System, program and information processing method

Abstract

An object of the present invention is to provide a system, a program, and an information processing method that can accurately estimate the fair price of real estate. [Solution] The present invention is a system including one or more computers, whose control circuitry acquires target property information including details of a target property for which a price is to be calculated, acquires one or more competing property information including prices and details of one or more competing properties competing with the target property, outputs prompts to an LLM including instructions to cause the LLM to calculate a price for the target property based on the target property information and the one or more competing property information, and acquires calculation information generated by the LLM including the price of the target property.

Description

[Technical Field]

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

[0002] A known example of a conventional real estate transaction system is the real estate appraisal support system described in Patent Document 1. This real estate appraisal support system includes a cluster classification unit that classifies property information, including calculation information stored in a property database, into multiple clusters based on property attribute information, including proximity information regarding the property's proximity to a base station, newness information regarding the property's newness, and size information regarding the property's size, for each of the multiple clusters; a base rent calculation unit that calculates the base price (base rent) of properties in the cluster from the calculation information for each property; and a station market price index calculation unit that extracts the property information sharing a common nearest station and calculates a station market price index corresponding to the difference between the calculation information for the property and the base price of the cluster to which the property belongs. The real estate appraisal support system described in Patent Document 1 appropriately calculates market price indexes for location factors related to the location of the property, enabling the system to be used for objective and highly valid real estate price calculations. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2021-111161 Summary of the Invention [Problem to be solved by the invention]

[0004] In the field of systems relating to real estate transactions, there is a demand for accurate estimation of the fair price of real estate.

[0005] Therefore, an object of the present invention is to provide a system, a program, and an information processing method that can accurately estimate the fair price of real estate. [Means for solving the problem]

[0006] The first type of system is A system comprising one or more computers, The one or more computer control circuits Obtain target real estate information including details of the target real estate for which the price is to be calculated; Obtain one or more competitive property information items including prices and details of one or more competitive properties that compete with the target property; outputting a prompt to the LLM (Large Language Model) including an instruction to cause the LLM to calculate the price of the target property based on the target property information and the one or more competing property information; Obtaining calculated information generated by the LLM, including the price of the target property; It is a system.

[0007] The second form is The one or more computer control circuits generating the prompt using the target property information and the one or more competing property information; In the process of outputting the prompt, the prompt generated in the process of generating the prompt is output to the LLM. The system according to the first aspect.

[0008] The third form is The one or more computer control circuits Obtain target real estate identification information used to identify the target real estate; In the process of acquiring the target real estate information, the target real estate information is acquired based on the target real estate identification information. The system according to the first aspect.

[0009] The fourth form is The one or more computer control circuits Obtain competitive real estate identification information used to identify the one or more competitive real estate properties; In the process of acquiring one or more pieces of competing real estate information, the one or more pieces of competing real estate information are acquired based on the competing real estate identification information. The system according to the third aspect.

[0010] The fifth form is The one or more computer control circuits generating list information including one or more candidate target real estate properties that are candidates for the target real estate based on the target real estate identification information; In the process of acquiring the target real estate information, the target real estate information of a target candidate real estate selected by a user from one or more target candidate real estates included in the list information is acquired. The system according to the fourth aspect.

[0011] The sixth form is The competing real estate information includes the set price set for the competing real estate and / or the closing price when the competing real estate is sold. The system according to the fourth or fifth aspect.

[0012] The seventh form is the one or more computer memory circuits store a plurality of property information items including property prices and descriptions; The plurality of pieces of real estate information include one or more pieces of real estate information with closing prices, the one or more real estate information with closing prices includes the closing prices associated with one or more accounts; The one or more accounts include a first account; The one or more computer control circuits In the process of acquiring the competing real estate identification information, the competing real estate identification information input by the user having the first account is acquired, In the process of acquiring the competing real estate information, the one or more competing real estate information are extracted from the plurality of real estate information based on the competing real estate identification information; causing the first account to display the contract price associated with the first account, and not allowing accounts other than the first account to display the contract price associated with the first account; The system according to the sixth aspect.

[0013] The eighth form is the one or more computer memory circuits store a plurality of property information items including property prices and descriptions; The plurality of pieces of real estate information include one or more pieces of real estate information with closing prices, the one or more real estate information with closing prices includes the closing prices associated with one or more accounts; The one or more accounts include a first account; The one or more computer control circuits In the process of acquiring the competing real estate identification information, the competing real estate identification information input by the user having the first account is acquired, In the process of acquiring competitive real estate information, acquiring the one or more pieces of competing real estate information from the plurality of pieces of real estate information based on the competing real estate identification information, and extracting one or more pieces of real estate information with a closing price including the closing price associated with the first account from the one or more pieces of competing real estate information, while not extracting one or more pieces of real estate information with a closing price including the closing price associated with an account other than the first account from the one or more pieces of competing real estate information; Alternatively, extracting the one or more pieces of competing real estate information from the plurality of pieces of real estate information based on the competing real estate identification information, and extracting one or more pieces of real estate information with closing prices including the closing prices associated with accounts other than the first account, while excluding the closing prices. The system according to the sixth aspect.

[0014] The ninth form is The memo information is text data or audio data including the name of the property and the contract price, The one or more computer control circuits Acquire the memo information entered by the user having the first account; storing the memo information converted into the format of real estate information with closing prices in a storage circuit as real estate information with closing prices; The system according to the eighth aspect.

[0015] The tenth form is the target real estate information includes image information showing an image of the target real estate, The prompt further includes instructions to extract one or more first features of the target property included in the image information and cause the LLM to calculate a price for the target property based on the one or more first features. The system according to any one of the first to ninth aspects.

[0016] The 11th form is the prompt further includes instructions to cause the LLM to estimate one or more second characteristics of the view of the subject property based on the subject property information, and to calculate a price for the subject property based on the one or more second characteristics. The system according to any one of the first to tenth aspects.

[0017] The 12th form is The prompt further includes instructions to cause the LLM to identify facilities that can be added or modified to the target property based on the target property information, and to calculate the price of the target property if the facilities are added or modified, and the costs required to add or modify the facilities. The system according to any one of the first to eleventh aspects.

[0018] The 13th form is The facilities are facilities whose addition or modification will increase the price of the target real estate. The system according to the twelfth aspect.

[0019] The 14th form is the prompt includes instructions to cause the LLM to generate a description of the subject property based on the subject property information. The system according to any one of the first to thirteenth aspects.

[0020] The 15th form is the prompt includes an instruction to cause the description to describe advantages of the target property over the one or more competing properties based on the target property information and the one or more competing property information. The system according to the fourteenth aspect.

[0021] The 16th form is The target real estate is an unbuilt property, The prompt includes instructions to cause the LLM to calculate the value of the subject property if a virtual building is constructed on the subject property. The system according to any one of the first to fifteenth aspects.

[0022] The 17th form is one or more computer control circuits, Obtain target real estate information including details of the target real estate for which the price is to be calculated; obtaining one or more competitive property information items including prices and details of one or more competitive properties that compete with the target property; outputting a prompt to an LLM (Large Language Model) including an instruction to cause the LLM to calculate the price of the target property based on the target property information and the one or more competing property information; Obtaining calculated information including the price of the target real estate generated by the LLM; It is a program.

[0023] The 18th form is The control circuitry of one or more computers Obtain target real estate information including details of the target real estate for which the price is to be calculated; Obtain one or more competitive property information items including prices and details of one or more competitive properties that compete with the target property; outputting a prompt to a Large Language Model (LLM) including an instruction to cause the LLM to calculate the price of the target property based on the target property information and the one or more competing property information; Obtaining calculated information generated by the LLM, including the price of the target property; It is an information processing method. [Effects of the Invention]

[0024] According to the present invention, the fair price of real estate can be estimated with high accuracy. [Brief explanation of the drawings]

[0025] [Figure 1] FIG. 1 is a block diagram of the system 1. [Figure 2] FIG. 2 is a block diagram of the user terminal 10. [Figure 3] FIG. 3 is a block diagram of the server 110. [Figure 4] FIG. 4 is a block diagram of the LLM server 210. [Figure 5] FIG. 5 is a diagram showing a real estate information table. [Figure 6] FIG. 6 shows a conversion memo information table. [Figure 7] FIG. 7 shows an image displayed on the display 20 of the user terminal 10. [Figure 8] FIG. 8 shows an image displayed on the display 20 of the user terminal 10. [Figure 9] FIG. 9 shows an image displayed on the display 20 of the user terminal 10. [Figure 10] FIG. 10 shows an image displayed on the display 20 of the user terminal 10. [Figure 11] FIG. 11 shows an image displayed on the display 20 of the user terminal 10. [Figure 12]FIG. 12 shows an image displayed on the display 20 of the user terminal 10. [Figure 13] FIG. 13 is a flowchart showing the registration process executed by the control circuit 12 of the user terminal 10, the control circuit 112 of the server 110, and the control circuit 212 of the LLM server 210. [Figure 14] FIG. 14 is a flowchart of the calculation process executed by the control circuit 12 of the user terminal 10, the control circuit 112 of the server 110, and the control circuit 212 of the LLM server 210. [Figure 15] FIG. 15 is a flowchart of the calculation process executed by the control circuit 12 of the user terminal 10, the control circuit 112 of the server 110, and the control circuit 212 of the LLM server 210. [Figure 16] FIG. 16 is a flowchart of the calculation process executed by the control circuit 12 of the user terminal 10, the control circuit 112 of the server 110, and the control circuit 212 of the LLM server 210. [Figure 17] FIG. 17 shows the prompt P1. [Figure 18] FIG. 18 is a diagram showing the image information I13. [Figure 19] FIG. 19 shows the view of the subject property. [Figure 20] FIG. 20 is a diagram showing the image information I13. DETAILED DESCRIPTION OF THE INVENTION

[0026] (Embodiment) A system 1 according to an embodiment of the present disclosure will be described with reference to the drawings.

[0027] [System 1 Structure] First, the overall configuration of the system 1 will be described with reference to the drawings.

[0028] 1 includes a user terminal 10, a server 110, and an LLM (Large Language Models) server 210. The user terminal 10, the server 110, and the LLM server 210 are communicably connected to each other via a communication network. The network may be the Internet, an intranet, or the like.

[0029] [Structure of user terminal 10] The structure of the user terminal 10 will be described with reference to the drawings.

[0030] The user terminal 10 is an information processing device used by a user, such as a computer. The user may be, for example, a real estate agent. The user terminal 10 may be, for example, a smartphone, a tablet terminal, or a personal computer. As shown in FIG. 2 , the user terminal 10 includes a control circuit 12, a memory circuit 14, a network interface 16, a graphics processing unit 18, a display 20, and an operation unit 26.

[0031] The memory circuit 14 stores programs and data and is, for example, a combination of a read-only memory (ROM), a random access memory (RAM), and a storage (for example, a flash memory or a hard disk).

[0032] The programs include, for example, the following programs: OS (Operating System) programs - Programs for applications that process information (e.g., web browsers or target apps described below)

[0033] The data includes, for example, the following data: Databases referenced in information processing Data obtained by performing information processing (i.e., the results of performing information processing)

[0034] The control circuit 12 executes a program stored in the memory circuit 14 to realize the functions of the user terminal 10. The control circuit 12 is, for example, a circuit including at least one of the following: ·CPU(Central Processing Unit) ·GPU(Graphic Processing Unit) ·ASIC(Application Specific Integrated Circuit) ·FPGA(Field Programmable Array)

[0035] The network interface 16 is a transmitting circuit and a receiving circuit that controls communication between the user terminal 10 and an external device, which is a server 110.

[0036] The graphics processing unit 18 displays an image on the display 20 based on the image data generated by the control circuit 12. The display 20 is a liquid crystal display or an organic EL (Electro Luminescence) display.

[0037] The operation unit 26 generates an operation signal based on a user's operation and outputs the operation signal to the control circuit 12. The operation unit 26 is, for example, a mouse, a keyboard, a touch panel, or the like.

[0038] [Server 110 Structure] The structure of the server 110 will be described with reference to the drawings.

[0039] The server 110 is a computer that can be accessed by the user terminal 10 via a network. As shown in Fig. 3, the server 110 includes a control circuit 112, a memory circuit 114, and a network interface 116.

[0040] The memory circuit 114 stores the program PG1 and data. The memory circuit 114 is, for example, a combination of a read-only memory (ROM), a random access memory (RAM), and a storage (for example, a flash memory or a hard disk).

[0041] The control circuit 112 executes the program PG1 stored in the memory circuit 114 to realize the functions of the server 110. The control circuit 112 is, for example, a circuit including at least one of the following: ·CPU(Central Processing Unit) ·GPU(Graphic Processing Unit) ·ASIC(Application Specific Integrated Circuit) ·FPGA(Field Programmable Array)

[0042] As shown in Figure 3, the control circuit 112 includes, as functional blocks, a target real estate information acquisition means 120, a competing real estate information acquisition means 122, a prompt output means 124, a calculation information acquisition means 126, a prompt generation means 128, a target real estate identification information acquisition means 130, a competing real estate identification information acquisition means 132, a list information generation means 134, a memo information acquisition means 136, a memory control means 138, and a display control means 140.

[0043] The network interface 116 is a transmitting circuit and a receiving circuit that controls communication between the server 110 and external devices, such as the user terminal 10 and the LLM server 210.

[0044] [LLM Server 210 Structure] The structure of the LLM server 210 will be described with reference to the drawings. Figure 4 is a block diagram of the LLM server 210.

[0045] The LLM server 210 is a computer that can be accessed by the server 110 via a network. As shown in FIG. 4 , the LLM server 210 includes a control circuit 212, a memory circuit 214, and a network interface 216.

[0046] The memory circuit 214 stores programs and data and is, for example, a combination of a read-only memory (ROM), a random access memory (RAM), and a storage (for example, a flash memory or a hard disk).

[0047] The control circuit 212 executes the program stored in the memory circuit 214 to realize the functions of the LLM server 210. The control circuit 212 is, for example, a circuit including at least one of the following: ·CPU(Central Processing Unit) ·GPU(Graphic Processing Unit) ·ASIC(Application Specific Integrated Circuit) ·FPGA(Field Programmable Array)

[0048] The network interface 216 is a transmitting and receiving circuit that controls communication between the LLM server 210 and an external device, such as the server 110.

[0049] The LLM server 210 also includes a first LLM 218 and a second LLM 220. The LLM is a natural language processing model trained using a large amount of text data. For example, the LLM can perform natural language processing such as text generation by fine-tuning a pre-trained model.

[0050] [System 1 operation] Next, the operation of the system 1 will be described with reference to the drawings. Fig. 5 is a diagram showing a real estate information table. Fig. 6 is a diagram showing a conversion memo information table.

[0051] When the control circuit 112 of the server 110 reads out the program PG1 stored in the memory circuit 114, the program PG1 causes the control circuit 112 of the server 110 to execute the operations described below. The program PG1 causes the control circuit 112 of the server 110 to execute a target real estate information acquisition means 120, a competing real estate information acquisition means 122, a prompt output means 124, a calculation information acquisition means 126, a prompt generation means 128, a target real estate specific information acquisition means 130, a competing real estate specific information acquisition means 132, a list information generation means 134, a memo information acquisition means 136, a memory control means 138, and a display control means 140. In addition, in the information processing method, the control circuit 112 of the server 110 executes the target real estate information acquisition means 120, the competing real estate information acquisition means 122, the prompt output means 124, the calculation information acquisition means 126, the prompt generation means 128, the target real estate identification information acquisition means 130, the competing real estate identification information acquisition means 132, the list information generation means 134, the memo information acquisition means 136, the memory control means 138, and the display control means 140.

[0052] The memory circuit 114 of the server 110 stores the real estate information table shown in FIG. 5. The real estate information table includes multiple pieces of real estate information I1, including the rent (price) and details of office properties for rent. More specifically, the real estate information I1 includes multiple items K1. The multiple items K1 include item names and contents. The multiple item names include the property name, address, nearest station, completion date, floor number, total rentable area, standard floor area, air conditioning, OA floor, ceiling height, structure, asking area, asking rent, current vacancy rate, contracted area, contracted rent, company, and account. Each of the multiple contents is described in the multiple item name columns. The asking rent is the set price set for the property. In other words, it is the rent for the property previously set by the user or another user. The contracted rent is the contracted price when a lease agreement for the property is concluded. The company is the buyer (tenant) when the property is contracted. The price of the property is the asking rent and the contracted rent. Property details include property name, address, nearest station, completion date, number of floors, total rental area, standard floor area, air conditioning, OA floor, ceiling height, structure, available area, current vacancy rate, sold area, company and account.

[0053] The details of the real estate name, address, nearest station, completion date, floors, total rentable area, standard floor area, air conditioning, OA floor, ceiling height, structure, available area, available rent, and current vacancy rate are input by the operating company of server 110. Meanwhile, the contracted area, contracted rent, and company details are input by the user. The account is the account of the user who inputs the contracted area, contracted rent, and company details. Since there are one or more users, there are one or more accounts in system 1. The one or more accounts include an account that is AAA (first account).

[0054] Here, as shown in FIG. 5, the multiple pieces of real estate information I1 include one or more pieces of real estate information I6 with contract prices. The one or more pieces of real estate information I6 with contract prices include contract rents (contract prices) associated with one or more accounts. In the real estate information table shown in FIG. 5, the real estate information I6 with contract prices is real estate information I1 for Building B and Building C. The contracted area, contract rent, and company details of real estate information I1 for Building B and Building C are entered by a user with an AAA account. Therefore, the real estate information I6 with contract prices for Building B and Building C includes contract rents (contract prices) associated with an AAA account. Also, although not shown, the one or more pieces of real estate information I6 with contract prices also include contract rents (contract prices) associated with accounts other than AAA. Real estate information I1 for Building A does not include contract rents, so it is not real estate information I6 with contract prices.

[0055] The storage circuit 114 of the server 110 stores the conversion memo information table shown in Fig. 6. The conversion memo information table shown in Fig. 6 is a conversion memo information table for a user who has an AAA account. The storage circuit 114 of the server 110 also stores conversion memo information tables for users whose accounts are other than AAA.

[0056] The conversion memo information table includes multiple pieces of conversion memo information I7. The multiple pieces of conversion memo information I7 are information related to real estate that has been sold. The multiple pieces of conversion memo information I7 are information obtained by converting the format of memo information I8, which will be described later. The memo information I8 is input by a user with an AAA account. As shown in FIG. 6, the conversion memo information I7 includes multiple items K2. The multiple items K2 include item names and contents. The multiple item names include the information acquisition date, real estate name, floor, area, rent, company, and remarks. Each of the multiple contents is an item described in the multiple item name columns.

[0057] The contents of the conversion memo information table shown in Figure 6 are reflected in the real estate information table shown in Figure 5. Specifically, the details of the contracted area, contracted rent, and company in the real estate information I6 with contracted price for Building B shown in Figure 5 match the details of the contracted area, contracted rent, and company in the conversion memo information I7 for Building B shown in Figure 6. Similarly, the details of the contracted area, contracted rent, and company in the real estate information I6 with contracted price for Building C shown in Figure 5 match the details of the contracted area, contracted rent, and company in the conversion memo information I7 for Building C shown in Figure 6.

[0058] Next, an overview of the operation of the system 1 will be described with reference to the drawings. Figures 7 to 12 show images displayed on the display 20 of the user terminal 10.

[0059] The system 1 executes a registration process and a calculation process. The registration process is a process in which the server 110 reflects information collected by the user in the real estate information I1. In the registration process, the user terminal 10 displays the memo information input image shown in FIG. 7. The memo information input image includes a memo information display section 300 and a memo information input section 302. The memo information display section 300 includes multiple pieces of converted memo information I7 input by a user with an account that is an AAA.

[0060] The memo information input section 302 is an input section for a user with an AAA account to input converted memo information 17 or memo information 18. More specifically, the memo information input section 302 includes a plurality of boxes B1. The plurality of boxes B1 include a property name box B1, a floor box B1, an area box B1, a rent box B1, a company box B1, a remarks box B1, and a memo box B1.

[0061] The property name box B1, floor box B1, area box B1, rent box B1, company box B1, and remarks box B1 are boxes for users with AAA accounts to enter conversion memo information I7. The property name of the concluded property is entered in the property name box B1. The floor of the concluded property is entered in the floor box B1. The area of ​​the concluded property (contracted area) is entered in the area box B1. The rent box B1 is entered with the rent of the concluded property (contracted rent). The company box B1 is entered with the tenant of the concluded property. The remarks box B1 is used to enter supplementary information that cannot be entered in the property name box B1, floor box B1, area box B1, rent box B1, or company box B1. A user with an AAA account enters conversion memo information I7 into the property name box B1, floor box B1, area box B1, rent box B1, company box B1, and remarks box B1, and presses the register button 304. Then, the server 110 displays the converted conversion memo information I7 in the memo information display section 300, and also registers it in the conversion memo information table shown in FIG.

[0062] The memo box B1 is a box for a user with an AAA account to input memo information I8. The memo information I8 is text data including a real estate name and a contract price. For example, a user with an AAA account inputs text such as "Apparently, 152.2 m2 on the 19th floor of Building B was rented to Company X for 42,000 yen" into the memo box B1 as memo information I8. The user with an AAA account then presses the register button 304. This causes the server 110 to convert the memo information I8 into the format of converted memo information I7 in the LLM server 210. Specifically, the LLM server 210 identifies the real estate name as Building B, the floor as the 19th floor, the contracted area as 152.2 m2, the contracted rent as 42,000 yen / m2, and the company as Company X, and converts the memo information I8 into converted memo information I7. Then, the server 110 displays the converted conversion memo information I7 in the memo information display section 300, and also registers it in the conversion memo information table shown in FIG.

[0063] As described above, the server 110 registers the conversion memo information I7 registered in the conversion memo information table shown in FIG. 6 to the real estate information I1 registered in the real estate information table shown in FIG. 5. More specifically, the server 110 registers the contracted area, contracted rent, and company details included in the conversion memo information I7 with the contracted area, contracted rent, and company details included in the real estate information I1 corresponding to the real estate name included in the conversion memo information I7. In this embodiment, the server 110 registers 42,000 yen / tsubo, 1528 tsubo, and Company X in the contracted area, contracted rent, and company details included in the real estate information I1 for Building B. Furthermore, the server 110 registers the user's account, AAA, in the account details included in the real estate information I1 for which the real estate name is Building B. This reflects the information collected by the user in the real estate information I1.

[0064] The calculation process is a process in which the server 110 calculates the price of a target property selected by a user based on the prices of competing properties. In the calculation process, the user terminal 10 displays the target property input image shown in FIG. 8. The target property input image includes a box B2. A user with an AAA account inputs target property identification information I2 into box B2. The target property identification information I2 is information used to identify the target property. In this embodiment, the target property identification information I2 is a keyword. A keyword is, for example, all or part of the property name of the target property. A part of the property name of the target property is, for example, a place name or the name of the building owner. In this embodiment, a user with an AAA account inputs Building A into box B2. As a result, the user terminal 10 displays the target property input image including the property information I1 of Building A on the display 20. Then, a user with an AAA account checks the checkbox for the property information I1 of Building A, thereby selecting Building A as the target property.

[0065] Next, the user terminal 10 displays the competing real estate input image shown in FIG. 9. The competing real estate input image includes real estate information I1 for Building B, Building C, Building D, Building E, and Building F. Buildings B, C, D, E, and F are competing candidate real estate that may compete with Building A, the target real estate. Buildings B, C, D, E, and F are selected by the server 110. For example, the server 110 identifies as competing candidate real estate a real estate that has an address and standard floor area similar to the address and standard floor area of ​​Building A. Then, a user with an AAA account checks the checkboxes for the real estate information I1 for Building B and Building C, thereby selecting Buildings B and C as competing real estate.

[0066] The competing real estate input image also includes a box B3, as shown in FIG. 10 . A user with an AAA account inputs competing real estate identification information I4 into box B3. The competing real estate identification information I4 is information used to identify competing real estate. In this embodiment, the competing real estate identification information I4 is a keyword. A keyword is, for example, all or part of the real estate name of a competing real estate. A part of the real estate name of a competing real estate is a place name, the name of the building owner, or the like. In this embodiment, a user with an AAA account inputs X Building into box B3. As a result, the user terminal 10 displays the competing real estate input image including real estate information I1 for X Building. Then, a user with an AAA account checks the checkbox for the real estate information I1 for X Building, thereby selecting X Building as a competing real estate.

[0067] A user with an AAA account presses the Create Calculation Report button 400 included in the competing real estate input image shown in FIG. 9 or 10. In response, the user terminal 10 displays the calculation request image shown in FIG. 11. The calculation request image includes a calculated rent box B4 and a calculation basis box B5. Box B4 displays the rent of Building A. Box B5 displays the calculation basis for the rent of Building A. However, since the price of Building A has not been calculated in the calculation request image shown in FIG. 11, boxes B4 and B5 are blank. The calculation request image shown in FIG. 11 also includes target real estate information I3 and one or more competing real estate information I5. The target real estate information I3 is real estate information I1 including details of the target real estate for which a price is to be calculated. The one or more competing real estate information I5 is real estate information I1 including the rent (price) and details of one or more competing real estate that competes with the target real estate. The rent (price) of a competing property is the asking rent (set price) set for the competing property, or the agreed rent (agreement price) when a deal is concluded for the competing property.

[0068] A user with an AAA account presses the calculation request button 402 included in the calculation request image shown in Figure 11. This causes the user terminal 10 to have the LLM server 210 generate calculation information I9 including the rent (price) of the target real estate and the calculation basis for the rent (price) based on the target real estate information I3 and one or more pieces of competing real estate information I5. The user terminal 10 then displays the calculation result image shown in Figure 12, in which the calculation information I9 is ​​written in boxes B4 and B5. This allows the user with an AAA account to know the rent and the calculation basis for the rent of Building A, which is the target real estate.

[0069] Next, a description will be given of a flowchart of the registration process executed by the system 1. Fig. 13 is a flowchart of the registration process executed by the control circuit 12 of the user terminal 10, the control circuit 112 of the server 110, and the control circuit 212 of the LLM server 210.

[0070] First, the control circuit 12 of the user terminal 10 and the control circuit 112 of the server 110 execute a login process (steps S1, S101). In the login process, a user who has an AAA account operates the operation unit 26 of the user terminal 10 to input account information including an account name and password. The control circuit 12 of the user terminal 10 and the control circuit 112 of the server 110 execute the login process using the account information. However, since the login process is a common process, further explanation will be omitted.

[0071] Next, the control circuit 12 of the user terminal 10 causes the display 20 to display the memo information input image shown in FIG. 7 (step S2). In response, the user with an AAA account operates the operation unit 26 of the user terminal 10 to input memo information I8 into the memo box B1. In this embodiment, the user with an AAA account inputs "Apparently, 1528 m2 on the 19th floor of Building B has been rented to Company X for 42,000 yen" into the memo box B1. As a result, the control circuit 12 (memo information acquisition means 136) of the user terminal 10 acquires the memo information I8 input by the user with an AAA account (step S3).

[0072] Next, the control circuit 12 of the user terminal 10 transmits the memo information I8 to the server 110 via the network interface 16 (step S4). In response, the network interface 116 of the server 110 receives the memo information I8 and outputs the memo information I8 to the control circuit 112. As a result, the control circuit 112 (memo information acquisition means 136) of the server 110 acquires the memo information I8 input by the user who has an account (first account) that is AAA (step S102).

[0073] Next, the control circuit 112 of the server 110 generates a prompt P2 (step S103). The memory circuit 114 of the server 110 stores a prompt P12. The prompt P12 includes an instruction to cause the LLM server 210 to convert the memo information I8 into converted memo information I7. More specifically, the prompt P12 includes the following instructions: Extract property name, floor, contracted area, contracted rent, company, and notes from memo information I8. Generate conversion memo information I7 including the extracted property name, floor, contracted area, contracted rent, company, and remarks. The control circuit 112 of the server 110 generates the prompt P2 by including the memo information I8 in the prompt P12.

[0074] Next, the control circuit 112 of the server 110 transmits a prompt P2 including the memo information I8 to the LLM server 210 via the network interface 116 (step S104). In response, the network interface 216 of the LLM server 210 receives the prompt P2 including the memo information I8 and outputs the prompt P2 including the memo information I8 to the control circuit 212. As a result, the control circuit 212 of the LLM server 210 acquires the prompt P2 including the memo information I8 (step S201).

[0075] Next, the control circuit 212 of the LLM server 210 generates converted memo information I7 using the second LLM 220 based on the prompt P2 including the memo information I8 (step S202). The control circuit 212 of the LLM server 210 converts memo information I8 indicating "Apparently, 1528 m2 on the 19th floor of Building B has been rented to Company X for 42,000 yen" into converted memo information I7 indicating that the real estate name is Building B, the floor is the 19th floor, the contracted area is 1528 m2, the contracted rent is 42,000 yen / m2, and the company is Company X.

[0076] Next, the control circuit 212 of the LLM server 210 transmits the conversion memo information I7 to the server 110 via the network interface 216 (step S203). In response, the network interface 116 of the server 110 receives the conversion memo information I7 and outputs the conversion memo information I7 to the control circuit 112. As a result, the control circuit 112 of the server 110 acquires the conversion memo information I7 (step S105).

[0077] Next, the control circuit 112 of the server 110 stores the conversion memo information I7 in the memory circuit 114 (step S106). In this embodiment, the control circuit 112 of the server 110 registers the conversion memo information I7 in the conversion memo information table shown in FIG.

[0078] Furthermore, the control circuit 112 of the server 110 generates real estate information I6 with a closing price based on the conversion memo information I7 (step S107). More specifically, the control circuit 112 of the server 110 extracts real estate information I1, including the real estate name included in the conversion memo information I7, from the real estate information table shown in FIG. 5. In this embodiment, the control circuit 112 of the server 110 extracts the real estate information I1 for Building B. The control circuit 112 of the server 110 then registers 42,000 yen / tsubo, 1528 tsubo, and Company X in the contracted area, contracted rent, and company details included in the real estate information I1 for Building B, and registers the user's account AAA in the account details included in the real estate information I1 for Building B. As a result, the control circuit 112 (storage control means 138) of the server 110 stores the memo information I8, converted into the format of the real estate information I6 with a closing price, in the memory circuit 114 as the real estate information I6 with a closing price (step S108). With the above processing, the registration processing is completed.

[0079] Next, a description will be given of a flowchart executed by the system 1 in the calculation process. Figures 14 to 16 are flowcharts executed by the control circuit 12 of the user terminal 10, the control circuit 112 of the server 110, and the control circuit 212 of the LLM server 210 in the calculation process. Figure 17 is a diagram showing prompt P1.

[0080] First, the control circuit 12 of the user terminal 10 and the control circuit 112 of the server 110 execute a login process (steps S11, S111). In the login process, a user who has an AAA account operates the operation unit 26 of the user terminal 10 to input account information including an account name and password. The control circuit 12 of the user terminal 10 and the control circuit 112 of the server 110 execute the login process using the account information. However, since the login process is a common process, further explanation will be omitted.

[0081] Next, the control circuit 12 of the user terminal 10 displays the target real estate input image shown in FIG. 8 on the display 20 (step S12). However, at the stage of step S12, the target real estate input image does not include the real estate information I1 of Building A. In response, a user with an AAA account operates the operation unit 26 of the user terminal 10 to input the target real estate identification information I2 into box B2. In this embodiment, a user with an AAA account inputs Building A as a keyword into box B2. As a result, the control circuit 12 of the user terminal 10 acquires the target real estate identification information I2 (step S13).

[0082] Next, the control circuit 12 of the user terminal 10 transmits the target real estate identification information I2 to the server 110 via the network interface 16 (step S14). In response, the network interface 116 of the server 110 receives the target real estate identification information I2 and outputs the target real estate identification information I2 to the control circuit 112. As a result, the control circuit 112 of the server 110 (target real estate identification information acquisition means 130) acquires the target real estate identification information I2 used to identify the target real estate (step S112).

[0083] Next, the control circuit 112 (list information generation means 134) of the server 110 generates first list information I10 including one or more target candidate properties that are candidates for one or more target properties based on the target real estate identification information I2 (step S113). More specifically, the control circuit 112 of the server 110 extracts real estate information I1 that includes Building A, which is included in the target real estate identification information I2, from the real estate information table. In this embodiment, Building A is the target candidate property. Therefore, the control circuit 112 of the server 110 generates first list information I10 including the real estate information I1 of Building A.

[0084] Next, the control circuit 112 of the server 110 transmits the first list information I10 to the user terminal 10 via the network interface 116 (step S114). In response, the network interface 16 of the user terminal 10 receives the first list information I10 and outputs the first list information I10 to the control circuit 12. As a result, the control circuit 12 of the user terminal 10 acquires the first list information I10 (step S15).

[0085] Next, the control circuit 12 of the user terminal 10 displays the target real estate input image shown in FIG. 8 on the display 20 based on the first list information I10 (step S16). In this embodiment, the first list information I10 includes the real estate information I1 of Building A, and therefore the target real estate input image includes the real estate information I1 of Building A. A user with an account AAA checks the checkbox for the real estate information I1 of Building A. This selects Building A as the target real estate. Therefore, the control circuit 12 of the user terminal 10 acquires the real estate information I1 of Building A as the target real estate information I3 (step S17).

[0086] Next, the control circuit 12 of the user terminal 10 transmits the target real estate information I3 to the server 110 via the network interface 16 (step S18). In response, the network interface 116 of the server 110 receives the target real estate information I3 and outputs the target real estate information I3 to the control circuit 112. As a result, the control circuit 112 of the server 110 (target real estate information acquisition means 120) acquires the target real estate information I3 of the target candidate real estate selected by the user with an AAA account from one or more target candidate real estate included in the first list information I10 (step S115). In this embodiment, the target real estate information I3 is the real estate information I1 of Building A. Since the first list information I10 is generated based on the target real estate identification information I2 in step S113, the control circuit 112 of the server 110 (target real estate information acquisition means 120) acquires the target real estate information I3 based on the target real estate identification information I2 acquired in step S112.

[0087] Next, the control circuit 112 of the server 110 generates second list information I11 including one or more competing candidate properties that are candidates for one or more competing properties based on the target real estate information I3 (step S116). More specifically, the control circuit 112 of the server 110 identifies properties with addresses and standard floor areas similar to those of Building A as competing candidate properties. In this embodiment, the control circuit 112 of the server 110 identifies Building B, Building C, Building D, Building E, and Building F as competing candidate properties. Then, the control circuit 112 of the server 110 generates second list information I11 including real estate information I1 for Building B, Building C, Building D, Building E, and Building F.

[0088] Next, the control circuit 112 of the server 110 transmits the second list information I11 to the user terminal 10 via the network interface 116 (step S117). In response, the network interface 16 of the user terminal 10 receives the second list information I11 and outputs the second list information I11 to the control circuit 12. As a result, the control circuit 12 of the user terminal 10 acquires the second list information I11 (step S19).

[0089] Next, the control circuit 12 of the user terminal 10 displays the competing real estate input image shown in FIG. 9 on the display 20 based on the second list information I11 (step S20). In this embodiment, the second list information I11 includes real estate information I1 for Building B, Building C, Building D, Building E, and Building F, so the competing real estate input image includes real estate information I1 for Building B, Building C, Building D, Building E, and Building F. A user with an AAA account checks the checkboxes for real estate information I1 for Building B and Building C. This selects Building B and Building C as competing real estate. Therefore, the control circuit 12 of the user terminal 10 acquires real estate information I1 for Building B and Building C as competing real estate information I5 (step S21).

[0090] Next, the control circuit 12 of the user terminal 10 displays the competing real estate input image shown in FIG. 10 on the display 20 (step S22). However, at the stage of step S22, box B3 is blank. Furthermore, the competing real estate input image does not include the real estate information I1 of X Building. A user with an AAA account inputs competing real estate identification information I4 into box B3. The competing real estate identification information I4 is information used to identify competing real estate. In this embodiment, a user with an AAA account inputs X Building as a keyword into box B3. As a result, the control circuit 12 of the user terminal 10 acquires competing real estate identification information I4 that includes X Building as a keyword (step S23).

[0091] Next, the control circuit 12 of the user terminal 10 transmits the competing real estate identification information I4 to the server 110 via the network interface 16 (step S24). In response, the network interface 116 of the server 110 receives the competing real estate identification information I4 and outputs the competing real estate identification information I4 to the control circuit 112. As a result, the control circuit 112 of the server 110 (competitive real estate identification information acquisition means 132) acquires the competing real estate identification information I4 input by the user with an AAA account (step S118).

[0092] Next, the control circuit 112 of the server 110 generates third list information I12 including one or more competing candidate properties that are candidates for one or more competing properties based on the competing property identification information I4 (step S119). More specifically, the control circuit 112 of the server 110 extracts real estate information I1 that includes X Building in the property name included in the target real estate identification information I2 from the real estate information table. In this embodiment, X Building is the competing candidate property. Therefore, the control circuit 112 of the server 110 generates third list information I12 including the real estate information I1 of X Building.

[0093] Next, the control circuit 112 of the server 110 transmits the third list information I12 to the user terminal 10 via the network interface 116 (step S120). In response, the network interface 16 of the user terminal 10 receives the third list information I12 and outputs the third list information I12 to the control circuit 12. As a result, the control circuit 12 of the user terminal 10 acquires the third list information I12 (step S25).

[0094] Next, the control circuit 12 of the user terminal 10 displays the competing real estate input image shown in FIG. 10 on the display 20 based on the third list information I12 (step S26). In this embodiment, the third list information I12 includes the real estate information I1 of X Building, so the competing real estate input image includes the real estate information I1 of X Building. A user with an AAA account checks the checkbox for the real estate information I1 of X Building. This adds X Building to the target real estate. Therefore, the control circuit 12 of the user terminal 10 acquires the real estate information I1 of X as the competing real estate information I5 (step S27).

[0095] Next, the control circuit 12 of the user terminal 10 transmits three pieces of competing real estate information I5, including the real estate information I1 of Building B, Building C, and Building X, to the server 110 via the network interface 16 (step S28). In response, the network interface 116 of the server 110 receives the three pieces of competing real estate information I5, including the real estate information I1 of Building B, Building C, and Building X, and outputs the three pieces of competing real estate information I5, including the real estate information I1 of Building B, Building C, and Building X, to the control circuit 112. As a result, the control circuit 112 of the server 110 acquires the three pieces of competing real estate information I5, including the real estate information I1 of Building B, Building C, and Building X (step S121). As described above, the control circuit 112 of the server 110 acquires one or more pieces of competing real estate information I5 based on the competing real estate identification information I4 acquired in step S118.

[0096] Next, the control circuit 112 of the server 110 extracts one or more pieces of real estate information I6 with closing prices from the three pieces of competing real estate information I5 (step S122). The real estate information I6 with closing prices is real estate information I1 including closing prices associated with the AAA account (first account). Therefore, the control circuit 112 of the server 110 extracts the three pieces of competing real estate information I5, namely, Building B, Building C, and Building X, including closing prices associated with the AAA account (first account), as the three pieces of real estate information I6 with closing prices. In this way, the control circuit 112 (competitive real estate information acquisition means 122) of the server 110 extracts one or more pieces of real estate information I6 with closing prices including closing rents (closing prices) associated with the AAA account (first account) from the three pieces of competing real estate information I5, while not extracting one or more pieces of real estate information I6 with closing prices including closing rents (closing prices) associated with accounts other than the AAA account (first account) from the three pieces of competing real estate information I5.

[0097] Next, the control circuit 112 of the server 110 transmits the target real estate information I3, which is the real estate information I1 of Building A, and the three competing real estate information I5 (three pieces of real estate information I6 with contract prices) of Building B, Building C, and Building X to the user terminal 10 via the network interface 16 (step S123). In response, the network interface 16 of the user terminal 10 receives the target real estate information I3, which is the real estate information I1 of Building A, and the three competing real estate information I5 (three pieces of real estate information I6 with contract prices) of Building B, Building C, and Building X, and outputs the target real estate information I3, which is the real estate information I1 of Building A, and the three competing real estate information I5 (three pieces of real estate information I6 with contract prices) of Building B, Building C, and Building X to the control circuit 12. As a result, the control circuit 12 of the user terminal 10 acquires the target real estate information I3, which is the real estate information I1 of Building A, and three competing real estate information I5 (three real estate information with closing prices I6), which are the three real estate information I1 of Building B, Building C, and Building X (step S29).

[0098] Next, the control circuit 12 of the user terminal 10 displays the calculation request image shown in FIG. 11 on the display 20 based on the target real estate information I3, which is the real estate information I1 of Building A, and the three competing real estate information I5 (three real estate information with contract prices I6), which are the three real estate information I1 of Building B, Building C, and Building X (step S30). The calculation request image shown in FIG. 11 includes a calculated rent box B4 and a calculation basis box B5. Box B4 displays the rent of Building A. Box B5 displays the calculation basis for the rent of Building A. However, since the price of Building A has not been calculated in the calculation request image shown in FIG. 11, boxes B4 and B5 are blank. The calculation request image shown in FIG. 11 also includes the target real estate information I3, which is the real estate information I1 of Building A, and the three competing real estate information I5 (three real estate information with contract prices I6), which are the three real estate information I1 of Building B, Building C, and Building X. The three competing real estate information I5, which are the three pieces of real estate information I1 for Building B, Building C, and Building X, are real estate information I6 with contract prices including contract rents. Therefore, the calculation request image includes the contract rents for Building B, Building C, and Building X. This allows users with AAA accounts to view the contract rents associated with their AAA accounts. As described above, the control circuit 112 (display control means 140) of the server 110 causes the AAA account (first account) to display the contract rent (contract price) associated with the AAA account (first account), and does not cause accounts other than the AAA account (first account) to display the contract rent (contract price) associated with the AAA account (first account) on the display 20 of the user terminal 10.

[0099] A user with an account that is AAA presses the calculation request button 402 included in the calculation request image shown in Fig. 11. This causes the control circuit 12 of the user terminal 10 to send calculation request information to the server 110 via the network interface 16 (step S31). The calculation request information indicates that the user terminal 10 requests the server 110 to calculate the rent for Building A. In response, the network interface 116 of the server 110 receives the calculation request information and outputs the calculation request information to the control circuit 112. This causes the control circuit 112 of the server 110 to acquire the calculation request information (step S124).

[0100] Next, the control circuit 112 (prompt generation means 128) of the server 110 generates a prompt P1 using the target real estate information I3, which is the real estate information I1 of Building A, and the three competing real estate information I5, which is the three real estate information I1 of Building B, Building C, and Building X (step 125). More specifically, the memory circuit 114 of the server 110 stores a prompt P11 shown in FIG. 17. The prompt P11 includes instructions to have the first LLM 218 calculate the rent (price) of the target real estate based on the target real estate information I3, which is the real estate information I1 of Building A, and the three competing real estate information I5, which is the three real estate information I1 of Building B, Building C, and Building X. In this embodiment, the prompt P11 includes instructions to have the first LLM 218 calculate an up scenario, a main scenario, and a down scenario for the rent of the target real estate. The up scenario is the rent that would result in the best possible outcome for the plan or forecast. The main scenario is the most likely rent or the most reasonable rent considering current conditions and existing trends. The down scenario is the rent that would result from the worst-case scenario in the plan or forecast. Furthermore, prompt P11 includes instructions to have the first LLM 218 output the basis for calculating the rent. However, prompt P11 shown in FIG. 12 does not include target real estate information I3, which is real estate information I1 for Building A, or three competing real estate information I5, which is real estate information I1 for Building B, Building C, and Building X. Therefore, the control circuit 112 of server 110 includes target real estate information I3, which is real estate information I1 for Building A, and three competing real estate information I5, which is real estate information I1 for Building B, Building C, and Building X, in prompt P11. As a result, the control circuit 112 of server 110 generates prompt P1 shown in FIG. 12.

[0101] Next, the control circuit 112 of the server 110 transmits the prompt P1 shown in Figure 17 to the LLM server 210 via the network interface 116 (step S126). In response, the network interface 216 of the LLM server 210 acquires the prompt P1 shown in Figure 12 and outputs the prompt P1 shown in Figure 12 to the control circuit 212. As a result, the control circuit 212 of the LLM server 210 acquires the prompt P1 shown in Figure 12 (step S301).

[0102] Next, the control circuit 212 of the LLM server 210 outputs the prompt P1 shown in FIG. 17 to the first LLM 218, thereby causing the first LLM 218 to generate calculation information I9 including the rent and the calculation basis for the rent for Building A, the target real estate (step S302). The three competing real estate information I5 included in the prompt P1 include the asking rents (set prices) set for Buildings B, C, and X, which are competing real estate, and the contracted rents (contract prices) when the competing real estate, Buildings B, C, and X, are sold. Therefore, the control circuit 212 of the LLM server 210 causes the first LLM 218 to generate calculation information I9 based on the asking rents and contracted rents for Buildings B, C, and X. Furthermore, the control circuit 212 of the LLM server 210 causes the first LLM 218 to generate calculation information I9 based on the specifications (type and structure of air conditioning) of Buildings B, C, and X. Thus, in steps S125, S126, S301, and S302, the control circuit 112 (prompt output means 124) of the server 110 outputs to the first LLM 218 the prompt P1 generated in the process of generating the prompt P1 (step S125).

[0103] Next, the control circuit 212 of the LLM server 210 transmits the calculation information I9 to the server 110 via the network interface 216 (step S303). In response, the network interface 116 of the server 110 receives the calculation information I9 and outputs the calculation information I9 to the control circuit 112. As a result, the control circuit 112 (calculation information acquisition means 126) of the server 110 acquires the calculation information I9 including the rent (price) and the calculation basis for the rent (price) of Building A, which is the target real estate, generated by the first LLM 218 (step S127).

[0104] Next, the control circuit 112 of the server 110 transmits the calculation information I9 to the user terminal 10 via the network interface 116 (step S128). In response, the network interface 16 of the user terminal 10 receives the calculation information I9 and outputs the calculation information I9 to the control circuit 12. As a result, the control circuit 12 of the user terminal 10 acquires the calculation information I9 (step S32).

[0105] Next, the control circuit 12 of the user terminal 10 uses the calculation information I9 to display the calculation result image shown in FIG. 12 on the display 20 (step S33). More specifically, the control circuit 12 of the user terminal 10 displays the rent for Building A included in the calculation information I9 in box B4, and also displays the calculation basis for the rent for Building A included in the calculation information I9 in box B5. This allows a user with an AAA account to know the rent for Building A, which is the target real estate, and the calculation basis for the rent. The calculation process is completed with the above processing.

[0106] [effect] System 1 can accurately estimate the appropriate rent for a property. More specifically, server 110's control circuit 112 outputs a prompt P1 to the first LLM 218, instructing the first LLM 218 to calculate the rent for the target property based on target property information I3 for Building A and competing property information I5 for Building B, Building C, and Building X. Furthermore, server 110's control circuit 112 acquires calculation information I9 generated by the first LLM 218, which includes the rent for Building A and the basis for calculating the rent. In this way, system 1 generates calculation information I9, including the rent for Building A, based on competing property information I5 for Building B, Building C, and Building X, which compete with Building A. Therefore, the three competing property information I5 for Building B, Building C, and Building X is reflected in Building A's rent. As a result, system 1 can accurately estimate the appropriate rent for a property.

[0107] In system 1, the first LLM 218 is used to calculate the rent for Building A. Therefore, instructions for calculating the rent for Building A are given by a prompt P1 written in natural language. This allows the user or the operating company of system 1 to include detailed instructions in the prompt P1. As a result, the control circuit 112 of server 110 can accurately set a rent close to that set by a human in a short amount of time.

[0108] In system 1, the control circuit 112 of the server 110 generates a prompt P1 using target real estate information I3, which is real estate information I1 for Building A, and three competing real estate information I5, which are real estate information I1 for Building B, Building C, and Building X. This allows the control circuit 212 of the LLM server 210 to cause the first LLM 218 to generate calculation information using the prompt P1 without processing the prompt P1. In other words, the LLM server 210 does not need to execute a program for processing the prompt P1, and can therefore use a general-purpose LLM.

[0109] In the system 1, the control circuit 112 of the server 110 acquires the target real estate information 13 of Building A based on the target real estate specification information 12, which is the keyword (Building A) input by the user. This allows the control circuit 112 of the server 110 to set the real estate desired by the user as the target real estate.

[0110] In the system 1, the control circuit 112 of the server 110 acquires competing real estate information 15 for X Building based on the competing real estate specification information 14, which is a keyword (X Building) input by the user. This allows the control circuit 112 of the server 110 to set the real estate desired by the user as a competing real estate.

[0111] According to the system 1, even if the user does not accurately remember the details of the target property, the control circuit 112 of the server 110 is likely to be able to identify the target property. More specifically, the control circuit 112 of the server 110 generates first list information I10 including one or more target candidate properties that are candidates for one or more target properties based on target property identification information I2, which is a keyword (Building A) entered by the user. For example, the control circuit 112 of the server 110 extracts real estate information I1, which includes Building A in the property name included in the target property identification information I2, from the real estate information table. In this embodiment, Building A is the target candidate property. Therefore, the control circuit 112 of the server 110 generates first list information I10 including the real estate information I1 of Building A. Then, the control circuit 112 of the server 110 acquires target real estate information I3 of the target candidate property selected by the user from the one or more target candidate properties included in the first list information I10. In this manner, the user can select a target property from the target candidate properties included in the first list information I10. As a result, according to the system 1, even if the user does not remember the exact details of the target property, there is a high possibility that the control circuit 112 of the server 110 can identify the target property.

[0112] In System 1, the three competing real estate information I5 included in Prompt P1 includes the asking rents set for the competing properties B, C, and X, as well as the contracted rents when a deal is concluded for these properties. As a result, the rent for the target property, Building A, reflects the asking rents and contracted rents for Buildings B, C, and X. As a result, System 1 can accurately estimate the fair rent for a property.

[0113] In the system 1, the control circuit 112 of the server 110 causes an AAA account to display the contracted rent associated with the AAA account, and does not cause an account other than the AAA account to display the contracted rent (contract price) associated with the AAA account on the display 20 of the user terminal 10. As a result, the contracted rent entered by a user with an AAA account is not presented to users with accounts other than the AAA account.

[0114] In the system 1, the control circuit 112 of the server 110 extracts one or more pieces of real estate information with closing prices I6 that include closing rents associated with AAA accounts from the three pieces of competing real estate information I5, and does not extract one or more pieces of real estate information with closing prices I6 that include closing rents associated with accounts other than the AAA accounts from the three pieces of competing real estate information I5. As a result, closing rents entered by users with AAA accounts are not presented to users with accounts other than the AAA accounts.

[0115] In the system 1, the control circuit 112 of the server 110 converts the memo information I8 into the format of the real estate information with closing prices I6 and stores it as one or more pieces of real estate information with closing prices I6 in the memory circuit 114. This allows the user to input the memo information I8 in any format without having to conform the format of the memo information I8 to the format of the real estate information with closing prices I6.

[0116] (First Modification) The system 1a according to the first modification will be described below. Fig. 18 is a diagram showing image information I13.

[0117] System 1a differs from System 1 in that the real estate is a residential property and that real estate information I1 includes image information I13 showing an image of the real estate. Image information I13 shown in FIG. 18 shows an image of a room in the real estate. Target real estate information I3 also includes image information I13 showing an image of the target real estate. Competing real estate information I5 also includes image information I13 showing an image of a competing real estate.

[0118] The prompt P1 also includes instructions to extract one or more first features A1 of the target property included in the image information I13 and to cause the first LLM 218 to calculate the price of the target property based on the one or more first features A1. More specifically, the first LLM 218 extracts one or more first features A1 that affect the rent of the target property from the image of the property represented by the image information I13 based on the prompt P1. Such one or more first features A1 are, for example, equipment installed in the target property. In this embodiment, the first LLM 218 extracts a ceiling-mounted air conditioner 500 and an island kitchen 502 as the one or more first features A1. At this time, the first LLM 218 may also extract one or more third features A3 that affect the rent of competing properties from the image of the property supported by the image information I13 included in the competing property information I5. The first LLM 218 then calculates the rent of the target property based on the one or more first features A1 and one or more third features A3. According to the system 1a, the first feature A1 of the property is taken into consideration in calculating the rent, so that the appropriate rent of the property can be estimated with high accuracy.

[0119] (Second Modification) A system 1b according to the second modification will be described below. Fig. 19 is a diagram showing a view of the target real estate.

[0120] System 1b differs from system 1a in that prompt P1 further includes instructions to instruct the first LLM 218 to estimate one or more second features A2 of the view of the target property based on target property information I3 and to calculate the rent (price) of the target property based on the one or more second features A2. In the second variant, target property information I3 includes information about the address, floor, and window orientation of the room. Based on the address, floor, and window orientation of the room, the first LLM 218 generates view information I14 showing the view of the target property, as shown in FIG. 19. Based on prompt P1, the first LLM 218 then extracts one or more second features A2 that affect the rent of the target property from the view of the target property indicated by view information I14. Examples of such one or more second features A2 include Tokyo Tower 504. Based on the one or more second features A2, the first LLM 218 calculates the rent of the target property. According to such a system 1b, the second feature A2 of the property is taken into consideration in calculating the rent, so that the appropriate rent of the property can be estimated with high accuracy.

[0121] (Third Modification) A system 1c according to the third modification will be described below. Fig. 20 is a diagram showing image information I13.

[0122] System 1c differs from System 1a in that prompt P1 further includes instructions to cause the first LLM 218 to identify facilities that can be added or modified to the target property based on the target property information I3 and calculate the rent (price) of the target property if the facilities are added or modified, as well as the costs required to add or modify the facilities. More specifically, based on prompt P1, the first LLM 218 extracts one or more facilities of the target property from the image of the property represented by image information I13. The first LLM 218 then extracts from the extracted one or more facilities facilities that, if added or modified, would increase the price of the target property. In this embodiment, the first LLM 218 determines that changing the kitchen 506 to an island kitchen will increase the rent of the target property. Therefore, the first LLM 218 identifies the kitchen 506 as a facility that can be added or modified to the target property. Furthermore, the first LLM 218 calculates the costs required to change the kitchen 506 to an island kitchen. System 1c as described above can increase the value of the target property.

[0123] (Other embodiments) The system according to the present invention is not limited to the systems 1, 1a to 1c, and can be modified within the scope of the invention. Furthermore, the configurations of the systems 1, 1a to 1c may be combined in any manner.

[0124] The prompt P1 may include an instruction to cause the first LLM 218 to generate a description of the target property based on the target property information I3. The description of the target property is displayed in the calculation result image shown in FIG.

[0125] The prompt P1 may include an instruction to have the description describe advantages of the target property over one or more competing properties based on the target property information I3 and one or more competing property information I5. The description of the target property is displayed in the calculation result image shown in FIG.

[0126] The target real estate may be an unbuilt property. In this case, prompt P1 includes an instruction to cause the first LLM 218 to calculate the rent (price) of the target real estate if a virtual building is constructed on the target real estate. This allows the user to know the rent of the unbuilt property. Therefore, the user can estimate the profits if the property is constructed.

[0127] In the system 1, the control circuit 112 of the server 110 extracts one or more pieces of real estate information I6 with closing prices that include closing rents associated with AAA accounts from the three pieces of competing real estate information I5, but does not extract one or more pieces of real estate information I6 with closing prices that include closing rents associated with accounts other than the AAA accounts from the three pieces of competing real estate information I5. Therefore, in the system 1, the control circuit 112 of the server 110 does not transmit one or more pieces of real estate information I6 with closing prices that include closing rents associated with accounts other than the AAA accounts to the user terminal 10 via the network interface 116. However, the control circuit 112 of the server 110 may transmit one or more pieces of real estate information I6 with closing prices that exclude the closing rents to the user terminal 10 via the network interface 116. In this case, the control circuit 112 of the server 110 extracts one or more competing real estate information I5 from the multiple real estate information I1 based on the competing real estate identification information I4, and extracts one or more real estate information I6 with closing prices that includes the closing rent (closing price) associated with an account other than the AAA account (first account), excluding the closing price.

[0128] The real estate may be for sale rather than for rent. If the real estate is for sale, the price is the sales price rather than the rent.

[0129] The control circuit 112 of the server 110 may convert the memo information 18 into the format of the real estate information with closing prices 16 without using the second LLM 220. In this case, the control circuit 112 of the server 110 may execute a program for converting the format of the memo information 18.

[0130] The server 110 may specify, as a competing candidate property, a property having similar content in the item K1 to the content in the item K1 other than the address and the standard floor area of ​​Building A.

[0131] It should be noted that the LLM server on which the second LLM 220 is provided may be different from the LLM server on which the first LLM 218 is provided. In other words, the LLM server 210 may include multiple servers.

[0132] Each of the first, second, and third listings I10, I11, and I12 may not include all of the items K1 in the real estate listing I1. For example, each of the first, second, and third listings I10, I11, and I12 may not include the contracted area, contracted rent, and company details in the real estate listing I1.

[0133] The memo information I8 may be voice data including the name of the property and the contract price.

[0134] The competing real estate information I5 may include either the set price set for the competing real estate or the contract price at the time of the contract for the competing real estate.

[0135] The system 1 may further include a machine learning server equipped with a machine learning model. The machine learning server includes a machine learning model trained using multiple pieces of real estate information I1 as training data. Specifically, the machine learning model learns the relationship between the details of the real estate contained in the real estate information I1 and the rent of the real estate. This allows the machine learning model to output the amount that each feature of the real estate contributes to the rent of the real estate. The feature of the real estate is an item included in the details of the real estate. Specifically, the machine learning model can output, for example, that the rent of a real estate with an OA floor is 1,000 yen per tsubo higher than the rent of a real estate without an OA floor.

[0136] Before generating prompt P1 in step S125, server 110 acquires the amount that each feature of the property contributes to the rent of the property from the machine learning server. Then, server 110 writes in prompt P1 that the amount that each feature of the property contributes to the rent of the property will be taken into consideration when calculating the rent of the target property.

[0137] In the above, instead of the machine learning server, a server that holds data on the items included in the property details and the amount that each item contributes to the property's rent may be used.

[0138] The user may be able to specify the items to be emphasized when calculating the rent for the target real estate. Specifically, the user operates the user terminal 10 to select the items to be emphasized when calculating the rent for the target real estate from multiple items such as property specifications, location, and transaction history. The property specifications are items related to the facilities in the property details. The location is items such as the property's address and the nearest station. The transaction history is items such as the contracted area and the contracted rent. The user terminal 10 transmits selected item information indicating the items selected by the user to the server 110. Based on the selected item information, the server 110 writes the items to be emphasized when calculating the rent for the target real estate in the prompt P1. The user may select the items to be emphasized when calculating the rent for the target real estate by allocating parameters to multiple items such as property specifications, location, and transaction history.

[0139] In addition, a user may wish to set a rent for a target property that is higher than the market rent (a rent calculated based on market information on offers and successful transactions, etc.) in consideration of their company's management plan, company portfolio, etc. In this case, the server 110 compares the details of the target property with the details of the target property and extracts rare items from the details of the target property. The server 110 may then write in the prompt P1 an instruction to calculate the rent for the target property higher than the market rent based on the rare items. In this case, the server 110 may write in the prompt P1 an instruction to output the basis for calculating the rent for the target property higher than the market rent. The server 110 may also obtain information indicating the company's budget, portfolio, and user intentions from the user terminal 10, and write in the prompt P1 an instruction to use the information indicating the company's budget, portfolio, and user intentions to calculate the rent for the target property. The user's intention may be, for example, whether or not they want to set the rent of the target property higher than the market rent, whether or not they want to set the rent of the target property equal to the market rent, or whether or not they want to set the rent of the target property lower than the market rent.

[0140] At least some of the processing described in this specification as being performed by a specific device may be performed by any information processing device.

[0141] The series of processes performed by each device described herein may be implemented using software, hardware, or a combination of software and hardware. The software programs may be stored in advance, for example, on a recording medium (more specifically, a non-transitory computer-readable storage medium) internal or external to each device. Each program is then loaded into RAM and executed by a processing circuit such as a CPU when executed by a computer controlling each device described herein. The recording medium may be, for example, a magnetic disk, an optical disk, a magneto-optical disk, or a flash memory. The computer program may also be distributed, for example, via a network, without using a recording medium. The computer may be, for example, an application-specific integrated circuit (ASIC), a general-purpose processor that executes functions by loading a software program, or a computer on a server used in cloud computing. The series of processes performed by each device described herein may be centrally processed by a single computer, or may be distributed across multiple computers. Furthermore, in each of the above embodiments, two or more communication means present in one device may be physically implemented on a single medium.

[0142] Furthermore, the processes described herein using flowcharts or sequence diagrams do not necessarily have to be performed in the order shown. Some process steps may be performed in parallel. Furthermore, additional process steps may be employed, and some process steps may be omitted.

[0143] The present invention is not limited to the above-described embodiment, and various modifications are possible, and it goes without saying that these modifications are also included within the scope of the present invention. [Explanation of symbols]

[0144] 1, 1a to 1c: System 10: User terminal 12: Control circuit 14:Memory circuit 16: Network interface 18: Graphics processing unit 20: Display 26:Operation unit 110: Server 112: Control circuit 114:Memory circuit 116: Network interface 120:Means for obtaining target real estate information 122:Method of obtaining competitive real estate information 124: Prompt output means 125: Step 126: Calculation information acquisition means 128: Prompt generation means 130: Means for obtaining target real estate specific information 132:Means for obtaining specific information on competing real estate 134: List information generating means 136:Memory information acquisition means 138: Storage control means 140: Display control means 212: Control circuit 214:Memory circuit 216: Network interface 300: Memo information display section 302: Memo information input section 304:Register button 400, 402: Button 500:Air conditioner 502: Island kitchen 504: Tokyo Tower 506: Kitchen A1: First feature A2: Second feature A3: Third feature B1~B5: Box I1: Real Estate Information I2: Target real estate specific information I3: Target real estate information I4: Competitive property specific information I5: Competitive real estate information I6: Real estate information with closing prices I7: Conversion memo information I8:Memo information I9: Calculation information I10: First list information I11: Second list information I12: Third List Information I13: Image information I14: View Information K1,K2:Item 218: 1st LLM 220: 2nd LLM P1, P2, P11, P12: Prompt PG1: Program

Claims

1. A system comprising one or more computers, The one or more computer control circuits include: Obtain target real estate information including details of the target real estate for which the price is to be calculated; acquiring competing real estate identification information used to identify one or more competing real estate properties that compete with the target real estate; obtaining one or more pieces of competitive real estate information including prices and details of the one or more competitive real estate properties based on the competitive real estate identification information; outputting a prompt to a Large Language Model (LLM) including an instruction to cause the LLM to calculate a price of the target real estate based on the target real estate information and the one or more competing real estate information; Obtaining calculation information including the price of the target real estate generated by the LLM; system.

2. A system comprising one or more computers, The one or more computer control circuits include: Obtain target real estate identification information used to identify the target real estate for which the price is to be calculated, generating list information including one or more candidate real estate properties that are candidates for the target real estate based on the target real estate identification information; acquiring target real estate information including details of the target real estate, the target real estate information being of a target candidate real estate selected by a user from one or more target candidate real estates included in the list information; obtaining one or more competitive property information items including prices and details of one or more competitive properties that compete with the target property; outputting a prompt to a Large Language Model (LLM) including an instruction to cause the LLM to calculate a price of the target real estate based on the target real estate information and the one or more competing real estate information; Obtain calculation information including the price of the target real estate generated by the LLM; In the process of acquiring the target real estate information, the target real estate information is acquired based on the target real estate identification information. system.

3. A system comprising one or more computers, The one or more computer control circuits include: Obtain target real estate information including details of the target real estate for which the price is to be calculated; obtaining one or more competitive property information items including prices and details of one or more competitive properties that compete with the target property; outputting a prompt to a Large Language Model (LLM) including an instruction to cause the LLM to calculate a price of the target real estate based on the target real estate information and the one or more competing real estate information; Obtain calculation information including the price of the target real estate generated by the LLM; The competing real estate information includes the set price set for the competing real estate and / or the closing price when the competing real estate is sold. system.

4. The one or more computer control circuits include: generating the prompt using the target property information and the one or more competing property information; In the process of outputting the prompt, the prompt generated in the process of generating the prompt is output to the LLM. A system according to any one of claims 1 to 3.

5. The one or more computer control circuits include: acquiring competitive real estate identification information used to identify the one or more competitive real estate properties; In the process of acquiring one or more pieces of competing real estate information, the one or more pieces of competing real estate information are acquired based on the competing real estate identification information, the one or more computer memory circuits store a plurality of property information items including property prices and descriptions; The plurality of pieces of real estate information include one or more pieces of real estate information with closing prices, the one or more real estate information with closing prices includes the closing prices associated with one or more accounts; The one or more accounts include a first account; The one or more computer control circuits include: In the process of acquiring the competing real estate identification information, the competing real estate identification information input by the user having the first account is acquired, In the process of acquiring the competing real estate information, the one or more competing real estate information are extracted from the plurality of real estate information based on the competing real estate identification information; causing the first account to display the contract price associated with the first account, and not allowing accounts other than the first account to display the contract price associated with the first account; The system of claim 3.

6. The one or more computer control circuits include: acquiring competitive real estate identification information used to identify the one or more competitive real estate properties; In the process of acquiring one or more pieces of competing real estate information, the one or more pieces of competing real estate information are acquired based on the competing real estate identification information, the one or more computer memory circuits store a plurality of property information items including property prices and descriptions; The plurality of pieces of real estate information include one or more pieces of real estate information with closing prices, the one or more real estate information with closing prices includes the closing prices associated with one or more accounts; The one or more accounts include a first account; The one or more computer control circuits include: In the process of acquiring the competing real estate identification information, the competing real estate identification information input by the user having the first account is acquired, In the process of acquiring competitive real estate information, acquiring the one or more pieces of competing real estate information from the plurality of pieces of real estate information based on the competing real estate identification information, and extracting one or more pieces of real estate information with a closing price including the closing price associated with the first account from the one or more pieces of competing real estate information, while not extracting one or more pieces of real estate information with a closing price including the closing price associated with an account other than the first account from the one or more pieces of competing real estate information; Alternatively, the one or more pieces of competing real estate information are extracted from the plurality of pieces of real estate information based on the competing real estate identification information, and one or more pieces of real estate information with closing prices, including the closing prices, associated with accounts other than the first account are extracted in a state where the closing prices are excluded. The system of claim 3.

7. The memo information is text data or audio data including the name of the property and the contract price, The one or more computer control circuits include: Acquire the memo information input by the user having the first account; storing the memo information converted into the format of real estate information with closing prices in a storage circuit as real estate information with closing prices; The system of claim 6.

8. the target real estate information includes image information showing an image of the target real estate, The prompt further includes instructions to extract one or more first characteristics of the target property included in the image information and cause the LLM to calculate a price for the target property based on the one or more first characteristics. A system according to any one of claims 1 to 3.

9. the prompt further includes instructions to cause the LLM to estimate one or more second characteristics of the view of the subject property based on the subject property information, and to calculate a price for the subject property based on the one or more second characteristics. A system according to any one of claims 1 to 3.

10. The prompt further includes instructions to cause the LLM to identify facilities that can be added to or modified in the target property based on the target property information, and to calculate the price of the target property when the facilities are added or modified, and the costs required to add or modify the facilities. A system according to any one of claims 1 to 3.

11. The facilities are facilities whose addition or modification will increase the price of the target real estate, The system of claim 10.

12. the prompt includes instructions to cause the LLM to generate a description of the subject property based on the subject property information. A system according to any one of claims 1 to 3.

13. The prompt includes an instruction to cause the description to describe advantages of the target property over the one or more competing properties based on the target property information and the one or more competing property information. The system of claim 12.

14. The target real estate is an unbuilt property, The prompt includes an instruction to cause the LLM to calculate the value of the target property if a virtual building is constructed on the target property. A system according to any one of claims 1 to 3.

15. one or more computer control circuits, Obtain target real estate information including details of the target real estate for which the price is to be calculated; acquiring competing real estate identification information used to identify one or more competing real estate properties that compete with the target real estate; obtaining one or more pieces of competitive real estate information including prices and details of the one or more competitive real estate based on the competitive real estate identification information; outputting a prompt to a Large Language Model (LLM) including an instruction to cause the LLM to calculate a price of the target real estate based on the target real estate information and the one or more competing real estate information; obtaining calculation information including the price of the target real estate generated by the LLM; program.

16. one or more computer control circuits, Obtain target real estate identification information used to identify the target real estate for which the price is to be calculated; generating list information including one or more candidate real estate properties that are candidates for the target real estate based on the target real estate identification information; acquiring target real estate information including details of the target real estate, the target real estate information being of a target candidate real estate selected by a user from one or more target candidate real estates included in the list information; obtaining one or more competitive property information items including prices and details of one or more competitive properties that compete with the target property; outputting a prompt to a Large Language Model (LLM) including an instruction to cause the LLM to calculate a price of the target real estate based on the target real estate information and the one or more competing real estate information; Obtaining calculation information including the price of the target real estate generated by the LLM; In the process of acquiring the target real estate information, the target real estate information is acquired based on the target real estate identification information. program.

17. one or more computer control circuits, Obtain target real estate information including details of the target real estate for which the price is to be calculated; obtaining one or more competitive property information items including prices and details of one or more competitive properties that compete with the target property; outputting a prompt to a Large Language Model (LLM) including an instruction to cause the LLM to calculate a price of the target real estate based on the target real estate information and the one or more competing real estate information; Obtaining calculation information including the price of the target real estate generated by the LLM; The competing real estate information includes the set price set for the competing real estate and / or the closing price when the competing real estate is sold. program.

18. The one or more computer control circuits Obtain target real estate information including details of the target real estate for which the price is to be calculated; acquiring competing real estate identification information used to identify one or more competing real estate properties that compete with the target real estate; obtaining one or more pieces of competitive real estate information including prices and details of the one or more competitive real estate properties based on the competitive real estate identification information; outputting a prompt to a Large Language Model (LLM) including an instruction to cause the LLM to calculate a price of the target real estate based on the target real estate information and the one or more competing real estate information; Obtaining calculation information including the price of the target real estate generated by the LLM; Information processing methods.

19. The one or more computer control circuits Obtain target real estate identification information used to identify the target real estate for which the price is to be calculated, generating list information including one or more candidate real estate properties that are candidates for the target real estate based on the target real estate identification information; acquiring target real estate information including details of the target real estate, the target real estate information being of a target candidate real estate selected by a user from one or more target candidate real estates included in the list information; obtaining one or more competitive property information items including prices and details of one or more competitive properties that compete with the target property; outputting a prompt to a Large Language Model (LLM) including an instruction to cause the LLM to calculate a price of the target real estate based on the target real estate information and the one or more competing real estate information; Obtain calculation information including the price of the target real estate generated by the LLM; In the process of acquiring the target real estate information, the target real estate information is acquired based on the target real estate identification information. Information processing methods.

20. The one or more computer control circuits Obtain target real estate information including details of the target real estate for which the price is to be calculated; obtaining one or more competitive property information items including prices and details of one or more competitive properties that compete with the target property; outputting a prompt to a Large Language Model (LLM) including an instruction to cause the LLM to calculate a price of the target real estate based on the target real estate information and the one or more competing real estate information; Obtain calculation information including the price of the target real estate generated by the LLM; The competing real estate information includes the set price set for the competing real estate and / or the closing price when the competing real estate is sold. Information processing methods.

Images