided to each customer and the value of the customer experience can be enhanced.

[0128] Furthermore, in this embodiment, the customer characteristics analysis unit 102 classifies customers into multiple customer groups with quantitative characteristics based on the questionnaire information and smart meter information, and converts them into qualitative language descriptions. In addition, according to this embodiment, by classifying customers into multiple customer groups with quantitative characteristics based on at least one of the questionnaire information 4 and smart meter information 2, and converting them into qualitative language descriptions, the lifestyle of each customer can be easily understood by a person.

[0129] Furthermore, in this embodiment, the digital service generation unit 103 associates the correlation between the qualitative language description and the performance data of the service information based on similarity, and complements the difference of the service information 3 to generate new service information. According to this configuration, it becomes possible to accurately provide digital services that further enhance the customer experience value to each customer.

[0130] Furthermore, in this embodiment, the customer characteristic analysis unit 102 appropriately classifies customers using a clustering method and creates a customer group with similar characteristics. By doing this, since it is possible to appropriately classify customer groups with similar characteristics, it is possible to accurately determine the customers who are the targets of digital service provision.

[0131] Furthermore, in this embodiment, the digital service generation unit 103 utilizes the correlation between the qualitative language description and the performance data in order to generate customized service information. By doing this, since it utilizes the correlation between the qualitative language description and the performance data, it is possible to catch up with the changes in the lifestyle caused by the changes in each customer's family attribute information and device information based on the questionnaire information regularly implemented.

[0132] Furthermore, due to the addition of the characteristics of power consumption, which is the demand amount of the smart meter information 2 according to the usage status of home appliances, and the possibility of combining the quantitative characteristics with the customer's family attribute information, it becomes possible to efficiently update the marketing accompanying the changes in the lifestyle.

Explanation of Signs

[0133] 100: Digital service design device 101: Reception / transmission unit 102: Customer characteristic analysis unit 103: Digital service generation unit 104: Demand prediction unit 105: MDMS 106: CIS / CRM 107: Internal communication network 108: External communication network 109: Needer's family 110: Needer 111: Information terminal 112: Smart meter 113: User terminal device 114: Solar power generation device 115: Storage battery 116: Service provider system

Claims

1. In a digital service design device that generates service information indicating services for multiple energy consumers, A storage unit that stores customer information indicating the attributes of the plurality of customers and equipment, demand information indicating the amount of energy demanded by the customers, and service information indicating the characteristics of the services that the service provider provides to the plurality of customers in a qualitative language description, Using the customer information and / or the demand information, cluster analysis is performed on the quantitative characteristics of each of the multiple customers according to weight parameters that have been set in advance to associate the service information with each of the multiple customers, thereby classifying the multiple customers into multiple customer groups. A customer characteristic analysis unit identifies customer characteristic information that represents the characteristics of each of the classified customer groups in qualitative language descriptions, based on the characteristic values ​​of each customer group and the language description of the customer information. A digital service design apparatus having a digital service generation unit that identifies service information containing a qualitative language description similar to the qualitative language description of the identified customer characteristics information, compares the customer characteristics information and the identified service information to calculate the difference, and generates new service information indicating a new service that complements the difference.

2. In the digital service design apparatus according to claim 1, The digital service generation unit is a digital service design device that identifies the service information based on performance data indicating the service performance indicated by the service information.

3. In the digital service design apparatus according to claim 2, The aforementioned customer information is survey information, which is the customer's response to a questionnaire, and includes the customer's residential area, family structure, hobbies and preferences, load, power generation, energy storage, and building, as part of a digital service design device.

4. In the digital service design apparatus according to claim 1, Furthermore, it has a demand forecasting unit that predicts the demand volume of each of the aforementioned multiple customers, The aforementioned customer characteristics analysis unit is a digital service design device that further uses the predicted demand volume information to classify the plurality of customers into a plurality of customer groups.

5. In the digital service design apparatus according to claim 4, The demand forecasting unit is a digital service design device that forecasts the amount of demand using the learning results from one of the following: multiple regression analysis using environmental data and smart meter information as input, a Kalman filter, or an ensemble learning function that combines offline and online forecasting.

6. In a digital service design method executed by a digital service design device that generates service information indicating services for multiple energy consumers, The digital service design device comprises a storage unit, a customer characteristics analysis unit, and a digital service generation unit. The memory unit stores customer information indicating the attributes of the plurality of customers and equipment, demand information indicating the amount of energy demanded by the customers, and service information indicating the characteristics of the services that the service provider provides to the plurality of customers in qualitative language descriptions. The aforementioned customer characteristics analysis unit, Using the customer information and / or the demand information, cluster analysis is performed on the quantitative characteristics of each of the multiple customers according to weight parameters that have been set in advance to associate the service information with each of the multiple customers, thereby classifying the multiple customers into multiple customer groups. Based on the characteristic values ​​of each classified customer group and the linguistic description of the customer information, customer characteristic information is identified that represents the characteristics of each customer group in a qualitative linguistic description. A digital service design method comprising: a digital service generation unit identifying service information that includes a qualitative language description similar to the qualitative language description of the customer characteristics information; comparing the customer characteristics information with the identified service information to calculate the difference; and generating new service information indicating a new service that complements the difference.

7. In the digital service design method described in claim 6, The digital service generation unit identifies the service information based on performance data indicating the service performance indicated by the service information, and is a digital service design method.

8. In the digital service design method described in claim 7, The aforementioned customer information is survey information, which is the customer's response to a questionnaire, and includes the customer's residential area, family structure, hobbies and preferences, load, power generation, energy storage, and building.

9. In the digital service design method described in claim 6, The digital service design device further includes a demand forecasting unit that predicts the demand volume of each of the multiple customers, The aforementioned customer characteristics analysis unit further uses the predicted demand volume information to classify the plurality of customers into a plurality of customer groups in a digital service design method.

10. In the digital service design method described in claim 9, The demand forecasting unit predicts the amount of demand using the learning results obtained from one of the following: multiple regression analysis using environmental data and smart meter information as input, a Kalman filter, or an ensemble learning function that combines offline and online forecasting. This is a digital service design method.

11. In a digital service design system having a digital service design device, an information terminal, and a smart meter that generate service information indicating services for multiple energy consumers, The aforementioned digital service design device is A receiving and distributing unit that receives customer information indicating the attributes of the multiple customers and equipment from the information terminal and the smart meter via an external communication network, A storage unit that stores the aforementioned customer information, demand information indicating the amount of energy the customer demands, and service information that describes the characteristics of the services that the service provider provides to the plurality of customers in a qualitative language. Using the customer information and / or the demand information, cluster analysis is performed on the quantitative characteristics of each of the multiple customers according to weight parameters that have been set in advance to associate the service information with each of the multiple customers, thereby classifying the multiple customers into multiple customer groups. A customer characteristic analysis unit identifies customer characteristic information that represents the characteristics of each of the classified customer groups in qualitative language descriptions, based on the characteristic values ​​of each customer group and the language description of the customer information. A digital service design system having a digital service generation unit that identifies service information containing a qualitative language description similar to the qualitative language description of the aforementioned customer characteristics information, compares the aforementioned customer characteristics information with the identified service information to calculate the difference, and generates new service information indicating a new service that complements the difference.

12. In the digital service design system according to claim 11, The digital service generation unit is a digital service design system that identifies the service information based on performance data indicating the service performance indicated by the service information.

13. In the digital service design system according to claim 12, The aforementioned customer information is survey information, which is the customer's response to a questionnaire, and includes the customer's residential area, family structure, hobbies and preferences, load, power generation, energy storage, and building, as part of a digital service design system.

14. In the digital service design system according to claim 11, The digital service design device further includes a demand forecasting unit that predicts the demand volume of each of the multiple customers, The aforementioned customer characteristics analysis unit is a digital service design system that further uses the predicted demand volume information to classify the multiple customers into multiple customer groups.

15. In the digital service design system according to claim 14, The demand forecasting unit is a digital service design system that forecasts the amount of demand using the learning results from one of the following: multiple regression analysis using environmental data and smart meter information as input, Kalman filtering, or an ensemble learning function that combines offline and online forecasting.

16. A digital service design apparatus according to any one of claims 1 to 5, A digital service device that provides the service information to multiple consumers who utilize the aforementioned energy, A digital service system comprising a receiving and distributing unit that receives the customer information and / or the demand quantity information with individual customers and / or distributes the service information, wherein the digital service device distributes the service information generated by the digital service design device to customer groups identified by the customer characteristic information in the customers in the predetermined region, either periodically or irregularly, via an external communication network in a predetermined region.

Images