Information processing device, information processing method, and program

Abstract

To efficiently and accurately conduct surveys on the classification of a large number of building uses. [Solution] The nameplate information acquisition unit 52 acquires nameplate information and address information from a residential map and links it with building data. The business type estimation unit 53 performs a web search using the nameplate information and address information to estimate the business type of the building. The usage classification determination unit 55 converts the estimated business type into a usage classification based on classification criteria. The confidence calculation unit 56 calculates a confidence score for the usage classification and identifies ambiguous areas. The map display control unit 57 displays the usage classification in color on the map in relation to the building shape and creates a current status map by building usage. This realizes the automation of determining the building's usage classification.

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an information processing method, and a program.

Background Art

[0002] Conventionally, in municipalities across the country, basic urban planning surveys are generally conducted once every five years. This is a survey for periodically grasping the current situation and future prospects of cities based on the Urban Planning Law, providing important data for promoting urban development and optimizing land use, and being utilized for formulating and reviewing urban plans. Among them, there is a classification survey of building uses, and it is necessary to investigate what uses all buildings within the urban planning area are utilized for. In this survey, it is necessary to target tens of thousands of buildings even in a single city. Conventionally, surveyors of contractors entrusted with the survey confirmed signs and nameplates of each building on-site to conduct the use survey. However, subsequently, desk research was combined with reference to materials such as residential maps to improve efficiency. At that time, the residential map was visually confirmed, each building was classified based on the implementation guidelines for basic urban planning surveys, and then the results were manually transcribed onto the drawing, or directly input as attribute information of building data by operating GIS (Geographic Information System) or the like. At that time, there was a technique of creating a database in advance with characteristic keywords that could be included as the name of the building, collating the building name information with the keywords, and automatically estimating the building use on the system (see, for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Thus, in urban planning basic surveys, it is necessary to classify the use of tens of thousands of buildings in a single city, and this required a great deal of time and effort using traditional manual methods. Furthermore, it was observed that the actual use of a building could not be determined solely from its name on the residential map, that standards were not standardized among multiple workers, and that incorrect usage classifications were sometimes made. Therefore, there is a need to solve the problem of not being able to efficiently and accurately conduct surveys of the use classification of a large number of buildings, but conventional technologies, including the technology described in Patent Document 1, are unable to adequately meet this need.

[0005] This invention was made in consideration of these circumstances, and aims to enable efficient and accurate surveys of the use classification of a large number of buildings. [Means for solving the problem]

[0006] To achieve the above objective, an information processing apparatus according to one aspect of the present invention is: A means for acquiring building data, including building location information, A means for acquiring nameplate information of the aforementioned building and associating it with the aforementioned building data, A classification criteria setting means for setting classification criteria for determining the use classification from the nameplate information of the aforementioned building, A means for determining the use of a building based on the classification criteria set by the classification criteria setting means, and based on the nameplate information acquired by the nameplate information acquisition means, A map display control means that performs control to display the usage classification determined by the usage classification determination means on a map in association with the building data, It is equipped with.

[0007] Each of the information processing method and program according to one aspect of the present invention corresponds to each of the method and program corresponding to the information processing apparatus according to one aspect of the present invention. [Effects of the Invention]

[0008] According to the present invention, by automating the entire process from acquiring building data to associating it with nameplate information, determining the usage classification based on classification criteria, and displaying it on a map, it becomes possible to efficiently and accurately investigate the usage classification of a large number of buildings. [Brief explanation of the drawing]

[0009] [Figure 1] This figure shows an overview of the service that can be realized by an information processing system to which a server according to one embodiment of the information processing device of the present invention is applied. [Figure 2] This figure shows an example of the configuration of an information processing system to which a server according to one embodiment of the information processing device of the present invention is applied. [Figure 3] Figure 2 is a block diagram showing an example of the server hardware configuration in the information processing system. [Figure 4] This is a functional block diagram showing an example of the functional configuration of the server in Figure 3 that constitutes the information processing system in Figure 2. [Figure 5] This flowchart shows an example of the processing flow for this service. [Figure 6] This diagram shows an example of linking building data with nameplate information. [Figure 7] This figure shows an example of a conversion table from business type to usage classification. [Figure 8] This diagram shows an example of how to display a current status map categorized by building use. [Figure 9] This figure shows an example of automatic recognition of residential maps. [Figure 10] This figure shows an example of its application to basic urban planning surveys. [Modes for carrying out the invention]

[0010] Embodiments of the present invention will be described below with reference to the drawings.

[0011] In addition, the main terms used in this specification are defined as follows. The "residential map" is a commercially available map on which the situation of a town (mainly the terrain such as roads and buildings, addresses, etc.) is described and which specifies what buildings are there. Specifically, the issuing company of the residential map sends investigators to visit the site and investigates the so-called "nameplate information" such as the nameplate installed on each building or the name of the main entrance in the case of a general residence, and uses a method of marking it on the map. In major cities, it is surveyed and issued every year, and in other regions, at intervals of about two to five years. The "nameplate information" refers to the name information described on the nameplate, name of the main entrance, signs, etc. installed on the building, and is the identification information of the building described in the residential map. The "business type" refers to information indicating the type of business conducted in the building or the actual usage form of the building (e.g., ramen shop, dental clinic, condominium, etc.). The "usage classification" refers to the classification of buildings by usage purpose based on the guidelines for urban planning basic surveys or the criteria set by the user (e.g., residence, commercial facility, factory, etc.). The "basic urban planning map" is a drawing (topographic map) prepared by municipalities, etc. in which roads, rivers, railways, building shapes, and contour lines representing the undulation of the land are described with accurate position information for the entire area of the municipality or within the urban planning area. It is the basic map for municipalities, etc. to implement each measure of urban planning. The "building data" is digital data representing the position and shape of a building in GIS, and includes at least coordinate information indicating the outer shape of the building. The "polygon figure" refers to graphic data of a closed polygon formed by connecting a plurality of coordinate points. The "confidence level" refers to a numerical index indicating the certainty of the determination result of business type estimation or usage classification. The "public coordinate" refers to a coordinate system based on the plane rectangular coordinate system of Japan.

[0012] First, referring to FIG. 1, an overview of a service (hereinafter referred to as "this service") that can be realized by an information processing system to which a server according to an embodiment of the information processing apparatus of the present invention is applied (see FIG. 2 described later) will be described. Figure 1 shows an overview of the service that can be realized by an information processing system to which a server according to one embodiment of the information processing device of the present invention is applied.

[0013] This service enables the automatic determination of the land use classification for a large number of buildings.

[0014] Specifically, Figure 1 shows the processing flow of this service.

[0015] First, in step S1, Server 1 acquires building data, nameplate information, and address information from digitized urban planning base maps and other topographic map data, as well as residential maps (digital version), and links them using location information. Server 1 acquires (extracts) building data representing the shape and location information of buildings included in topographic map data from GIS data server 6, and acquires nameplate information and address information including location information included in residential maps from data provision server 4. Server 1 stores nameplate information and address information corresponding to each building, based on the location information of the building data.

[0016] Next, in step S2, server 1 performs a web search using the nameplate information and address information for each building, and obtains search results such as customer reviews on store websites and social media, and real estate information (rental and sales). Server 1 then uses a generative AI, for example, employing RAG (Retrieval Augmented Generation) technology, to estimate the type of business each building operates from web information. Furthermore, if the nameplate information contains a common name, the generation AI recognizes it as a name and automatically determines it to be a private residence without performing a web search. This demonstrates a significant efficiency improvement in processing private residences, which make up the majority of buildings, through the use of generation AI. In this process, the generating AI is asked to calculate a confidence score representing the degree of confidence in the estimation result, and ambiguous areas are identified. Additionally, the generating AI may be asked to provide the basis for its estimation and the reason for any errors that occurred if the confidence score was lower than a predetermined standard. Furthermore, by adding address information to the search in addition to nameplate information, it becomes possible to accurately determine the type of business. For example, even if the nameplate information is the same, "Maison Pasco," Server 1 can determine that the "Maison Pasco" in Meguro Ward is a "restaurant" because it is found in restaurant review websites, while the "Maison Pasco" in Adachi Ward is a "apartment building" because it is found in apartment rental information. In this way, by using address information in conjunction with other information, Server 1 can accurately identify the relevant web information even if there are different buildings with the same nameplate information, and can accurately determine the actual type of business for each building.

[0017] Next, in step S3, Server 1 converts the business type of each building, estimated by a generative AI using, for example, RAG technology, into a usage classification based on the standard classification items and criteria of the urban planning basic survey or classification items and criteria set independently by the user, and at the same time calculates a confidence score to identify any ambiguous areas. Once Server 1 can identify the type of business in each building, it assigns each type of business to the corresponding classification item, similar to how classification items are used in the basic urban planning survey. In other words, Server 1 formulates classification criteria for each type of business, feeds these criteria into a generating AI, and then the generating AI sorts each building into its appropriate usage category according to its type of business. Alternatively, the generating AI may be made to formulate the classification criteria for each type of business. Server 1 can accept custom classification items and criteria from users. This makes it possible to adapt the system not only to standard classification criteria for urban planning basic surveys, but also to various other surveys of current conditions, such as surveying the distribution of apartment buildings in real estate market research, or extracting the locations of stores of specific types of businesses for trade area analysis when opening restaurants or other stores.

[0018] Finally, in step S4, Server 1 adds the classification results as attribute information to the building data on the topographic map, displays the map, creates a current status map by building use, and displays it in different colors on the map. Once the usage classification of each building is determined, Server 1 assigns these classifications as attribute information to the shape of each building in the building data. Specifically, for example, Server 1 displays the shape of each building in the building data in conjunction with other map items of the topographic map, such as the urban planning base map, by color-coding and classifying them based on the building's use classification, thereby enabling a visual understanding of the distribution of building uses.

[0019] Although not shown in Figure 1, this service may also involve Server 1 comparing nameplate information from a predetermined point in the past with current nameplate information and performing business type estimation processing only for buildings where changes have occurred. In other words, Server 1 can obtain nameplate information from a predetermined point in the past, compare it with the current nameplate information, and identify buildings that have undergone changes. As a result, the use classification of buildings whose nameplate information has not changed does not need to be revised; the classification results of the urban planning basic survey at the same point in time are obtained and used as is, and only the use classification of buildings whose nameplate information has changed needs to be revised. This will allow for even shorter processing times.

[0020] In this service, although not shown in Figure 1, Server 1 may, instead of using a generating AI to estimate the type of business from the nameplate information of each building, acquire or formulate classification criteria such as urban planning basic surveys for the nameplate information, and determine the usage classification directly from the nameplate information. In other words, Server 1 may directly determine the usage classification using a generative AI that utilizes RAG technology.

[0021] In this service, although not shown in Figure 1, Server 1 may acquire nameplate information from past residential maps and information on building use classifications already classified in the current status map of building use from the urban planning basic survey at the same time as actual data, generate a machine learning model that uses conversion examples such as "in the case of this nameplate information, it will be this use classification" as training data, and use the generated machine learning model as the classification criterion (conversion criterion) from nameplate information to use classification. In other words, Server 1 can reproduce the classification criteria for land use classification using a learning model based on actual data generated from historical residential maps and information from urban planning basic surveys. This makes it possible to determine land use classifications with high accuracy by applying current nameplate information from residential maps to classification criteria recreated using past information (residential maps and results from urban planning basic surveys).

[0022] Although not shown in Figure 1, in this service, Server 1 may display on a map ambiguous areas where the mechanical judgment could not accurately determine the type of business or usage classification, allowing workers to visually confirm these areas. In other words, Server 1 can output a map displaying, in an identifiable format, building shapes whose confidence scores, calculated by machine learning models or generative AI, are lower than a predetermined standard. Furthermore, it can also refer to the basis for the estimated business type and usage classification by the generative AI, as well as the reasons for the low confidence scores. This makes it possible to visualize areas that could not be determined mechanically (ambiguous areas) and the basis for those determinations, making it easier for workers to verify them, thus further improving accuracy.

[0023] In this way, compared to the conventional method which was done manually, it is possible to improve accuracy, increase operational efficiency, speed up operations, and reduce costs.

[0024] Next, with reference to Figure 2, we will describe the configuration of an information processing system to which an information processing system that realizes the provision of the above-mentioned service, i.e., an information processing system to which a server according to one embodiment of the information processing device of the present invention is applied. Figure 2 shows an example of the configuration of an information processing system to which a server according to one embodiment of the information processing device of the present invention is applied.

[0025] The information processing system shown in Figure 2 is configured to include Server 1, User Terminal 2, Administrator Terminal 3, Data Provision Server 4, Web Information Server 5, and GIS Data Server 6. Server 1, user terminal 2, administrator terminal 3, data provision server 4, web information server 5, and GIS data server 6 are interconnected via a network N such as the Internet. Furthermore, it is preferable that Server 1, which handles prompt input and response result acquisition for the generation AI, and Web information server 5 support MCP (Model Context Protocol), which is a common protocol for generation AI.

[0026] Server 1 is an information processing device managed by the service provider of this service (Figure 1). Server 1 performs various processes necessary to realize this service while communicating as needed with user terminal 2, administrator terminal 3, data provision server 4, web information server 5, and GIS data server 6. User terminal 2 is an information processing device operated by user U, and consists of a smartphone, tablet, personal computer, etc. Administrator terminal 3 is an information processing device operated by administrator M, and consists of a smartphone, tablet, personal computer, etc. Note that administrator terminal 3 may also be operated directly by user U. Data provision server 4 is an information processing device that provides residential map data, etc. Web information server 5 is an information processing device that provides search APIs, etc. GIS data server 6 is an information processing device that provides building data and other information.

[0027] Figure 3 is a block diagram showing an example of the server hardware configuration in the information processing system shown in Figure 2.

[0028] Server 1 comprises a CPU (Central Processing Unit) 11, ROM (Read Only Memory) 12, RAM (Random Access Memory) 13, a bus 14, an input / output interface 15, an input unit 16, an output unit 17, a storage unit 18, a communication unit 19, and a drive 20.

[0029] The CPU 11 executes various processes according to the program recorded in the ROM 12 or the program loaded from the storage unit 18 into the RAM 13. RAM13 also stores data and other information necessary for the CPU11 to perform various processes.

[0030] The CPU 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output interface 15 is also connected to this bus 14. The input / output interface 15 is connected to an input unit 16, an output unit 17, a storage unit 18, a communication unit 19, and a drive 20.

[0031] The input unit 16 is composed of, for example, a keyboard, mouse, touch panel, or voice input device, and is used to input various types of information. The output unit 17 consists of a display such as an LCD and a speaker, and outputs various information as images and sounds. The memory unit 18 is part of the memory device and consists of DRAM (Dynamic Random Access Memory), etc., and stores various types of data. The communication unit 19 communicates with other devices (for example, the user terminal 2, administrator terminal 3, data provision server 4, web information server 5, and GIS data server 6 in Figure 2) via a network N including the Internet.

[0032] The drive 20 is appropriately fitted with removable media 21, such as a magnetic disk, optical disk, magneto-optical disk, or semiconductor memory. Programs read from the removable media 21 by the drive 20 are installed in the storage unit 18 as needed. Furthermore, the removable media 21 can store various types of data stored in the storage unit 18, just as the storage unit 18 does.

[0033] Although not shown in the diagram, the user terminal 2, administrator terminal 3, data provision server 4, web information server 5, and GIS data server 6 in Figure 2 can also have a configuration that is basically the same as the hardware configuration shown in Figure 3. Therefore, the explanation of the hardware configuration of user terminal 2, administrator terminal 3, data provision server 4, web information server 5, and GIS data server 6 will be omitted.

[0034] Through the cooperation of various hardware and software components that make up the information processing system in Figure 2, including Server 1 in Figure 3, various processes for providing the service in Figure 1 can be executed.

[0035] Figure 4 is a functional block diagram showing an example of the functional configuration of the server in Figure 3 within the information processing system shown in Figure 2.

[0036] As shown in Figure 4, the CPU 11 of server 1 functions as follows: building data acquisition unit 51, nameplate information acquisition unit 52, business type estimation unit 53, classification criterion setting unit 54, usage classification determination unit 55, confidence level calculation unit 56, map display control unit 57, and memory control unit 58. Furthermore, one area of ​​the storage unit 18, which is a storage device of server 1, is provided with building data DB71, nameplate information DB72, business type information DB73, classification criteria DB74, usage classification DB75, learning data DB76, and historical data DB77.

[0037] In step S1 of Figure 1, the building data acquisition unit 51 acquires building data representing the shape and location information of buildings included in the topographic map data from the GIS data server 6, and stores the acquired building data in the building data DB 71. The building data consists of a sequence of coordinate points of polygon figures representing the building shape, and is assumed to be GIS data. The building data acquisition unit 51 generally uses digitized topographic map data (such as 1 / 2500 urban planning base maps or 1 / 1000 house maps) issued by public institutions such as local governments. Alternatively, building data included in residential map data may be used as is. Building data requires location coordinates such as latitude, longitude, and public coordinates, but relative coordinates within the target area are also acceptable.

[0038] In step S1 of Figure 1, the nameplate information acquisition unit 52 acquires nameplate information, including location information, and address information from the data provision server 4, and stores the acquired nameplate information and address information in the nameplate information DB 72. The nameplate information consists of location information (coordinates) that represents the location displayed on the residential map, and a string representing the building name (building name) and nameplate name, and also includes address information consisting of all or part of the city / ward / town / village name, town / district, house number, and residential address. The address information does not require any specific notation method as long as it specifies a region, such as a common name that is not the official address (e.g., XX district or ▲▲ station area). Next, from the residential map data, nameplate information containing location information (coordinates) within the polygon shapes representing the building shapes in the building data is extracted and linked, then added as attribute information for each building shape in the building data and stored in the building data DB71. Alternatively, a linking key that allows linking each building shape record in the building data DB71 with the nameplate information record in the nameplate information DB72 may be assigned to the records in each DB for management. These processes can be carried out using GIS functions.

[0039] In step S2 of Figure 1, the business type estimation unit 53 obtains nameplate information and address information from the attribute information of the building data DB 71 or from the nameplate information DB 72, collects information from external information sources, and estimates the business type of the building from the search results. Examples of information gathering from external sources include performing web searches via the Web information server 5. Specifically, the business type estimation unit 53 uses the nameplate and address information as search keywords to collect information from the store's website, social media, real estate information, etc., and uses a generating AI (e.g., RAG) to estimate the business type from this information. The business type estimation unit 53 generates a search query based on the nameplate and adds address information to improve accuracy before performing the search. The business type estimation unit 53 instructs the generating AI to extract information and determines the business type (type of store, business hours, ratings, presence or absence of queues, characteristics) based on the web information. The business type estimation unit 53 stores the estimated business type information as attribute information for each building shape in the building data DB 71. Alternatively, the estimated business type information may be stored in the business type information DB 73, and a linking key that can link the building shape records in the building data DB 71 with the business type information records in the business type information DB 73 may be assigned to the records in each DB for management.

[0040] The classification criteria setting unit 54 receives the classification items and classification criteria from the user terminal 2 and stores them in the classification criteria DB 74. The classification criteria setting unit 54 allows users to set their own classification items and criteria, in addition to the standard classification items and criteria for urban planning basic surveys, for use in real estate market surveys, trade area analysis, and other purposes. Specifically, for example, in the case of a basic urban planning survey, the classification criteria setting unit 54 adopts the standard classification items of the basic urban planning survey, such as "office facilities," "commercial facilities," "accommodation facilities," "commercial-use mixed facilities," "residential buildings," and "apartment buildings." However, it may also obtain classification items for the basic urban planning survey that are independently determined by each municipality. In addition, the classification criteria setting unit 54 acquires classification criteria corresponding to each classification item. Specifically, this is a conversion table of classification items corresponding to business types, which serves as the basis for assigning the business types estimated by the business type estimation unit 53 to the corresponding classification items. For example, it is a conversion table that allows conversion of "family restaurant" to "restaurant," "ramen shop" to "restaurant," "convenience store" to "retail store," "dentist" to "medical facility," "nursery school" to "educational facility," "general name" to "housing," and "apartment name" to "apartment building." Business types corresponding to classification items are determined in advance and acquired as a conversion table. The classification criteria setting unit 54 may perform data mapping transformation (rule-based) to create a map table (correspondence table (dictionary) from business type to classification item, and create a transformation table that serves as the classification criteria for land use classification so that each business type matches the classification item of the urban planning basic survey. Alternatively, the classification criterion setting unit 54 may use a machine learning model to determine classification criteria that can be converted from nameplate information to usage classification. Specifically, historical data is acquired from nameplate information in past residential maps and information on building use classifications from the current status maps of building use in urban planning basic surveys at the same time. By performing machine learning using this data as training data, a learning model is generated that associates nameplate information with building use classifications. This learning model serves as the classification criterion for determining building use classifications from nameplate information. Using this learning model based on historical data, building use classifications can be determined directly from nameplate information. The learning model generated here may be either a machine learning model or a rule-based inference model. The learning model generated here will be stored in the training data DB76. Furthermore, classification criteria may be established by using generative AI, etc., and assigning appropriate classification items corresponding to each type of business, taking into account guidelines issued by the national government, prefectural governments, etc. Alternatively, classification criteria may be generated (acquired) using generative AI, etc., that allow for the direct assignment of classification items for usage classification from the nameplate information without going through the type of business. In this case, the conversion model from nameplate information to classification items generated within the generative AI becomes the classification criterion, and its structure depends on the estimation method and specifications of the generative AI used, but its format is not restricted. The classification criteria setting unit 54 may also acquire user-defined classification items other than those from the basic urban planning survey. Specifically, for example, classifications for a market area analysis system (e.g., medium classifications, ramen shops, family restaurants, stationery stores, etc.) can be obtained, and the estimated business types (industries [small classifications: ramen shops (pork bone broth), curry shops (Indian), used car dealerships (foreign cars), etc.]) can be grouped together and generated as classification items. If you want to classify only specific types of businesses, you can use classification items such as "ramen shops, others." Based on the obtained classification items, classification criteria that can be converted from business types and sign information can be obtained or formulated and used to classify the use of each building.

[0041] In step S3 of Figure 1, the usage classification determination unit 55 obtains business type information from the building data DB 71 or the business type information DB 73, obtains the applicable classification items and classification criteria from the classification criteria DB 74, and performs the conversion from business type to usage classification. The usage classification determination unit 55 converts the business type estimated by the generating AI into a usage classification (classification item) according to the classification criteria set by the classification criteria setting unit 54. The classification criteria is a conversion table of usage classifications corresponding to each business type. For example, "family restaurant" can be converted to "restaurant," "ramen shop" to "restaurant," "convenience store" to "retail store," "dentist" to "medical facility," "nursery school" to "educational facility," "general name" to "residential," and "apartment name" to "apartment building." The usage classification determination unit 55 performs data mapping transformation (rule-based) and converts each business type into a classification item of the urban planning basic survey based on the transformation criteria of the map table (correspondence table (dictionary) of classification items selected from the business type). Preprocessing to converge inconsistencies in business type notation may be performed using generation AI. Alternatively, the usage classification determination unit 55 may directly determine the usage classification from the building's nameplate information without having to estimate the type of business. The usage classification determination unit 55 obtains nameplate information and address information for each building from the attribute information of the building data DB 71 or from the nameplate information DB 72, and determines the usage classification of each building using classification criteria generated from a machine learning learning model for directly determining the usage classification from the nameplate information, or classification criteria which are conversion models from nameplate information to classification items generated by the generation AI. The usage classification determination unit 55 stores the determined (converted) usage classification as attribute information for each building shape in the building data DB 71. Alternatively, the determined (converted) usage classification may be stored in the usage classification DB 75, and a linking key that can link the building data records in the building data DB 71 and the usage classification records in the usage classification DB 75 may be assigned to the records in each DB for management.

[0042] In step S3 of Figure 1, the confidence calculation unit 56 calculates confidence scores for the business type and usage classification determined by the business type estimation unit 53 and the usage classification determination unit 55, and also generates the estimation basis determined by the generating AI. The confidence calculation unit 56 causes machine learning models, generative AI, etc., to output a confidence score that expresses the likelihood of success. The confidence score is determined, for example, by pre-setting one or more criteria and having a generating AI count the number and degree of relevance of the search results against those criteria. Alternatively, the generating AI may be allowed to set arbitrary criteria and calculate the confidence score, but this is not the only option. Furthermore, the confidence score may be calculated as a score (in the range of 0 to 1) by weighting multiple indicators such as the number of search results, the reliability of the information source, and the degree of keyword match, but is not limited to this. In that case, you may also ask the user to provide the reasoning behind the AI's estimation (e.g., "We chose this business model for reason ●●"). For example, if the confidence score is low, indicating an error, the reason for the low confidence score (error reason) may be output in natural language or similar. For example, it can provide information such as, "We were unable to obtain information on XX Store, but based on the name and other details, it appears to be a general retail store." The output, such as confidence scores and error reasons, will differ depending on the classification method (dictionary, machine learning, or generative AI). The confidence level calculation unit 56 can identify ambiguous areas where accurate mechanical judgment was not possible and, if necessary, prompt a worker to confirm the information via the user terminal 2. By generating a confidence score for the machine-generated judgment, it becomes possible to visualize areas where the machine could not make a judgment (ambiguous areas) along with the reason for the error, and have the worker review them. This can lead to further improvements in accuracy. The confidence calculation unit 56 stores the calculated confidence score along with the building data DB 71 as attribute information for each building shape. Alternatively, if the determined (converted) usage classification is stored in the usage classification DB 75, the calculated confidence score, estimation basis, error reason, etc., may also be stored in the usage classification DB 75.

[0043] In step S4 of Figure 1, the map display control unit 57 obtains the usage classification information for each building from the building data DB 71 or the usage classification DB 75, associates it with the building data in the building data DB 71, and executes control to display the building shapes on the map in different colors. The map display control unit 57 assigns the building use classification result as attribute information to the building data, and together with other map items on the topographic map such as the urban planning base map, it displays the building shape of the building data in a color determined according to the building use classification result, thereby creating a current status map by building use. If there is no building shape data, the map display control unit 57 may simply display symbols such as ○□ at the location information (coordinates) of the nameplate information on the map, colored in a color determined according to the building use classification result. The map display control unit 57 outputs various error locations in an identifiable manner, such as discrepancies where there is building data but no corresponding nameplate information, nameplate information but no corresponding building data, cases where the corresponding use classification could not be determined, and cases with a low confidence score. Based on the results, the map display control unit 57 can create a building use-specific current status map for the urban planning basic survey by coloring the building shapes of the urban planning basic map with colors classified by use on the screen or on paper using GIS. Furthermore, using functions such as GIS, the results of classifying building uses can be aggregated by statistical units such as town / district or 1km grid, based on the number of buildings or total floor area, and then output as statistical information or displayed on a map.

[0044] The memory control unit 58 executes control to save past building data, nameplate information, and classification results of usage classification at the same time to the past data DB 77. The memory control unit 58 can track changes over time and improve processing efficiency. First, the memory control unit 58 can acquire historical residential map data and obtain data for time-series analysis, such as extracting locations where the building's use has changed based on differences in nameplate information at two points in time. As a result, the use classification of buildings whose nameplate information has not changed does not need to be revised; the classification results of the urban planning basic survey at the same point in time are acquired and used as is, and only the use classification of buildings whose nameplate information has changed needs to be revised. This reduces processing time. Furthermore, the memory control unit 58 can perform the same usage classification process as the present based on past nameplate information, thereby extracting buildings whose usage classification differs from those previously classified manually or otherwise, and verifying the accuracy of the past and present building usage classification results. For example, it can also extract errors in past usage classifications.

[0045] Next, please refer to Figure 5 to explain the details of the processing flow of this service. Figure 5 is a flowchart showing an example of the processing flow of this service. This processing flow is executed through the coordinated action of each functional unit shown in Figure 4.

[0046] As shown in Figure 5, Server 1 first acquires building data (Step S5: Processing by Building Data Acquisition Unit 51), and then acquires nameplate information (Step S6: Processing by Nameplate Information Acquisition Unit 52). Specifically, for example, Server 1 acquires building data for each building, which consists of a sequence of coordinate points of a polygon shape representing the building's form, assuming that the data is GIS data (processing by Building Data Acquisition Unit 51). Server 1 obtains nameplate information and address information from residential map data and links them to building data for each building (processing by nameplate information acquisition unit 52). Specifically, for example, Server 1 obtains nameplate information, such as the building name and nameplate name of a building with location information, and associated address information from residential map data. It then links buildings that have the same location information with each other and stores the nameplate information and address information as attribute information for each building (processing by Nameplate Information Acquisition Unit 52).

[0047] Next, Server 1 obtains business type information from an external information source based on the nameplate information of each building (Step S7), and performs the estimation of the business type of each building (Step S8: Processing by Business Type Estimation Unit 53). Server 1 collects unstructured information, including social media data, and organizes it using AI, including generative AI and search AI that utilize RAG technology (processing by business type estimation unit 53). Server 1 then stores the business type information and other data as attribute information for each building (step S9).

[0048] Subsequently, Server 1 uses a generation AI or the like to acquire and generate classification items and classification criteria for usage classification (Step S10: Processing of Classification Criteria Setting Unit 54), sets them as classification items and classification criteria for usage classification, and determines the usage classification of each building based on the set classification items and classification criteria (Step S11: Processing of Usage Classification Determination Unit 55). Server 1 is capable of not only using the classification items of the basic urban planning survey, but also generating its own classifications (classification items from estimated business types) (processing by the use classification determination unit 55).

[0049] Furthermore, Server 1 outputs a confidence score when estimating the type of business and usage classification of each building (Step S12: Processing by Confidence Calculation Unit 56), and saves it as attribute information for each building along with the usage classification result, confidence score, estimation basis, error reason, etc. (Step S13). Finally, Server 1 displays the classification results of the use classification for each building on a map along with the background topographic map (Step S14: Processing by Map Display Control Unit 57). Server 1 displays the classification results of the use classification of each building in color, and can be displayed dynamically on the screen, on paper, as a PDF®, or through the system as appropriate (processing by map display control unit 57).

[0050] Next, referring to Figure 6, we will explain the process of linking building data and nameplate information. Figure 6 shows an example of linking building data and nameplate information. The process shown in Figure 6 is mainly performed by the building data acquisition unit 51 and the nameplate information acquisition unit 52.

[0051] As shown in Figure 6, the building data acquisition unit 51 acquires building data consisting of a sequence of coordinate points of a polygon shape representing the building's shape, assuming that the data is GIS data. The building data requires location information, including latitude, longitude, and public coordinates.

[0052] The nameplate information acquisition unit 52 extracts nameplate information (nameplate name, store name, building name, etc.) and address information from the residential map data. The nameplate information also includes location information for where the nameplate name is displayed on the residential map. The linking of building data and nameplate information is done by searching for nameplate information with coordinates contained within the polygon figures representing the shape of each building in the building data such as topographic maps, and linking it as attribute information for each building in the building data. Before linking the two sets of data, it is necessary to unify the coordinate systems of both sets of data. For example, since building data is generally in public coordinates, the location information of the nameplates should be converted to public coordinates. Furthermore, if there is a slight discrepancy between the coordinates of the building shape in the building data and the coordinates of the nameplate information in the residential map due to errors in accuracy during the preparation of each map, the linking process may be performed by connecting buildings without corresponding nameplate information with the closest nameplate information among those buildings without corresponding nameplate information.

[0053] Next, we will explain the conversion from business type to usage classification, referring to Figure 7. Figure 7 shows an example of a conversion table that illustrates the classification criteria from business type to usage classification. The process shown in Figure 7 is mainly performed by the usage classification determination unit 55.

[0054] As shown in Figure 7, the usage classification determination unit 55 converts the business type of each building estimated by the generation AI of the business type estimation unit 53 into usage classifications (classification items) based on the classification criteria set by the classification criteria setting unit 54. For example, "family restaurants" and "ramen shops" are classified as "restaurants." "Convenience stores" are classified as "retail stores." "Dentists" are classified as "medical facilities." "Nursery schools" are classified as "educational facilities." "General names" are classified as "residences." "Apartment names" are classified as "apartment buildings." By converting various types of businesses into a unified usage classification in this way, it becomes possible to perform statistics and analyses by building use, which are necessary for urban planning basic surveys and market area analyses.

[0055] Next, we will explain how to display the current status diagram by building use, referring to Figure 8. Figure 8 shows an example of a display of a building usage-based current status map. The process for displaying the example shown in Figure 8 is mainly performed by the map display control unit 57.

[0056] As shown in Figure 8, the map display control unit 57 assigns the building use classification result as attribute information to the building data, and displays it on the screen along with the background topographic map, and also performs printing and PDF output. Figure 8 shows the following as representative classification items in the legend: "1. Business facilities," "2. Commercial facilities," "3. Accommodation facilities," "4. Mixed-use commercial facilities," "5. Residential buildings," "6. Apartment buildings," "7. Residential buildings with shops, etc.," and "8. Apartment buildings with shops, etc.." However, in actual classification, in addition to these, classifications such as "9. Residential buildings with workshops," "10. Government facilities," "11. Educational and welfare facilities," "12. Transportation and warehousing facilities," "13. Factories," "14. Agricultural, forestry, and fishery facilities," "15. Supply and processing facilities," "16. Defense facilities," "17. Others," "18. Unknown," and "19. Vacant houses" are color-coded.

[0057] The map display control unit 57 displays building shapes in color and classification along with other map items on topographic maps such as urban planning base maps. As shown in the lower right frame of Figure 8, buildings with low confidence scores can be highlighted with a red frame or the like to encourage workers to verify them. Workers can refer to the building's nameplate information, confidence score, and error reasons using GIS, and can correct the classification results of the usage classification determined on the spot as needed.

[0058] Next, referring to Figure 9, we will explain the automatic recognition of printed residential maps distributed in booklets, etc. Figure 9 shows an example of automatic recognition of residential maps. The process shown in Figure 9 is mainly performed by the nameplate information acquisition unit 52.

[0059] As shown in Figure 9, the nameplate information acquisition unit 52 acquires image data of printed residential maps distributed in booklets, etc., by scanning them, instead of residential map data, and automatically recognizes strings corresponding to nameplate information from the residential map image where building names, nameplate names, etc., are written inside the houses. These characters are recognized as nameplate information for each building, and the purpose of each building is automatically determined, similar to the digital data of the residential map. For example, "○○ School" is identified as an educational facility, "△△ Building" as an office, etc., "×× Shop" as a commercial facility, "Restaurant ※※" as a food and beverage establishment, and "Yamada Taro" as a residence. Similarly, the building shape is also automatically recognized from the residential map image and converted into vector data, so that the building shape, nameplate information, and classification results of the usage classification can be converted into vector data. When conducting usage surveys on a drawing-by-drawing basis without coordinate information, or when coordinates are assigned to each drawing, processing similar to that of a digitized residential map is possible.

[0060] Next, referring to Figure 10, we will explain how to utilize the results of past urban planning basic surveys. Figure 10 shows examples of how the results of the urban planning basic survey conducted in previous years have been utilized. The processing shown in Figure 10 is mainly performed using the memory control unit 58 and the learning data DB 76.

[0061] As shown in Figure 10, the basic urban planning survey includes the creation of a current status map by building use, which is conducted once every five years. While the basic procedure involves contractors hired by municipalities to visit sites and survey building uses, to improve efficiency, sometimes building uses are identified from nameplate information on residential maps, and buildings on the map are color-coded according to their use. As a result, every five years, a combination of nameplate information from residential maps and building use classification information is created for each building. Therefore, one way to improve the accuracy of the building use classification determined by the business type estimation unit 53, classification criterion setting unit 54, building use classification determination unit 55, and confidence calculation unit 56 is to utilize the results of basic surveys already conducted in previous years.

[0062] There are two main ways to utilize the results of surveys conducted in previous years: one is to extract changes in land use over time, specifically identifying locations where the land use has changed; the other is to perform fully automated recognition, automatically recognizing everything from residential maps. First, the current nameplate information is compared with past information, and only those with changes are extracted and classified. If there are no changes, the past classification results can be used as is. Alternatively, it is possible to obtain nameplate information from past residential maps and information from the current building use maps of the urban planning basic survey at the same time as historical data, and use these as training data to create a learning model that determines land use classification from nameplate information, which serves as the classification criterion for land use classification. By utilizing this learning model based on historical data, it becomes possible to determine land use classification with higher accuracy by inputting the latest nameplate information.

[0063] Although one embodiment of the present invention has been described above, the present invention is not limited to the embodiments described above, and any modifications, improvements, etc. that can achieve the objectives of the present invention are considered to be included in the present invention.

[0064] Furthermore, the system configuration shown in Figure 2 and the hardware configuration of Server 1 shown in Figure 3 are merely illustrative examples for achieving the objectives of the present invention and are not particularly limited.

[0065] Furthermore, the functional block diagram shown in Figure 4 is merely illustrative and not particularly limiting. In other words, it is sufficient that the information processing system in Figure 2 has the functionality to execute the various processes described above as a whole, and the functional blocks and databases used to realize this functionality are not particularly limited to the example in Figure 4.

[0066] Furthermore, the location of the functional blocks and database is not limited to Figure 4 and can be arbitrary. For example, at least a portion of the functional blocks and database located on Server 1 may be provided on User Terminal 2, Administrator Terminal 3, or other information processing devices not shown.

[0067] Furthermore, the series of processes described above can be executed by hardware or by software. Furthermore, a single functional block may consist of hardware alone, software alone, or a combination of both.

[0068] When a series of processes are executed by software, the programs that make up that software are installed on a computer or other device from a network or storage medium. The computer may be a computer that is built into dedicated hardware. Furthermore, a computer can be any computer capable of performing various functions by installing various programs, such as a server, or it could be a general-purpose smartphone or personal computer.

[0069] Such recording media containing programs may consist not only of removable media (not shown) distributed separately from the main unit of the device to provide the program to the user, but also of recording media provided to the user in a state where they are pre-installed in the main unit of the device.

[0070] In this specification, the step of describing a program to be recorded on a recording medium includes not only processes that are performed chronologically in that order, but also processes that are not necessarily performed chronologically, but are executed in parallel or individually.

[0071] In summary, the information processing device to which the present invention applies only needs to have the following configuration, and can take various forms. That is, the information processing device to which the present invention is applied (for example, Server 1 in Figures 2 to 4) is: A building data acquisition means (for example, the building data acquisition unit 51 in Figure 4) that acquires building data including building location information, A means for acquiring nameplate information of the aforementioned building and associating it with the building data (for example, the nameplate information acquisition unit 52 in Figure 4), A classification criteria setting means (for example, the classification criteria setting unit 54 in Figure 4) sets classification criteria for determining the usage classification from the nameplate information of the aforementioned building, Based on the classification criteria set by the classification criteria setting means, a usage classification determination means (for example, the usage classification determination unit 55 in Figure 4) determines the usage classification of the building from the nameplate information acquired by the nameplate information acquisition means, A map display control means (for example, the map display control unit 57 in Figure 4) that performs control to display the usage classification determined by the usage classification determination means on the map in association with the building data, Having that will suffice.

[0072] In this way, compared to the conventional manual method, it becomes possible to improve operational efficiency, speed up operations, reduce costs, and also improve the accuracy of survey results.

[0073] Furthermore, this allows for the visual understanding of the distribution of building uses by color-coding and classifying the shape of each building in the building data, in conjunction with other map items on topographic maps such as urban planning base maps.

[0074] Furthermore, the system includes a means for estimating the type of business of a building by collecting information from external sources based on the sign information of the building (for example, searching for store website / SNS information and real estate information in step S2 of Figure 1, and estimating the type of business in step S8 of Figure 5), The classification criteria setting means sets classification criteria for determining the use classification from the type of business of the building estimated by the business type estimation means based on the sign information, The aforementioned usage classification determination means can determine the usage classification from the business type estimated by the business type estimation means based on the classification criteria.

[0075] This allows for the determination of the actual type of business based on web searches using nameplate information, as well as external information such as customer reviews on the store's website and social media, and real estate transaction-related websites. This enables the accurate determination of the building's intended use.

[0076] Furthermore, the nameplate information acquisition means further acquires the address information of the building, The aforementioned usage classification determination means can determine the usage classification of the building based on the nameplate information and the address information (for example, the linking process in step S1 of Figure 1 where address information is added to the nameplate information, and the acquisition of nameplate information in step S6 of Figure 5).

[0077] This allows for the distinction and determination of the building use classification corresponding to each address, even if there are different buildings with the same nameplate information, enabling more accurate determination of building use classification.

[0078] Furthermore, the classification criterion setting means accepts the setting of classification criteria from the user (for example, the classification criterion setting unit 54 in Figure 4, the acquisition and generation of classification items and classification criteria in step S10 in Figure 5, and the classification criteria in Figure 7), The aforementioned usage classification determination means can determine the usage classification from the nameplate information based on the classification criteria set by the classification criteria setting means (for example, application of user-defined classification criteria in step S3 of Figure 1, and classification based on classification items and classification criteria in step S11 of Figure 5).

[0079] This makes it possible to repurpose the data not only for land use classification in basic urban planning surveys, but also for surveying the distribution of apartment buildings in real estate market research, and for conducting market area analysis surveys when opening restaurants, shops, and other businesses.

[0080] Furthermore, the system may further include a confidence calculation means for calculating the confidence level of the application classification determined by the application classification determination means (for example, the confidence calculation unit 56 in Figure 4, and the output of the confidence score in step S12 in Figure 5).

[0081] This allows for the system to visualize areas where mechanical judgment was not possible (ambiguous areas) for human verification, further improving accuracy. Furthermore, by providing information on the basis for business type estimation and usage classification, as well as the reasons for errors with low confidence scores, it also leads to more efficient verification and updating of classification results.

[0082] Furthermore, the business type estimation means can estimate the business type from the sign information using a generating AI (for example, business type estimation by a generating AI utilizing RAG technology in step S2 of Figure 1, and business type estimation by a generating AI in step S8 of Figure 5).

[0083] This allows for cost reduction and increased development speed compared to conventional software development by utilizing generative AI, while also improving the accuracy of research results.

[0084] Furthermore, the system further includes a storage control means (for example, a storage control unit 58 in Figure 4) that executes control to store nameplate information at a predetermined point in the past in a predetermined storage device (for example, the past data DB77 in Figure 4), The nameplate information acquisition means can compare the current nameplate information with the nameplate information at a predetermined point in the past, and perform the processing by the usage classification determination means only for buildings where changes have occurred (extracting only the locations where the usage has changed in Figure 10).

[0085] This reduces processing time and eliminates the need to classify every building.

[0086] Furthermore, the classification criterion setting means sets classification criteria for usage classification using a learning model (for example, stored in the learning data DB76 in Figure 4) for estimating usage classification from the nameplate information. The aforementioned learning model can generate training data from past nameplate information and corresponding building use classification results (using the residential map data for the same point in time shown in Figure 10 is sufficient as training data).

[0087] This allows for the accurate reproduction of classification criteria by using historical information (residential maps and results from basic urban planning surveys), and also improves the accuracy of the classification results for land use based on current nameplate information.

[0088] Furthermore, the system includes building data acquisition means (for example, the building data acquisition unit 51 in Figure 4, and the acquisition of building data in step S5 in Figure 5) that acquire building data including building shape and location information. The nameplate information acquisition means can acquire the corresponding nameplate information based on the location information of the building data (for example, linking with location information in step S1 of Figure 1, and acquiring nameplate information in step S6 of Figure 5).

[0089] In this way, compared to the conventional manual method, it becomes possible to improve operational efficiency, speed up operations, reduce costs, and also improve the accuracy of survey results. [Explanation of Symbols]

[0090] 1...Server, 2...User terminal, 3...Administrator terminal, 4...Data provision server, 5...Web information server, 6...GIS data server, N...Network, U...User, M...Administrator, 11...CPU, 12...ROM, 13...RAM, 14...Bus, 15...Input / Output interface, 16...Input unit, 17...Output unit, 18...Storage unit, 19...Communication unit, 20...Drive, 21...Removable media, 51...Building data acquisition unit, 52...Nameplate information acquisition unit, 53...Business type estimation unit, 54...Classification criterion setting unit, 55...Usage classification determination unit, 56...Confidence level calculation unit, 57...Map display control unit, 58...Storage control unit, 71...Building data DB, 72...Nameplate information DB, 73...Business type information DB, 74...Classification criteria DB, 75...Usage classification DB, 76...Learning data DB, 77...Past data DB, S1~S14...Steps

Claims

1. A means for acquiring building data, including building location information, A means for acquiring nameplate information of the aforementioned building and associating it with the aforementioned building data, A classification criteria setting means for setting classification criteria for determining the use classification from the nameplate information of the said building, A means for determining the use of a building, which collects external information from an external information source based on the nameplate information of the building, and determines the use classification of the building based on the nameplate information and the external information, A map display control means that performs control to display the usage classification determined by the usage classification determination means on a map in association with the building data, Equipped with, The aforementioned usage classification determination means determines the usage classification from the nameplate information using generating AI. Information processing device.

2. A means for acquiring building data, including building location information, A means for acquiring nameplate information of the aforementioned building and associating it with the aforementioned building data, A business type estimation means that collects information from an external information source based on the nameplate information of the said building and estimates the type of business of the said building, A classification criteria setting means for setting classification criteria for determining the use classification of the aforementioned building from the aforementioned business type, A means for determining the use classification of a building based on the business type estimated by the business type estimation means, based on the classification criteria, A map display control means that performs control to display the usage classification determined by the usage classification determination means on a map in association with the building data, An information processing device equipped with the following features.

3. A means for acquiring building data, including building location information, A means for acquiring nameplate information of the aforementioned building and associating it with the aforementioned building data, A classification criteria setting means for setting classification criteria for determining the use classification from the nameplate information of the aforementioned building, A means for determining the use classification of a building from the nameplate information based on the aforementioned classification criteria, A map display control means that performs control to display the usage classification determined by the usage classification determination means on a map in association with the building data, A storage control means that performs control to store the nameplate information at a predetermined point in the past in a predetermined storage device, Equipped with, The nameplate information acquisition means compares the current nameplate information with the nameplate information at a predetermined point in the past, The processing by the aforementioned use classification determination means is performed only for the buildings that have undergone changes. Information processing device.

4. The aforementioned nameplate information acquisition means further acquires the address information of the building, The aforementioned usage classification determination means determines the usage classification of the building based on the nameplate information and the address information. The information processing apparatus according to claim 1 or 2.

5. The system further comprises a confidence calculation means for calculating the confidence level of the use classification determined by the use classification determination means. The information processing apparatus according to any one of claims 1 to 3.

6. The aforementioned usage classification determination means determines the usage classification from the nameplate information using generating AI. The information processing apparatus according to claim 3.

7. The aforementioned business type estimation means estimates the business type from the sign information using generating AI. The information processing apparatus according to claim 2.

8. An information processing method performed by an information processing device, A building data acquisition step that acquires building data including building location information, A step of acquiring nameplate information for the aforementioned building and associating it with the aforementioned building data, A classification criteria setting step for setting classification criteria to determine the usage classification from the nameplate information of the aforementioned building, A usage classification determination step involves collecting external information from external sources based on the nameplate information of the said building, and determining the usage classification of the said building based on the classification criteria, using the nameplate information and the external information. A map display control step that performs control to display the usage classification determined in the usage classification determination step in relation to the building data on a map, Includes, The aforementioned step of determining the usage classification includes a step of determining the usage classification from the nameplate information using generating AI. Information processing methods.

9. On the computer, A building data acquisition step that acquires building data including building location information, A step of acquiring nameplate information for the aforementioned building and associating it with the aforementioned building data, A classification criteria setting step for setting classification criteria to determine the usage classification from the nameplate information of the aforementioned building, A usage classification determination step involves collecting external information from external sources based on the nameplate information of the said building, and determining the usage classification of the said building based on the classification criteria, using the nameplate information and the external information. A map display control step that performs control to display the usage classification determined in the usage classification determination step in relation to the building data on a map, The control process including this is executed, The aforementioned step of determining the usage classification involves using a generating AI to determine the usage classification from the nameplate information. To execute control processing that includes, program.

Images