Information processing device, information processing method, and program

Abstract

The goal is to provide technology for analyzing real estate using generative models. [Solution] One aspect of the present disclosure relates to an information processing device comprising: an acquisition unit for acquiring real estate-related information of real estate; a prompt creation unit for creating a prompt that causes a generation model to generate utilization plans for the real estate based on the real estate-related information; and an interface unit for transmitting the prompt to the generation model and acquiring the utilization plans from the generation model.

Description

Technical Field

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

Background Art

[0002] The use of information and communication technologies has spread to various industrial fields. For example, the introduction of information and communication technologies has also spread to the real estate industry, and such real estate-related information and communication technologies are called proptech. Currently, the use of information and communication technologies is also progressing in the mediation of real estate properties and remote property viewings.

[0003] With the recent evolution of machine learning technologies, a search technology for real estate properties using a machine learning model has been proposed.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] With the emergence of ChatGPT (registered trademark), machine learning models such as generative AI (Artificial Intelligence) and large language models (LLMs) (hereinafter collectively referred to as generative models) have been used in various business fields.

[0006] An object of the present disclosure is to provide a technique for analyzing real estate using a generative model.

Means for Solving the Problems

[0007] One aspect of this disclosure relates to an information processing device comprising: an acquisition unit for acquiring real estate-related information of real estate; a prompt creation unit for creating a prompt that causes a generation model to generate a plan for the use of the real estate based on the real estate-related information; and an interface unit for transmitting the prompt to the generation model and acquiring the plan for the use from the generation model. [Effects of the Invention]

[0008] According to this disclosure, we can provide technology for analyzing real estate using generative models. [Brief explanation of the drawing]

[0009] [Figure 1] Figure 1 is a schematic diagram showing the process of generating a real estate utilization plan according to one embodiment of the present disclosure. [Figure 2] Figure 2 is a block diagram showing the hardware configuration of an information processing device according to one embodiment of the present disclosure. [Figure 3] Figure 3 is a block diagram showing the functional configuration of an information processing device according to one embodiment of the present disclosure. [Figure 4] Figure 4 is an illustrative diagram showing real estate registration data according to one embodiment of the present disclosure. [Figure 5] Figure 5 shows an example of prompts for generating a real estate utilization plan according to one embodiment of the present disclosure. [Figure 6] Figure 6 shows an example of prompts for generating a real estate utilization plan according to one embodiment of the present disclosure. [Figure 7] Figure 7 shows a real estate utilization plan generated in response to an example prompt according to one embodiment of the present disclosure. [Figure 8] Figure 8 shows a real estate utilization plan generated in response to an example prompt according to one embodiment of the present disclosure. [Figure 9] Figure 9 shows a real estate utilization plan generated in response to an example prompt according to one embodiment of the present disclosure. [Figure 10]Figure 10 is a flowchart showing the process of generating a real estate utilization plan according to one embodiment of the present disclosure. [Modes for carrying out the invention]

[0010] Embodiments of this disclosure will be described below with reference to the drawings.

[0011] In the following embodiment, an information processing device for generating real estate utilization plans for real estate is disclosed.

[0012] [overview] According to one embodiment of the present disclosure, the information processing device 100 acquires real estate-related information, such as real estate registration data, of a property to be analyzed, and creates a prompt (question or instruction) to cause a generation model to generate a proposed use for the property based on the acquired real estate-related information. The information processing device 100 then sends the created prompt to a generation model such as ChatGPT, and obtains a proposed use for the property from the generation model.

[0013] Specifically, as shown in Figure 1, when a user inputs real estate registration data of the property to be analyzed (e.g., registration certificate, land value, land use designation, urban plan, etc.) into the information processing device 100, the information processing device 100 creates a prompt to cause the generation model 50 to generate utilization plans that suggest how the property to be analyzed can be utilized, based on the received real estate registration data, as will be explained in more detail below.

[0014] The users here may typically be staff members of real estate companies, financial institutions, construction companies, etc., who are authorized to use the information processing device 100, and they may use the real estate utilization plan generation process described later not only to analyze their clients' real estate but also to analyze the real estate of third parties.

[0015] Note that the user does not need to obtain the real estate registration data of the real estate to be analyzed by themselves. For example, when the user designates the real estate to be analyzed on the map information displayed on the display, the information processing device 100 may obtain the real estate registration data of the designated real estate from a database or the like.

[0016] Then, the information processing device 100 inputs the created prompt into the generation model 50 operating on a cloud server (not shown), and obtains a utilization plan for the real estate from the generation model 50. Then, the information processing device 100 provides the obtained utilization plan to the user. At this time, the information processing device 100 may provide the text data of the utilization plan obtained from the generation model 50 to the user as it is, or execute any appropriate interface processing for enhancing the user interface, and present the processing result on a user-friendly interface screen, etc.

[0017] According to such an information processing device 100, the user can easily and quickly obtain a utilization plan for the desired real estate. Also, it is possible to create a prompt to generate a plurality of utilization plans, or create a prompt by referring to unique reference data (for example, real estate transaction information, real estate price information, parking lot information, etc.) based on RAG (Retrieval Augmented Generation), and cause the generation model 50 to generate a more useful real estate utilization plan for the user.

[0018] Here, the information processing device 100 may have a hardware configuration as shown in FIG. 2, for example. That is, the information processing device 100 has a storage device 101, a memory device 102, a processor 103, and a communication device 104 that are interconnected via a bus B.

[0019] Programs and / or data for realizing various functions and processes in the information processing device 100 are downloaded from an external device and / or network to the storage device 101 or the memory device 102 via the communication device 104.

[0020] The storage device 101 is implemented using non-volatile memory or the like, and stores installed or downloaded programs or data (e.g., files).

[0021] The memory device 102 is implemented using random access memory, static memory, etc., and when a program or instruction is activated, it reads the program or instruction, data, etc. from the storage device 101 and stores it. The storage device 101, the memory device 102, and the removable storage medium may be collectively referred to as a non-transitory storage medium.

[0022] The processor 103 may be implemented by one or more CPUs (Central Processing Units), GPUs (Graphics Processing Units), processing circuits, etc., which may consist of one or more processor cores, and executes various functions and processes of the information processing device 100 according to data such as programs, instructions, and parameters necessary to execute said programs or instructions stored in the memory device 102.

[0023] The communication device 104 is implemented by various communication circuits that perform wired and / or wireless communication processing with external devices, the Internet, LAN (Local Area Network), cellular networks, and other communication networks.

[0024] However, the hardware configuration described above is merely an example, and the information processing device 100 according to this disclosure may be implemented by any other suitable hardware configuration.

[0025] [Information Processing Device] Next, an information processing device 100 according to one embodiment of the present disclosure will be described. Figure 3 is a block diagram showing the functional configuration of the information processing device 100 according to one embodiment of the present disclosure. As shown in Figure 3, the information processing device 100 has an acquisition unit 110, a prompt creation unit 120, and an interface unit 130. Each of the functional units of the acquisition unit 110, the prompt creation unit 120, and the interface unit 130 can be realized by the processor 103 of the information processing device 100 executing a program stored in the memory device 102.

[0026] The acquisition unit 110 acquires real estate-related information of the property. For example, real estate-related information is information related to the property being analyzed, and is typically real estate registration data (e.g., a certificate of registered matters) that identifies the property being analyzed. The acquisition unit 110 provides the acquired real estate-related information to the prompt creation unit 120.

[0027] For example, real estate registration data may be document data with a format as shown in Figure 4. As shown in the figure, the “Title Section (Description of Land)” includes the land number indicating the location of the land, the land category indicating its use, the land area indicating its area, the reason for registration, and the date thereof.

[0028] Furthermore, the "Rights Section (Section A)" contains information related to ownership, including the priority number, purpose of registration, date and number of registration, rights holder, and other details.

[0029] Furthermore, the "Rights Section (Section B)" contains information about rights other than ownership, including the priority number, purpose of registration, date and number of registration, rights holder, and other details.

[0030] Furthermore, the “Joint Collateral Register” will list other land and other properties that are jointly collateralized. The matters to be recorded in the “Title Section (Land Description),” “Rights Section (Section A),” “Rights Section (Section B),” and “Joint Collateral Register” are the same as the matters to be recorded in the Certificate of Registered Matters.

[0031] Furthermore, real estate-related information may include land value assessments, land use designations, and urban planning information for the area where the property is located. It may also include real estate usage information indicating the current or past usage history of the property.

[0032] Furthermore, in the embodiments described above, real estate-related information was provided by the user of the information processing device 100, but the acquisition of real estate-related information according to this disclosure is not necessarily limited to this. For example, if the user specifies a property to be analyzed on map information displayed on the display, the acquisition unit 110 may extract real estate-related information for the property specified by the user from a database or the like, and provide the extracted real estate-related information to the prompt creation unit 110. This allows the user to provide the prompt creation unit 120 with the real estate-related information necessary to create a prompt simply by specifying a property to be analyzed on the map information, without having to collect real estate-related information such as real estate registration data themselves.

[0033] In this case, the acquisition unit 110 may display real estate-related information collected from a database or the like to the user and allow the user to select the real estate-related information to be used to create a prompt. This makes it possible to exclude information from the collected real estate-related information that is not necessary to create a prompt (for example, outdated information, incorrect information, etc.) in advance.

[0034] The prompt creation unit 120 creates prompts that cause the generation model 50 to generate utilization plans for real estate based on real estate-related information. Specifically, after acquiring real estate-related information such as real estate registration data that identifies the real estate to be analyzed, land value for the area in which the real estate is located, land use designation, urban planning, and real estate usage information that shows the current usage status or past usage history of the real estate, the prompt creation unit 120 uses the acquired real estate-related information to create prompts that cause the generation model 50 to generate real estate utilization plans that propose how the real estate can be utilized.

[0035] Specifically, the prompt generation unit 120 may create prompts as shown in Figure 5. In the illustrated example prompt, the prompt generation unit 120 specifies the role of the generation model 50 regarding the perspective from which it should generate the response by instructing it with the statement, "This GPT functions as an expert for financial institutions." It is known that defining a role for the generation model 50 in this way improves the accuracy of the response.

[0036] Next, the prompt creation unit 120 attaches real estate-related information such as real estate registration data, land value, land use designation, urban planning, and real estate usage information to the prompt in order to have the generation model 50 consider it. As shown in the figure, it also writes an instruction to the prompt for the attached real estate-related information, stating, "First, receive the PDF file of the real estate registration and understand its contents. Then, investigate the surrounding area and check how real estate prices have fluctuated in the past. Based on that, make proposals to the real estate owner regarding asset utilization, taxation, and business succession," and has the generation model 50 consider the real estate-related information.

[0037] Next, the prompt generation unit 120 instructs the generation model 50 on specific proposals for the utilization plan by writing the following instruction in the prompt: "The following is a list of proposals that I, as an asset management expert, can offer to this property owner: • Offer asset management proposals • Develop a tax plan • Present strategies to maximize the value of the property."

[0038] Next, the prompt generation unit 120 may write the following instruction in the prompt to specify the output content of the utilization proposal from the generation model 50: "Create specific proposals based on the following requirements: ・Create 10 specific proposals that can be provided regarding asset management and tax planning. ・Provide specific forecasts of expected returns for each project. ・For each project, give specific examples for that region. ・Specifically list strategies to be avoided in this area (avoid abstract expressions). ・Explain why those forecasts were made. The proposals should be written in Japanese and in an easy-to-understand format."

[0039] The instructions for incorporating real estate-related information into the aforementioned generation model 50, the instructions for considering specific proposals in the utilization plan, and the instructions regarding the output content of the utilization plan are not necessarily limited to these.

[0040] For example, prompts corresponding to the types of real estate-related information may be provided, and the prompt creation unit 120 may use prompts corresponding to the types of real estate-related information acquired.

[0041] Furthermore, prompts corresponding to specific considerations regarding real estate are provided, and the prompt creation unit 120 may use prompts corresponding to specific considerations regarding real estate specified by the user.

[0042] Furthermore, prompts corresponding to the output format of the usage proposal are provided, and the prompt creation unit 120 may use prompts corresponding to the output format specified by the user.

[0043] Furthermore, the prompt generation unit 120 may also input an English prompt corresponding to the above-mentioned instruction sentence to the generation model 50, as shown in Figure 6. This is because, for many generation models 50 such as ChatGPT, the response to an English prompt is often superior to the response to a Japanese prompt, even if the prompts are identical.

[0044] Furthermore, the prompt creation unit 120 may create prompts using RAG. Specifically, the information processing device 100 can connect to a database that stores reference data for RAG, and the prompt creation unit 120 may extract reference data for areas where real estate exists from the database and create prompts using the extracted reference data.

[0045] Generative models such as ChatGPT are trained using information and data publicly available on the internet. Generally, information and data publicly available on the internet are not guaranteed to be accurate and may contain errors. This is a well-known problem that can lead to hallucination. To mitigate hallucination, a technique called RAG (Research Aggregation) is gaining attention, which improves the accuracy of responses from generative models by having them refer to highly accurate information and data.

[0046] For example, the reference data stored in the database may include real estate price information, parking information, railway line information, hazard maps, and real estate transaction information for each area. This information is typically verified for accuracy by the operator of the information processing device 100, and may include information that is publicly available on the internet or information that can only be viewed by specific users. The prompt creation unit 120 may use such reference data to create prompts.

[0047] Furthermore, the prompt generation unit 120 may determine the regularity of the real estate or its connectivity to public roads based on real estate-related information and generate a prompt based on the determination result. For example, real estate registration data may include information on the shape of the land and the public roads adjacent to the land, and this information is generally considered important in land utilization.

[0048] Here, the degree of regularity of the property is, for example, the perimeter of the land [m] / the area of ​​the land [m]. 2The calculation may be performed according to the formula described above. The higher the ratio of the perimeter to the area, the lower the degree of regularity is judged to be, and the lower the ratio of the perimeter to the area, the higher the degree of regularity is judged to be. However, the degree of regularity of land is not necessarily limited to the calculation formula described above, and may be judged by other ratios of perimeter to area, or by other calculation formulas using perimeter and area.

[0049] When the degree of regularity and / or public road connectivity of the land to be analyzed is determined from the real estate registration data, the prompt creation unit 120 may describe the determined degree of regularity and / or public road connectivity in the prompt, causing the generation model 50 to take the degree of regularity and / or public road connectivity into consideration.

[0050] Furthermore, the prompt generation unit 120 may create prompts based on real estate usage information that indicates the current usage status or past usage history of the real estate. The current usage status and past usage history of the real estate being analyzed can be important factors when considering how to utilize the real estate. Therefore, if real estate usage information is obtained from the user, the prompt generation unit 120 may describe in the prompt how the real estate is currently being used and / or how it has been used in the past, based on the obtained real estate usage information, and cause the generation model 50 to take the real estate usage information into consideration.

[0051] Furthermore, the prompt generation unit 120 may create prompts for owners who own multiple properties, prompting them to generate utilization plans for those properties based on a portfolio strategy. Portfolio strategies may include various investment strategies, such as a diversified approach that avoids concentrating on a specific utilization method, or a rebalancing approach that changes utilization methods according to the market value of the multiple properties owned. Once a portfolio strategy is selected by the user, the prompt generation unit 120 may write prompts to prompt the user to consider utilization plans for the multiple properties according to the selected portfolio strategy.

[0052] Furthermore, the prompt generation unit 120 may determine whether the property owner has borrowed money from a financial institution based on real estate-related information, and determine whether the property is for the owner's residence based on the determination result. For example, the prompt generation unit 120 may determine whether a mortgage on a loan from a financial institution is set in the real estate registration data, and if a mortgage is set, it may determine that there is a high possibility that the property owner is using the property for residential purposes rather than investment purposes. On the other hand, if no mortgage is set, it may determine that there is a possibility that the property owner is using the property for investment purposes.

[0053] If the property is being used for residential purposes, the prompt generation unit 120 may include in the prompt that the property being analyzed is currently being used for residential purposes, causing the generation model 50 to take into consideration that the property is for residential use.

[0054] On the other hand, if the property is being used for investment purposes, the prompt generation unit 120 may describe in the prompt that the property being analyzed is currently being used for investment purposes, causing the generation model 50 to take this fact into consideration.

[0055] Furthermore, the prompt generation unit 120 may create prompts to add the owner's attributes to the utilization plan based on the determination result of whether the property is for residential or investment use. For example, if the property is for investment use, the prompt generation unit 120 may write prompts to include the owner's nationality, etc., in the utilization plan.

[0056] Once a prompt is created in this manner, the prompt creation unit 120 provides the created prompt to the interface unit 130.

[0057] The interface unit 130 sends a prompt to the generation model 50 and obtains application suggestions from the generation model 50. Specifically, when the interface unit 130 obtains a prompt from the prompt generation unit 120, it sends the obtained prompt to a cloud server or the like that operates the generation model 50, and obtains the response generated by the generation model 50 in response to the prompt from the cloud server.

[0058] For example, when the interface unit 130 sends a prompt to the cloud server as shown in Figure 5 or Figure 6, it receives from the cloud server, as a result of generation from the generation model 50 in response to that prompt, the application suggestions shown in Figures 7 to 9.

[0059] The illustrated utilization plan presents 10 utilization options (Proposal 1 to Proposal 10) for the property being analyzed, as instructed by the prompt. For each proposal, the expected return and examples of nearby properties are provided along with the proposal content.

[0060] The interface unit 130 may forward the response received from the cloud server directly to the user, or it may format the received response into a format that is easy for the user to read and display it.

[0061] According to the information processing device 100 described above, by using the generation model 50, users who are not familiar with real estate utilization can easily obtain utilization proposals for the real estate being analyzed, thereby improving the efficiency of operations in the real estate business.

[0062] [Generation process] Next, the process for generating a real estate utilization plan according to one embodiment of the present disclosure will be described. This generation process can be realized by the information processing device 100, or more specifically, by the processor 103 of the information processing device 100 executing a program stored in the memory device 102. Figure 10 is a flowchart showing the process for generating a real estate utilization plan according to one embodiment of the present disclosure.

[0063] As shown in Figure 10, in step S101, the information processing device 100 acquires real estate-related information of the property. This real estate-related information may include, for example, real estate registration data, land value in the area where the property is located, land use designation, urban planning, and property usage information. The real estate-related information may be provided by the user, or...

[0064] In step S102, the information processing device 100 creates a prompt to cause the generation model 50 to generate a plan for the use of the real estate based on real estate-related information. Alternatively, the information processing device 100 may create a prompt to cause the generation model 50 to generate a plan for the use of the real estate based on highly accurate information and data stored in the database according to the RAG.

[0065] In step S103, the information processing device 100 inputs a prompt to the generation model 50. Specifically, the information processing device 100 sends a prompt to the cloud server that operates the generation model 50.

[0066] In step S104, the information processing device 100 obtains application ideas from the generated model 50. Specifically, the information processing device 100 obtains the generation results of the generated model 50 from the cloud server and provides the obtained generation results to the user.

[0067] According to the generation process described above, by using the generation model 50, users who are not familiar with real estate utilization can easily obtain utilization proposals for the properties being analyzed, thereby potentially improving the efficiency of operations in the real estate business.

[0068] Although embodiments of this disclosure have been described in detail above, this disclosure is not limited to the specific embodiments described above, and various modifications and changes are possible within the scope of the gist of this disclosure as described in the claims. [Explanation of Symbols]

[0069] 50 Generative Models 100 Information Processing Devices 110 Acquisition Department 120 Prompt Creation Section 130 Interface section

Claims

1. The acquisition unit acquires real estate-related information for real estate, A prompt generation unit creates prompts that cause a generation model to generate utilization plans for the real estate based on the real estate-related information, An interface unit that sends the prompt to the generation model and obtains the application proposal from the model, An information processing device having

2. The information processing apparatus according to claim 1, wherein the real estate-related information includes real estate registration data of the real estate.

3. The information processing device according to claim 1, wherein the aforementioned real estate-related information includes land value, land use designation, or urban plan for the area in which the real estate is located.

4. It is possible to connect to the database that stores the reference data, The information processing apparatus according to claim 1, wherein the prompt creation unit extracts reference data of the area where the real estate is located from the database and creates the prompt using the extracted reference data.

5. The information processing device according to claim 4, wherein the reference data includes real estate price information, parking information, railway line information, hazard maps, or real estate transaction information for each area.

6. The information processing apparatus according to claim 1, wherein the prompt generation unit determines the degree of shape or public road connectivity of the real estate based on the real estate-related information, and generates the prompt based on the determination result.

7. The acquisition unit acquires real estate usage information indicating the current usage status or past usage history of the real estate, The information processing apparatus according to claim 1, wherein the prompt generation unit generates the prompt based on the real estate usage information.

8. The information processing apparatus according to claim 1, wherein the prompt generation unit generates prompts for an owner who owns multiple properties to generate proposals for the use of the multiple properties based on a portfolio strategy.

9. The information processing device according to claim 1, wherein the prompt generation unit determines whether the owner of the real estate has borrowed money from a financial institution based on the real estate-related information, and determines whether the real estate is for the owner's residence based on the determination result.

10. The information processing apparatus according to claim 9, wherein the prompt generation unit generates a prompt to add the owner's attributes to the utilization plan based on the determination result.

11. To obtain real estate-related information, To create a prompt that causes the generation model to generate utilization plans for the real estate based on the aforementioned real estate-related information, Sending the prompt to the generation model and obtaining the proposed use from the generation model, An information processing method in which one or more computers perform the processing.

12. To obtain real estate-related information, To create a prompt that causes the generation model to generate utilization plans for the real estate based on the aforementioned real estate-related information, Sending the prompt to the generation model and obtaining the proposed use from the generation model, A program that causes one or more computers to execute a command.

Images