Information processing system, information processing device and program

Abstract

To provide an information processing system capable of calculating assessed price information of an assessment object property with further high accuracy.SOLUTION: An information processing system 10 includes a similarity calculation unit 435 and an assessment information calculation unit 438. The similarity calculation unit 435 calculates a similarity degree of each of a plurality of similar properties to an assessment object property. The assessment information calculation unit 438 selects a similar property for assessment to be used for calculating assessed price information of the assessment object property from the plurality of similar properties based on each similarity degree of the plurality of similar properties, and calculates the assessed price information of the assessment object property based on predicted assessed price information of the similar property for assessment.SELECTED DRAWING: Figure 2

Description

[Technical field]

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

[0002] A conventional information processing device is described in Patent Document 1 below. This information processing device calculates a score using two or more attribute values ​​related to a real estate property and outputs the score. The attribute values ​​include the contract price, the assessed price, etc. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2020-91567 A Summary of the Invention [Problem to be solved by the invention]

[0004] The information processing device described in Patent Document 1 has room for improvement in terms of calculation of the assessed price.

[0005] An object of the present invention is to provide an information processing system, an information processing device, and a program capable of calculating the appraisal price information of a real estate to be appraised with higher accuracy. [Means for solving the problem]

[0006] The information processing system for solving the above problem includes a first appraisal information prediction unit, a similarity calculation unit, and an appraisal information calculation unit. The first appraisal information prediction unit calculates, for each of the similar real estate data of the multiple similar real estates, predicted appraisal price information, which is predicted price information at the time of appraisal of the appraisal target real estate predicted from the similar real estate data of each of the multiple similar real estates, based on appraisal target data of the appraisal target real estate and similar real estate data of each of the multiple similar real estates, including information similar to information included in the appraisal target data. The similarity calculation unit calculates the similarity of each of the multiple similar real estates to the appraisal target real estate. The appraisal information calculation unit selects an appraisal similar real estate to be used in calculating appraisal price information of the appraisal target real estate from among the multiple similar real estates, based on the similarity of each of the multiple similar real estates, and calculates appraisal price information of the appraisal target real estate based on the predicted appraisal price information of the appraisal similar real estate.

[0007] The information processing device for solving the above problem includes a first appraisal information prediction unit, a similarity calculation unit, and an appraisal information calculation unit. The first appraisal information prediction unit calculates, for each of the similar real estate data of the multiple similar real estates, predicted appraisal price information, which is predicted price information at the time of appraisal of the appraisal target real estate predicted from the similar real estate data of each of the multiple similar real estates, based on appraisal target data of the appraisal target real estate and similar real estate data of each of the multiple similar real estates, including information similar to information included in the appraisal target data. The similarity calculation unit calculates the similarity of each of the multiple similar real estates to the appraisal target real estate. The appraisal information calculation unit selects an appraisal similar real estate to be used in calculating appraisal price information of the appraisal target real estate from among the multiple similar real estates, based on the similarity of each of the multiple similar real estates, and calculates appraisal price information of the appraisal target real estate based on the predicted appraisal price information of the appraisal similar real estate.

[0008] To solve the above problem, a computer is caused to calculate, for each of the similar real estate data of the multiple similar real estates, predicted appraisal price information, which is predicted price information at the time of appraisal of the real estate to be appraised predicted from the similar real estate data of each of the multiple similar real estates, based on the appraisal data of the real estate to be appraised and similar real estate data of each of the multiple similar real estates, which includes information similar to the information contained in the appraisal data, calculate the similarity of each of the multiple similar real estates to the real estate to be appraised, select from the multiple similar real estates a similar real estate to be used for calculating the appraisal price information of the real estate to be appraised based on the similarity of each of the multiple similar real estates, and calculate appraisal price information of the real estate to be appraised based on the predicted appraisal price information of the similar real estate to be appraised.

[0009] With this configuration, the appraisal price information of the real estate to be appraised can be calculated based on data on similar real estate for appraisal that is highly similar to the real estate to be appraised, making it possible to calculate the appraisal price information of the real estate to be appraised with greater accuracy. [Brief description of the drawings]

[0010] [Figure 1] FIG. 1 is a block diagram showing a schematic configuration of an information processing system according to a first embodiment. [Diagram 2] FIG. 2 is a block diagram showing a schematic configuration of an assessment server device according to the first embodiment. [Diagram 3] 4 is a table showing an example of information included in assessment target data according to the first embodiment. [Figure 4] FIG. 2 is a block diagram showing a partial configuration of a control unit according to the first embodiment. [Diagram 5] 5 is a table showing an example of information included in neighborhood case data according to the first embodiment. [Figure 6] FIG. 2 is a block diagram showing a partial configuration of a control unit according to the first embodiment. [Figure 7] 4 is a table showing an example of information included in similar case data according to the first embodiment. [Figure 8] 6 is a table showing an example of information included in similar case data after processing by the preprocessing unit according to the first embodiment. [Figure 9]4 is a table showing an example of information included in assessment target data according to the first embodiment. [Figure 10] 13A to 13C are tables showing examples of data on the provisional appraised price per square meter of the real estate to be appraised in the first embodiment, the predicted closing price per square meter of the real estate to be appraised, and the predicted closing price per square meter of similar real estate. [Figure 11] FIG. 2 is a block diagram showing a partial configuration of a control unit according to the first embodiment. [Figure 12] 6 is a table showing an example of information included in the similarity according to the first embodiment. [Figure 13] FIG. 2 is a block diagram showing a partial configuration of a control unit according to the first embodiment. [Figure 14] 13A and 13B are tables showing examples of data for the first estimated assessed price per square meter and the second estimated price per square meter in the first embodiment. [Figure 15] 4 is a sequence chart showing an example of the operation of the information processing system according to the first embodiment. [Figure 16] FIG. 2 is a block diagram showing the hardware configuration of a computer according to the first embodiment. [Figure 17] FIG. 11 is a block diagram showing the configuration of a reliability calculation unit of a control unit according to a second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, an embodiment of an information processing device, an information processing system, and a program will be described with reference to the drawings. In order to facilitate understanding of the description, the same components in each drawing are denoted by the same reference numerals as much as possible, and duplicated descriptions will be omitted.

[0012] First Embodiment First, an information processing system according to a first embodiment will be described.

[0013] (Outline of information processing system configuration) FIG. 1 shows a schematic configuration of an information processing system 10 of this embodiment. The information processing system 10 of this embodiment is used by, for example, an employee of a real estate company. When the employee of the real estate company inputs information on the location, transportation, etc. of a specific real estate, this information processing system 10 calculates the appraisal price of the specific real estate using information on other real estate. This allows the employee of the real estate company to present the appraisal price information calculated by the information processing system 10 to, for example, a customer who is considering selling the specific real estate. Note that the following will be described using an example where the real estate to be appraised is an apartment building.

[0014] As shown in FIG. 1, the information processing system 10 includes a terminal device 20, an information providing server device 30, and an assessment server device 40.

[0015] The terminal device 20 is used by users such as employees of a real estate company. The terminal device 20 can be used to input various information such as the location and transportation of the real estate to be appraised, and can display information on the appraised price of the real estate to be appraised. The terminal device 20 is a personal computer or the like. Personal computers include, for example, desktop computers and laptop computers.

[0016] The information providing server device 30 is a device that provides station information, location information, etc. Station information that the information providing server device 30 can provide includes, for example, information on the route names and station names of JR conventional lines, information on the route names and station names of private railways, and information on the number of passengers boarding and alighting at each station. In addition, location information that the information providing server device 30 can provide includes information on the latitude and longitude of a specific location, information on the publicly announced land prices of a specific location, etc. The information providing server device 30 is configured, for example, as a cloud system.

[0017] The assessment server device 40 is a device that calculates the assessed price of real estate based on the information input to the terminal device 20 and the station information, location information, etc. provided by the information providing server device 30. The assessment server device 40 is configured as, for example, a cloud system. In this embodiment, the assessment server device 40 corresponds to an information processing device.

[0018] The terminal device 20, the information providing server device 30, and the assessment server device 40 are communicably connected to each other via a communication network 11. The communication network 11 may be either a wired network or a wireless network.

[0019] (Configuration of the assessment server device) Next, a detailed description will be given of the configuration of the assessment server device 40. As shown in FIG.

[0020] The communication unit 41 is a part that transmits and receives various information to and from the terminal device 20 and the information providing server device 30 via the network 11. The information that the communication unit 41 receives from the terminal device 20 via the network 11 includes, for example, the appraisal data Da of the appraised real estate. The information that the communication unit 41 transmits to the terminal device 20 via the network 11 includes, for example, information on the appraisal price calculated by the appraisal server device 40. The information that the communication unit 41 receives from the information providing server device 30 via the network 11 includes station information and location information of any point, etc.

[0021] The storage unit 42 stores information such as various programs for operating the assessment server device 40. The storage unit 42 stores a real estate case database 420. The real estate case database 420 is a database of data on various real estate contract cases. The data on various real estate contract cases includes information such as property information, contract price, contract date, and contract price per tsubo. The property information includes information such as the building name, location, transportation, construction date, total number of units, number of above-ground floors, and exclusive area. Hereinafter, the data on various real estate contract cases stored in the real estate case database 420 is referred to as "real estate case data." The data on various real estate contract cases stored in the real estate case database 420 includes data on real estate contract cases manually input by employees of a real estate company, data on real estate contract cases handled in the past by a real estate company, and the like.

[0022] The control unit 43 controls the appraisal server device 40. The control unit 43 includes, as functional components realized by executing the programs stored in the storage unit 42, an appraisal information acquisition unit 430, a nearby case acquisition unit 431, a pre-processing unit 432, a provisional information calculation unit 433, a contract information prediction unit 434, a similarity calculation unit 435, a first appraisal information prediction unit 436, a second appraisal information prediction unit 437, and an appraisal information calculation unit 438.

[0023] (Assessment Information Acquisition Department) The assessment information acquisition unit 430 acquires the assessment target data Da of the real estate to be assessed from the terminal device 20 via the communication unit 41. The assessment target data Da includes, for example, the name, address, transportation, construction date, total number of units, number of above-ground floors, and exclusive area of ​​the building as shown in FIG.

[0024] (Neighborhood Case Collection Department) The neighboring case acquisition unit 431 acquires data on a number of other real estate properties that exist in the neighborhood of the real estate to be appraised.

[0025] For example, as shown in FIG. 4, the neighboring case acquisition unit 431 acquires the appraisal target data Da from the appraisal information acquisition unit 430, and then acquires information on the line name and station name existing near the appraisal target real estate based on the location of the appraisal target real estate included in the appraisal target data Da. When the line and station existing near the appraisal target real estate are respectively the "target line" and the "first target station", the neighboring case acquisition unit 431 extracts two stations existing before and after the first target station on the target line as the "second target station" and the "third target station", respectively, by using map information that can be acquired from the information providing server device 30, for example. Then, the neighboring case acquisition unit 431 identifies, from the multiple real estate case data stored in the real estate case database 420, real estate whose nearest station is any of the first target station, the second target station, and the third target station as neighboring real estate, and lists the data of the identified neighboring real estate. As a result, the neighboring case example acquisition unit 431 acquires data on a number of other real estate properties that exist in the neighborhood of the real estate to be assessed, for example, data Db(1) to Db(m) as shown in Fig. 5. Hereinafter, the data shown in Fig. 5 will be referred to as "neighboring case example data Db(1) to Db(m)." "m" is an integer of 2 or greater.

[0026] The neighboring case acquisition unit 431 may specify, from among the multiple real estate case data stored in the real estate case database 420, only data on real estate whose nearest station is the first target station as the neighboring real estate.

[0027] The neighboring case acquisition unit 431 may change the target station depending on the number of target lines existing near the real estate to be assessed. For example, when the number of target lines existing near the real estate to be assessed is one, the neighboring case acquisition unit 431 may specify, as the neighboring real estate, from the multiple real estate case data stored in the real estate case database 420, real estate whose nearest station is any one of the first target station, the second target station, and the third target station. When the number of target lines existing near the real estate to be assessed is two or more, the neighboring case acquisition unit 431 may specify, as the neighboring real estate, from the multiple real estate case data stored in the real estate case database 420, only real estate whose nearest station is the first target station.

[0028] (Preprocessing section) The preprocessing unit 432 performs a process of extracting data on similar real estate to be used in calculating the appraisal price of the real estate to be appraised, from the neighboring case data Db(1) to Db(m) acquired by the neighboring case acquisition unit 431.

[0029] 6, the pre-processing unit 432 acquires the appraisal target data Da from the appraisal information acquisition unit 430, and acquires the neighboring case data Db(1) to Db(m) from the neighboring case acquisition unit 431. The pre-processing unit 432 then compares the data of the appraisal target real estate contained in the appraisal target data Da with the data of multiple neighboring real estate contained in the neighboring case data Db(1) to Db(m), thereby extracting data of neighboring real estate that satisfies, for example, any of the following conditions (a1) to (a5) as an exclusion target.

[0030] (a1) The number of rooms in the floor plan differs from the number of rooms in the floor plan of the real estate to be appraised by more than a predetermined number. The predetermined number is set to, for example, "2 rooms." (a2) The date of the transaction is more than a specified number of years before the appraisal date of the real estate being appraised. The specified number of years is set to, for example, "1 year." (a3) The exclusive area differs from the exclusive area of ​​the real estate being assessed by a specified area or more. (a4) The number of units differs from the number of units in the real estate being assessed by more than a specified number. (a5) The hierarchical structure differs from the hierarchical structure of the real estate being assessed by more than a specified number.

[0031] In addition, whether or not to exclude a property may be determined based on whether or not the property is currently being rented, whether the transaction type is seller or agent, and the like.

[0032] The preprocessing unit 432 excludes the data of the neighboring real estate thus identified as the exclusion target from the neighboring case data Db(1)-Db(m). For example, when apartments A, B, E, and K are the exclusion targets in the neighboring case data Db(1)-Db(m) shown in FIG. 5, the preprocessing unit 432 creates the similar case data Dc(1)-Dc(n) as shown in FIG. 7 by excluding the data Db(1), Db(2), Db(5), and Db(11) of apartments A, B, E, and K, respectively, from the neighboring case data Db(1)-Db(m). By such processing, it is possible to create the similar case data Dc(1)-Dc(n) including only the data of similar real estate having information similar to the assessed real estate. Here, "n" is an integer of 2 or more that satisfies "n≦m", and if there are no neighboring real estate to be excluded, "n=m".

[0033] The pre-processing unit 432 creates the similar case data Dc(1) to Dc(n) in this manner, and also performs processing to link the station information and location information to the data of each similar real estate contained in the similar case data Dc(1) to Dc(n).

[0034] For example, the pre-processing unit 432 performs a process of linking station information obtainable from the information providing server device 30 to data of each similar real estate. The station information includes a JR flag, an access flag, and the number of passengers boarding and alighting. The JR flag is information indicating whether the nearest line to the similar real estate is a JR line. The access flag is information indicating whether the nearest station to the similar real estate is accessible between a JR line and a private railway.

[0035] The pre-processing unit 432 also performs a process of linking the location information obtainable from the information providing server device 30 to the data of each similar real estate. The location information includes latitude and longitude information and publicly announced land price information of the similar real estate. The latitude and longitude information is information indicating the latitude and longitude of the location of the similar real estate. The publicly announced land price information is information indicating the publicly announced land price of the location of the similar real estate. In this embodiment, the location information corresponds to position information.

[0036] Through the above processing, the preprocessing unit 432 creates similar real estate data Dd(1) to Dd(n) as shown in FIG.

[0037] The pre-processing unit 432 also performs processing to link the station information and location information to the appraisal target data Da in a similar manner, thereby creating appraisal target data De as shown in Fig. 9. Note that the information contained in the similar real estate data Dd(1) to Dd(n) described above is similar to the information contained in the appraisal target data De.

[0038] (Temporary information calculation unit) The provisional information calculation unit 433 calculates a provisional assessed price per square meter Pa of the real estate to be appraised from the appraisal target data De created by the pre-processing unit 432, as shown in FIG.

[0039] Specifically, the provisional information calculation unit 433 has an AI (Artificial Intelligence) model 433a. The AI ​​model 433a is a machine learning model capable of deriving a provisional assessed price per tsubo Pa of the assessed real estate using the assessed data De shown in FIG. 9 as input data. The AI ​​model 433a is constructed by learning the relationship between the location, transportation, construction date, total number of units, number of floors above ground, exclusive area, station information, location information, etc. of real estate previously assessed by a real estate company, and the price per tsubo assessed by an employee for the real estate. The station information includes information such as a JR flag, an access flag, and the number of passengers. The location information includes latitude and longitude information and publicly announced land price information. The provisional information calculation unit 433 uses the AI ​​model 433a to calculate a provisional assessed price per tsubo Pa of the assessed real estate from the assessed data De shown in FIG. 9, for example, as shown in FIG. 10(A).

[0040] (Deal Information Prediction Department) As shown in FIG. 6, the transaction information prediction unit 434 calculates a predicted transaction price per square meter Pb of the real estate to be appraised based on the appraisal target data De created by the pre-processing unit 432, and also calculates predicted transaction prices per square meter Pc(1) to Pc(n) of multiple similar real estate based on the similar real estate data Dd(1) to Dd(n) created by the pre-processing unit 432.

[0041] Specifically, the transaction information prediction unit 434 has an AI model 434a. The AI ​​model 434a is a machine learning model capable of deriving the predicted transaction price per tsubo Pb of the appraisal target real estate using the appraisal target data De shown in FIG. 9 as input data, and capable of deriving the predicted transaction price per tsubo Pc(1)-Pc(n) of similar real estate using the similar real estate data Dd(1)-Dd(n) shown in FIG. 8 as input data. The AI ​​model 434a is constructed by, for example, acquiring information on transaction cases of real estate nationwide or in a specified region from a specified server device, and learning the acquired information on transaction cases of real estate nationwide or in a specified region. The AI ​​model 434a is constructed by learning the relationship between the location, transportation, construction date, total number of units, number of floors above ground, exclusive area, station information, location information, etc. of the real estate of the transaction case and the actual transaction price per tsubo of the real estate. The station information includes information such as a JR flag, an access flag, and the number of passengers. The location information includes latitude and longitude information and publicly announced land price information. The learning data includes data on past sales cases. The sales information prediction unit 434 uses the AI ​​model 434a to calculate the predicted sales price per tsubo Pb of the appraisal target real estate as shown in FIG. 10(B) from the appraisal target data De shown in FIG. 9. The sales information prediction unit 434 also uses the AI ​​model 434a to calculate the predicted sales price per tsubo Pc(1)-Pc(n) of each similar real estate as shown in FIG. 10(C) from the similar real estate data Dd(1)-Dd(n) shown in FIG. 8.

[0042] As shown in Fig. 8, the similar real estate data Dd(1) to Dd(n) includes information on the actual contract price per tsubo, but the predicted contract prices Pc(1) to Pc(n) of similar real estate are calculated by the AI ​​model 434a based on information such as the location and transportation of real estate in successful transactions, and are not based on the actual contract price per tsubo. Therefore, the predicted contract prices Pc(1) to Pc(n) of similar real estate may differ from the actual contract price per tsubo.

[0043] (Similarity calculation unit) The similarity calculation unit 435 calculates the similarities Sd(1) to Sd(n) of a plurality of similar properties to the property to be appraised. In this embodiment, the similarities Sd(1) to Sd(n) are values ​​set in the range from "0" to "1", with the closer to "0" the lower the similarity, and the closer to "1" the higher the similarity.

[0044] Specifically, as shown in Fig. 11, the similarity calculation unit 435 obtains the appraisal target data De and the similar real estate data Dd(1) to Dd(n) from the pre-processing unit 432. In addition, the similarity calculation unit 435 obtains information on the provisional appraisal price per square meter Pa of the appraisal target real estate (provisional appraisal price information of the appraisal target real estate) from the provisional information calculation unit 433, and obtains information on the predicted contract price per square meter Pb of the appraisal target real estate (predicted contract price information of the appraisal target real estate) and information on the predicted contract prices per square meter Pc(1) to Pc(n) of each similar real estate (predicted contract price information of each of the multiple similar real estates) from the contract information prediction unit 434. The similarity calculation unit 435 has an AI model 435a. The AI ​​model 435a is a machine learning model capable of deriving similarities Sd(1) to Sd(n) using as input data the appraisal target data De, similar real estate data Dd(1) to Dd(n), the provisional appraisal price per square meter Pa of the appraisal target real estate, the predicted closing price per square meter Pb of the appraisal target real estate, and the predicted closing prices per square meter Pc(1) to Pc(n) of similar real estate.

[0045] For example, when calculating the similarity Sd(1) of the similar apartment C to the apartment A to be appraised, the AI ​​model 435a compares each item of the appraisal target data De with each item of the similar real estate data Dd(1), and sets the similarity Sd(1) to a smaller value the greater the difference between them. For example, the AI ​​model 435a sets the similarity Sd(1) to a smaller value the greater the difference between the exclusive area of ​​the apartment A to be appraised and the exclusive area of ​​the similar apartment C. The AI ​​model 435a also calculates the distance between the apartment A to be appraised and the similar apartment C from the latitude and longitude of the apartment A to be appraised and the latitude and longitude of the similar apartment C, and sets the similarity Sd(1) to a smaller value the longer the calculated distance.

[0046] Furthermore, the AI ​​model 435a sets the similarity Sd(1) by comparing the provisional assessed price per tsubo Pa of the real estate to be appraised, the predicted contract price per tsubo Pb of the real estate to be appraised, and the predicted contract price per tsubo Pc(1) of similar real estate. For example, the AI ​​model 435a sets the similarity Sd(1) to a smaller value the greater the difference between the predicted contract price per tsubo Pb of the real estate to be appraised and the predicted contract price per tsubo Pc(1) of similar condominium C. Furthermore, the AI ​​model 435a sets the similarity Sd(1) to a smaller value the greater the difference between the provisional assessed price per tsubo Pa of the real estate to be appraised and the actual contract price per tsubo of similar real estate.

[0047] The similarity calculation unit 435 uses such an AI model 435a to calculate the similarities Sd(1) to Sd(n) of the multiple similar properties with respect to the appraisal target property, as shown in FIG.

[0048] (First assessment information prediction unit) The first appraisal information prediction unit 436 calculates first predicted appraisal prices per square meter Pd(1)-Pd(n) of the real estate to be appraised, which are predicted from the similar real estate data Dd(1)-Dd(n).

[0049] 13, the first appraisal information prediction unit 436 acquires the appraisal target data De and similar real estate data Dd(1)-Dd(n) from the pre-processing unit 432. The first appraisal information prediction unit 436 has an AI model 436a. The AI ​​model 436a is a machine learning model capable of calculating the first predicted appraisal prices per square meter Pd(1)-Pd(n) of the appraisal target real estate using the appraisal target data De and the similar real estate data Dd(1)-Dd(n) as input data.

[0050] The AI ​​model 436a is constructed, for example, by learning the relationship between the data of a given real estate, the real estate case data of each of the real estate of the completed cases, and the information of the assessed price per tsubo of the given real estate. The data of the given real estate and the real estate case data of the real estate of the completed cases include information such as the building name, location, transportation, construction date, total number of units, number of floors above ground, exclusive area, contract price, contract date, and contract price per tsubo. The AI ​​model 436a is configured to compare, for example, each piece of information included in the data of the given real estate with each piece of information included in the real estate case data of the real estate of the completed cases, and calculate the assessed price per tsubo of the given real estate from the price per tsubo at the time of contract of the real estate of the completed cases while taking into account the magnitude of the difference between each piece of information.

[0051] The first assessment information prediction unit 436 uses the AI ​​model 436a to calculate a first predicted assessed unit price per square meter Pd(1) of the appraisal target real estate from the appraisal target data De and similar real estate data Dd(1). Similarly, the first assessment information prediction unit 436 uses the AI ​​model 436a to calculate first predicted assessed unit prices per square meter Pd(2) to Pd(n) of the appraisal target real estate from the appraisal target data De and similar real estate data Dd(2) to Dd(n). As a result, the first assessment information prediction unit 436 calculates the first predicted assessed unit prices per square meter Pd(1) to Pd(n) of the appraisal target real estate corresponding to each of the multiple similar real estates as shown in FIG. 14(A).

[0052] (Second Assessment Information Prediction Department) The second appraisal information prediction unit 437 calculates second predicted appraisal prices per square meter Pe(1)-Pe(n) of the real estate to be appraised, which are predicted from the similar real estate data Dd(1)-Dd(n).

[0053] Specifically, as shown in FIG. 13, the second appraisal information prediction unit 437 acquires the appraisal target data De and the similar real estate data Dd(1)-Dd(n) from the preprocessing unit 432, and acquires information on the predicted contract price per square meter Pb of the appraisal target real estate and the predicted contract prices per square meter Pc(1)-Pc(n) of each similar real estate from the contract information prediction unit 434. The second appraisal information prediction unit 437 has an AI model 437a. The AI ​​model 437a is a machine learning model capable of deriving the second predicted appraisal price per square meter Pe(1)-Pe(n) of the appraisal target real estate using the appraisal target data De, the similar real estate data Dd(1)-Dd(n), the predicted contract price per square meter Pb of the appraisal target real estate, and the predicted contract prices per square meter Pc(1)-Pc(n) of each similar real estate as input data.

[0054] The AI ​​model 437a is constructed, for example, by learning the relationship between the data of a given real estate, the real estate case data of each of the real estate of the completed cases, and the information of the assessed price per tsubo of the given real estate. The data of the given real estate and the real estate case data of the real estate of the completed cases include information such as the building name, location, transportation, construction date, total number of units, number of floors above ground, exclusive area, contract price, contract date, and contract price per tsubo. The AI ​​model 437a is configured to compare, for example, each piece of information included in the data of the given real estate with each piece of information included in the real estate case data of the real estate of the completed cases, and calculate the assessed price per tsubo of the given real estate from the price per tsubo at the time of contract of the real estate of the completed cases while taking into account the magnitude of the difference between each piece of information.

[0055] The second assessment information prediction unit 437 uses the AI ​​model 437a to calculate the second predicted assessment price per square meter Pe(1) of the assessment target real estate from the assessment target data De, the similar real estate data Dd(1), the predicted contract price per square meter Pb of the assessment target real estate, and the predicted contract price per square meter Pc(1) of the similar real estate. Similarly, the second assessment information prediction unit 437 uses the AI ​​model 437a to calculate the second predicted assessment price per square meter Pe(2) to Pe(n) of the assessment target real estate from the assessment target data De, the similar real estate data Dd(2) to Dd(n), the predicted contract price per square meter Pb of the assessment target real estate, and the predicted contract price per square meter Pc(2) to Pc(n) of the similar real estate. As a result, the second assessment information prediction unit 437 calculates the second predicted assessment price per square meter Pe(1) to Pe(n) of the assessment target real estate corresponding to each of the multiple similar real estates as shown in FIG. 14(B).

[0056] (Assessment Information Calculation Unit) The appraisal information calculation unit 438 calculates the appraisal price Pf of the real estate to be appraised.

[0057] Specifically, as shown in FIG. 13, the assessment information calculation unit 438 obtains information on the first predicted assessed unit price per square meter Pd(1)-Pd(n) of the assessed real estate from the first assessment information prediction unit 436, and obtains information on the second predicted assessed unit price per square meter Pe(1)-Pe(n) of the assessed real estate from the second assessment information prediction unit 437. Furthermore, the assessment information calculation unit 438 obtains information on similarities Sd(1)-Sd(n) from the similarity calculation unit 435. The assessment information calculation unit 438 also obtains similar real estate data Dd(1)-Dd(n) from the preprocessing unit 432, and obtains information on the actual contract unit price per square meter of each of the multiple similar real estates from the obtained similar real estate data Dd(1)-Dd(n). The assessment information calculation unit 438 selects multiple similar real estates corresponding to the top few similarities with the largest values ​​from the multiple similarities Sd(1)-Sd(n) as similar real estates for assessment. The number of properties selected as similar real estate for appraisal is set to, for example, five. The appraisal information calculation unit 438 calculates a weighted average of the first predicted appraisal price, the second predicted appraisal price, and the actual contract price per square meter of each of the multiple similar real estate for appraisal using a weighting coefficient that is a coefficient representing the similarity of each of the multiple similar real estate for appraisal. The appraisal information calculation unit 438 then outputs the calculated weighted average value as the appraisal price per square meter of the real estate to be appraised, and calculates and outputs the appraisal price Pf of the real estate to be appraised based on the appraisal price per square meter and the exclusive area of ​​the real estate to be appraised.

[0058] (Example of operation of information processing system) Next, an example of the operation of the information processing system 10 of this embodiment will be described.

[0059] As shown in FIG. 15, in the information processing system 10 of this embodiment, when an employee of a real estate company operates the terminal device 20 to input information on the appraisal target real estate, the information is transmitted from the terminal device 20 to the assessment server device 40 (step S10). When the assessment server device 40 receives the information on the appraisal target real estate transmitted from the terminal device 20 (step S20), the control unit 43 executes a series of processes as shown in FIG. 4, FIG. 6, FIG. 11, and FIG. 13 based on the information, thereby calculating the appraisal price Pf of the appraisal target real estate (step S21). Then, the assessment server device 40 transmits information on a plurality of appraisal similar real estates used in calculating the appraisal price Pf to the terminal device 20 together with the information on the appraisal price Pf of the appraisal target real estate (step S22). Therefore, when the number of properties selected as appraisal similar real estate is five, information on the five appraisal similar real estates is transmitted to the terminal device 20. The information on the similar real estate for appraisal includes information such as the name, location, transportation, year of construction, total number of units, number of above-ground floors, and exclusive area of ​​the similar real estate for appraisal.

[0060] When the terminal device 20 receives the assessed price Pf of the property to be appraised and information on multiple similar properties to be appraised transmitted from the appraisal server device 40 (step S11), it displays the information on a display or the like (step S12). This allows employees of the real estate company to know the assessed price Pf of the property to be appraised and to check the properties used to calculate the assessed price Pf. In addition, employees of the real estate company can present the assessed price Pf of the property to be appraised and the properties used in the appraisal to a customer who is requesting an appraisal of the property.

[0061] (Hardware configuration of information processing system) Next, an example of a hardware configuration of the computer 100 for implementing the terminal device 20 and the server devices 30 and 40 will be described with reference to FIG.

[0062] As shown in FIG. 16, the computer 100 includes a processor 101, a storage device 102, a communication device 103, an input device 104, and an output device 105. The processor 101 is a central processing unit (CPU) or a graphical processing unit (GPU). The storage device 102 is composed of at least one of a memory, a hard disk drive (HDD), and a solid state drive (SSD), for example. The communication device 103 performs wired or wireless communication. The input device 104 is a device that accepts an input operation, and is composed of at least one of a keyboard, a touch panel, a mouse, and a microphone, for example. The output device 105 is a device that outputs information, and is composed of at least one of a display, a touch panel, and a speaker, for example.

[0063] (Actions and Effects of the Information Processing System of the Present Embodiment) As described above, the information processing system 10 of this embodiment includes an assessment information acquisition unit 430, a first assessment information prediction unit 436, a similarity calculation unit 435, and an assessment information calculation unit 438. The assessment information acquisition unit 430 acquires the assessment target data Da of the assessment target real estate. The first assessment information prediction unit 436 calculates the first predicted assessment unit price Pd(1) to Pd(n) of the assessment target real estate predicted from the similar real estate data Dd(1) to Dd(n) of the similar real estate for each of the multiple similar real estates based on the assessment target data De, which has been pre-processed by the pre-processing unit 432 to link station information and location information to the assessment target data Da, and the similar real estate data Dd(1) to Dd(n) of the multiple similar real estates. In this embodiment, the first predicted assessment unit price Pd(1) to Pd(n) corresponds to predicted assessment price information (first predicted assessment price information), which is predicted price information at the time of appraisal of the assessment target real estate. The similarity calculation unit 435 calculates the similarities Sd(1)-Sd(n) of each of the multiple similar properties to the real estate to be appraised. The appraisal information calculation unit 438 selects a similar real estate for appraisal to be used in calculating the appraisal price per square meter and appraisal price of the real estate to be appraised based on the similarities Sd(1)-Sd(n) of each of the multiple similar properties from among the multiple similar properties, and calculates the appraisal price per square meter and appraisal price of the real estate to be appraised based on the first predicted appraisal price per square meter of the similar real estate for appraisal. In this embodiment, the appraisal price per square meter and appraisal price of the real estate to be appraised correspond to appraisal price information of the real estate to be appraised. With this configuration, the appraisal price per square meter and appraisal price of the real estate to be appraised can be calculated based on information on similar real estate for appraisal that is highly similar to the real estate to be appraised, making it possible to calculate the appraisal price per square meter and appraisal price of the real estate to be appraised with greater accuracy.

[0064] The information processing system 10 of this embodiment further includes a second assessment information prediction unit 437. The second assessment information prediction unit 437 calculates the second predicted assessment unit price Pe(1)-Pe(n) of the appraisal target real estate for each of a plurality of similar real estates based on the appraisal target data De, the similar real estate data Dd(1)-Dd(n), the predicted contract price Pb of the appraisal target real estate, and the predicted contract price Pc(1)-Pc(n) of the similar real estates. The assessment information calculation unit 438 calculates the appraisal unit price and appraisal price of the appraisal target real estate based on the first predicted appraisal unit price and the second predicted appraisal unit price of the appraisal similar real estates. In this embodiment, the second predicted appraisal unit price Pe(1)-Pe(n) corresponds to the second predicted appraisal price information. With this configuration, it becomes possible to calculate the assessed price per square meter and the assessed price of the real estate to be appraised with even greater accuracy.

[0065] The similarity calculation unit 435 of this embodiment calculates similarities Sd(1) to Sd(n) based on the appraisal target data De, similar real estate data Dd(1) to Dd(n), the provisional appraised price per square meter Pa of the appraisal target real estate, the predicted closing price per square meter Pb of the appraisal target real estate, and the predicted closing prices per square meter Pc(1) to Pc(n) of the similar real estate. According to this configuration, it is possible to calculate the similarities Sd(1) to Sd(n) with higher accuracy.

[0066] The information processing system 10 of this embodiment further includes a provisional information calculation unit 433. The provisional information calculation unit 433 calculates a provisional assessed price per square meter Pa of the appraisal target real estate based on the appraisal target data De. In this embodiment, the provisional assessed price per square meter Pa of the appraisal target real estate corresponds to provisional appraisal price information of the appraisal target real estate. With this configuration, it is possible to calculate the provisional assessed price per tsubo Pa of the real estate to be appraised with greater accuracy.

[0067] The information processing system 10 of this embodiment includes a contract information prediction unit 434. The contract information prediction unit 434 calculates a predicted contract price per square meter Pb of the appraisal target real estate based on the appraisal target data De. The contract information prediction unit 434 also calculates predicted contract prices per square meter Pc(1) to Pc(n) of each of a plurality of similar real estates based on the similar real estate data Dd(1) to Dd(n). According to this configuration, it is possible to calculate the estimated contract price per square meter Pb and the estimated contract prices per square meter Pc(1) to Pc(n) of the appraised real estate with higher accuracy.

[0068] As shown in Fig. 9, the appraisal target data De includes station information and location information of the appraisal target real estate. Also, as shown in Fig. 8, the similar real estate data Dd(1) to Dd(n) include station information and location information of the similar real estate. According to this configuration, the estimated closing price per square meter Pb and the estimated closing prices per square meter Pc(1) to Pc(n) of the real estate to be appraised are calculated based on the station information and location information of the real estate to be appraised and similar real estate, making it possible to calculate the estimated closing prices per square meter Pb, Pc(1) to Pc(n) with greater accuracy.

[0069] The information processing system 10 of this embodiment includes a neighboring case acquisition unit 431 and a pre-processing unit 432. The neighboring case acquisition unit 431 acquires neighboring case data Db(1) to Db(m) of multiple real estate. In this embodiment, the neighboring case acquisition unit 431 corresponds to a real estate information acquisition unit. The pre-processing unit 432 selects multiple similar real estate that is similar to the real estate to be assessed from among multiple neighboring real estate, based on the appraisal target data Da and the neighboring case data Db(1) to Db(m). In this embodiment, the pre-processing unit 432 corresponds to a similar real estate selection unit. According to this configuration, it is possible to select similar real estate that is similar to the real estate to be appraised with greater accuracy.

[0070] <Second embodiment> Next, an information processing system 10 according to a second embodiment will be described, focusing on differences from the information processing system 10 according to the first embodiment.

[0071] (Configuration of the assessment server device) As indicated by the dashed line in FIG. 2, the control unit 43 of this embodiment differs from the control unit 43 of the first embodiment in that it further includes a reliability calculation unit 439.

[0072] (Reliability calculation section) The reliability calculation unit 439 calculates the reliability Ra of the appraisal price Pf of the appraisal target real estate calculated by the appraisal information calculation unit 438.

[0073] Specifically, as shown in FIG. 17, the reliability calculation unit 439 acquires the appraisal target data De and the similar real estate data Dd(1) to Dd(n) from the preprocessing unit 432, and acquires information on the similarities Sd(1) to Sd(n) from the similarity calculation unit 435. The reliability calculation unit 439 also acquires information on the appraisal price Pf of the appraisal target real estate calculated by the appraisal information calculation unit 438. The reliability calculation unit 439 has an AI model 439a. The AI ​​model 439a is a machine learning model that can derive the reliability Ra of the appraisal price Pf using the appraisal target data De, the similar real estate data Dd(1) to Dd(n), the similarities Sd(1) to Sd(n), and the appraisal price Pf of the appraisal target real estate as input data. The reliability Ra in this embodiment is a value set in the range from "0" to "1", and the closer to "0" the lower the reliability, and the closer to "1" the higher the reliability.

[0074] For example, the AI ​​model 439a calculates the variance of the contract price per tsubo of multiple similar real estates included in the similar real estate data Dd(1) to Dd(n), and sets the reliability Ra to a higher value as the variance value is smaller. Also, the AI ​​model 439a compares the maximum and minimum values ​​of the contract price per tsubo of multiple similar real estates included in the similar real estate data Dd(1) to Dd(n), and sets the reliability Ra to a higher value as the difference between them is smaller. Furthermore, the AI ​​model 439a compares the location of the real estate to be assessed included in the appraisal target data De with the locations of multiple similar real estates included in the similar real estate data Dd(1) to Dd(n), and sets the reliability Ra to a higher value if the location of any of the multiple similar real estates matches the location of the real estate to be assessed. Also, the AI ​​model 439a sets the reliability Ra to a higher value as the similarities Sd(1) to Sd(n) include more values ​​close to "1". The AI ​​model 439a is constructed by learning so that the smaller the error between the appraisal price Pf of the real estate to be appraised calculated by the appraisal information calculation unit 438 and the actual appraisal price, the higher the reliability Ra.

[0075] The reliability calculation unit 439 uses such an AI model 439a to calculate the reliability Ra of the assessed price Pf of the assessed real estate. The reliability Ra calculated by the reliability calculation unit 439 is transmitted to the terminal device 20 together with the assessed price Pf. Therefore, the terminal device 20 can display the assessed price Pf and the reliability Ra side by side.

[0076] (Actions and Effects of the Information Processing System of the Present Embodiment) The information processing system 10 of this embodiment further includes a reliability calculation unit 439. The reliability calculation unit 439 calculates the reliability Ra of the appraisal price Pf of the appraisal target real estate based on the appraisal target data De, the similar real estate data Dd(1) to Dd(n), the similarities Sd(1) to Sd(n), and the appraisal price Pf of the appraisal target real estate.

[0077] According to this configuration, it is possible to determine whether the assessed price Pf of the real estate to be appraised is reliable by checking the value of the reliability Ra, thereby improving convenience.

[0078] <Other embodiments> The present disclosure is not limited to the above specific examples.

[0079] For example, the predicted appraisal price information of the real estate to be appraised is not limited to the predicted prices per tsubo at the time of appraisal Pd(1)-Pd(n), Pe(1)-Pe(n), but may also include the predicted price per square meter at the time of appraisal. Similarly, the appraisal price information of the real estate to be appraised is not limited to the assessed price per tsubo and the appraisal price Pf, but may also include the assessed price per square meter, etc. Furthermore, the predicted closing price information of the real estate to be appraised and similar real estate is not limited to the predicted closing prices per tsubo Pb, Pc(1)-Pc(n), but may also include the predicted price per square meter, etc. Furthermore, the provisional appraisal price information of the real estate to be appraised is not limited to the provisional assessed price per tsubo Pa, but may also include the provisional assessed price per square meter, etc.

[0080] The similarity calculation unit 435 may calculate the similarities Sd(1) to Sd(n) based on at least one of the appraisal target data De, the similar real estate data Dd(1) to Dd(n), the provisional appraisal price per square meter Pa of the appraisal target real estate, the predicted contract price per square meter Pb of the appraisal target real estate, and the predicted contract price per square meter Pc(1) to Pc(n) of the similar real estate. For example, the similarity calculation unit 435 may calculate the similarities Sd(1) to Sd(n) based on the appraisal target data De and the similar real estate data Dd(1) to Dd(n).

[0081] The appraisal information calculation unit 438 may calculate the appraisal price per square meter and appraisal price Pf of the real estate to be appraised based on the first estimated appraisal price per square meter Pd(1) to Pd(n) and the similarity Sd(1) to Sd(n) without using the second estimated appraisal price per square meter Pe(1) to Pe(n).

[0082] The real estate to be appraised, the neighboring example real estate, and similar real estate are not limited to condominiums, but may be detached houses, land, etc. For example, when using detached houses or land as real estate, the information for appraising them may include the land area, land shape (regular land, irregular land, etc.), road width, building coverage ratio, floor area ratio, etc.

[0083] The above specific examples, to which a person skilled in the art may make appropriate design modifications, are also included within the scope of the present disclosure as long as they have the features of the present disclosure. The elements of each of the above specific examples, as well as their arrangements, conditions, shapes, etc., are not limited to those exemplified and can be modified as appropriate. The elements of each of the above specific examples can be combined as appropriate as long as no technical contradictions arise. [Explanation of symbols]

[0084] 10: Information processing system, 40: Appraisal server device (information processing device), 100: Computer, 431: Neighboring case acquisition unit (real estate information acquisition unit), 432: Pre-processing unit (similar real estate selection unit), 433: Provisional information calculation unit, 434: Deal information prediction unit, 435: Similarity calculation unit, 436: First appraisal information prediction unit, 437: Second appraisal information prediction unit, 438: Appraisal information calculation unit, 439: Reliability calculation unit.

Claims

1. a first appraisal information prediction unit that calculates, based on appraisal data of the appraised real estate and similar real estate data of a plurality of similar real estates, predicted appraisal price information, which is predicted price information at the time of appraisal of the appraised real estate predicted from the similar real estate data of each of the plurality of similar real estates, for each of the similar real estates, the similar real estate data including information similar to the information included in the appraisal data; a similarity calculation unit that calculates a similarity between the appraisal target real estate and each of the plurality of similar real estates; an appraisal information calculation unit that selects an appraisal similar real estate to be used in calculating appraisal price information of the appraisal target real estate from among the plurality of similar real estates based on the respective similarities of the plurality of similar real estates, and calculates appraisal price information of the appraisal target real estate based on the predicted appraisal price information of the appraisal similar real estates. Information processing system.

2. When the predicted appraisal price information calculated by the first appraisal information prediction unit is set as first predicted appraisal price information, a second appraisal information prediction unit that calculates second predicted appraisal price information for each of the similar real estate data of the plurality of similar real estates based on the appraisal target data, the similar real estate data of each of the plurality of similar real estates, predicted closing price information of the appraisal target real estate, and predicted closing price information of each of the plurality of similar real estates, The appraisal information calculation unit calculates appraisal price information of the appraisal target real estate based on the first predicted appraisal price information and the second predicted appraisal price information corresponding to the appraisal similar real estate. The information processing system according to claim 1 .

3. The similarity calculation unit calculates the similarity for each of the plurality of similar real estate properties based on the appraisal target data and the similar real estate data. The information processing system according to claim 1 .

4. The similarity calculation unit calculates the similarity for each of the plurality of similar real estates based on the appraisal target data and the similar real estate data, as well as provisional appraisal price information for the appraisal target real estate, predicted closing price information for the appraisal target real estate, and predicted closing price information for each of the plurality of similar real estates. The information processing system according to claim 3 .

5. The method further includes a provisional information calculation unit that calculates provisional appraisal price information of the appraisal target real estate based on the appraisal target data.

5. The information processing system according to claim 4.

6. The method further includes a transaction information prediction unit that calculates predicted transaction price information of the appraisal target real estate based on the appraisal target data, and calculates predicted transaction price information of each of the plurality of similar real estates based on the similar real estate data.

5. The information processing system according to claim 2 or 4.

7. The appraisal target data includes station information and location information of the appraisal target real estate, The similar real estate data includes station information and location information of the similar real estate. The information processing system according to claim 1 .

8. a real estate information acquisition unit for acquiring neighborhood case data of a plurality of real estate properties existing in the neighborhood of the real estate to be appraised; and a similar real estate selection unit that selects the plurality of similar real estates from the plurality of real estates based on the appraisal target data and the neighborhood case data. The information processing system according to claim 1 .

9. a reliability calculation unit that calculates reliability of the appraisal price information of the appraisal target real estate based on the appraisal target data, the similar real estate data of each of the plurality of similar real estates, the similarity of each of the plurality of similar real estates to the appraisal target real estate, and appraisal price information of the appraisal target real estate. The information processing system according to claim 1 .

10. a first appraisal information prediction unit that calculates, based on appraisal data of the appraised real estate and similar real estate data of a plurality of similar real estates, predicted appraisal price information, which is predicted price information at the time of appraisal of the appraised real estate predicted from the similar real estate data of each of the plurality of similar real estates, for each of the similar real estates, the similar real estate data including information similar to the information included in the appraisal data; a similarity calculation unit that calculates a similarity between the appraisal target real estate and each of the plurality of similar real estates; an appraisal information calculation unit that selects an appraisal similar real estate to be used in calculating appraisal price information of the appraisal target real estate from among the plurality of similar real estates based on the respective similarities of the plurality of similar real estates, and calculates appraisal price information of the appraisal target real estate based on the predicted appraisal price information of the appraisal similar real estates. Information processing device.

11. On the computer, based on the appraisal target data of the appraisal target real estate and similar real estate data of each of a plurality of similar real estates, the similar real estate data including information similar to the information included in the appraisal target data, calculate predicted appraisal price information, which is predicted from the similar real estate data of each of the plurality of similar real estates at the time of appraisal of the appraisal target real estate, for each of the similar real estate data of the plurality of similar real estates; Calculating the similarity of each of the plurality of similar properties to the real estate to be appraised; A similar real estate for appraisal to be used in calculating appraisal price information of the real estate to be appraised is selected from the plurality of similar real estates based on the respective similarities of the plurality of similar real estates, and the appraisal price information of the real estate to be appraised is calculated based on the predicted appraisal price information of the similar real estate for appraisal. program.

Images