Information processing device, information processing method, and information processing program

Abstract

To improve the liquidity of real estate in the world. [Solution] An information processing device comprising: a calculation unit that takes real estate information as input data and calculates the probability of closing a property for each property sales price, to a property closing probability estimation model that is generated by learning past property closing records as training data for real estate closing probability estimation; and a generation unit that generates display data for displaying an image showing the change in closing probability due to differences in property sales price.

Description

Technical Field

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

Background Art

[0002] In the above technical field, Patent Document 1 discloses a technique for determining an empty house type based on measurement history information of energy consumption and generating an empty house rate map for each predetermined area on a map.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, in the technique described in the above document, although the existence of an empty house can be recognized, it does not show a strategy for circulating the empty house.

[0005] An object of the present invention is to provide a technique for solving the above problems.

Means for Solving the Problems

[0006] To achieve the above object, the apparatus according to the present invention inputs real estate information as input data to a contract probability estimation model that estimates a contract probability, which is generated by learning past contract results of real estate as teacher data, and calculates a contract probability for each real estate sales price; and a generation unit that generates display data for displaying an image showing a change in contract probability due to a difference in the real estate sales price. The information processing apparatus is provided with the above components.

[0007] To achieve the above objective, the method according to the present invention is The calculation unit generates a sales probability estimation model that estimates the probability of a sale by learning from past real estate transaction data as training data. This model is then input with real estate information as input data, and the calculation step involves calculating the probability of a sale for each real estate sales price. The generation unit generates display data for displaying an image showing the change in the probability of closing a deal due to the difference in real estate sales prices, This is an information processing method that includes [something].

[0008] To achieve the above objective, the program according to the present invention The calculation step involves inputting real estate information as input data into a sales probability estimation model, which is generated by learning from past real estate transaction data as training data, to calculate the probability of closing a sale for each real estate sales price. A generation step to generate display data for displaying an image showing the change in the probability of closing a deal due to the difference in real estate sales prices, It is an information processing program that causes a computer to execute something. [Effects of the Invention]

[0009] According to the present invention, the liquidity of real estate can be improved. [Brief explanation of the drawing]

[0010] [Figure 1] This is a block diagram showing the configuration of the information processing device according to the first embodiment. [Figure 2] This is a diagram showing the background of the information processing device according to the second embodiment. [Figure 3] This figure shows an overview of the processing of the information processing device according to the second embodiment. [Figure 4A] This is a block diagram showing the functional configuration of the information processing device according to the second embodiment. [Figure 4B] This figure shows the configuration of the training data storage unit according to the second embodiment. [Figure 5]It is a flowchart showing the processing procedure of the information processing apparatus according to the second embodiment. [Figure 6] It is a diagram showing the commitment probability distribution according to the second embodiment. [Figure 7] It is a diagram showing the prediction accuracy according to the second embodiment. [Figure 8] It is a diagram showing an example of a display screen according to the second embodiment. [Figure 9] It is a diagram showing an example of a display screen according to the second embodiment. [Figure 10] It is a block diagram showing the functional configuration of the information processing apparatus according to the third embodiment. [ [Figure 11] It is a block diagram showing the functional configuration of the information processing apparatus according to the fourth embodiment.

Modes for Carrying Out the Invention

[0011] Hereinafter, embodiments of the present invention will be exemplarily described in detail with reference to the drawings. However, the components described in the following embodiments are merely examples, and are not intended to limit the technical scope of the present invention thereto.

[0012] [First Embodiment] The information processing apparatus 100 as the first embodiment of the present invention will be described with reference to FIG. 1. The information processing apparatus 100 is a device for realizing shortening of the vacant house period.

[0013] As shown in FIG. 1, the information processing apparatus 100 includes a commitment probability calculation unit 101 and a display data generation unit 102.

[0014] The commitment probability calculation unit 101 inputs the real estate information 120 as input data into a commitment probability estimation model 111 that estimates the commitment probability, which is generated by learning the past commitment results 110 of real estate as teacher data, and calculates the commitment probability for each real estate sales price.

[0015] The display data generation unit 102 generates display data for displaying an image showing the change in the probability of closing a deal due to differences in real estate sales prices 130.

[0016] With the above configuration, it is possible to provide users with information on real estate and the resulting changes in the probability of closing a deal based on differences in real estate sales prices, thereby improving the liquidity of real estate.

[0017] [Second Embodiment] Next, an information processing device according to the second embodiment of the present invention will be described using Figure 2 and subsequent figures. Figure 2 is a diagram illustrating the necessity of the information processing device according to this embodiment. As shown in Figure 2, in the current situation in 201, there are many used houses that are known as "negative real estate" because there are no buyers. This is because the owners were unable to sell the vacant houses at the desired price, and in the meantime, damage progressed, making them even harder to sell. Both the owners and real estate agents profit more if they can sell at a high price, so the selling price tends to be high, but if the house remains unsold for a year or two, the damage worsens, making it even harder to sell. As the damage progresses, the burden on the owners increases, so they consult with local governments, etc., but by the time they consult, the damage has progressed and there is often no prospect of selling. There is no other option but demolition. Local governments also face a heavy workload, dealing with complaints from neighbors and handling repairs.

[0018] In the ideal scenario 202, owners would understand how different selling prices affect the probability of a sale, allowing them to review the property's selling price early, collaborate with real estate brokerage companies to register the property on real estate websites early, set a fair price, and carry out repairs. By enabling owners to take early and appropriate measures for the circulation of their homes, the vacancy period can be shortened, and the circulation of used homes can be promoted and standardized.

[0019] <Overview of information processing equipment> Figure 3 is a diagram showing the processing overview of the information processing device according to this embodiment. The data used by the information processing device can be broadly divided into real estate advertising data 301, lifeline data 302, resident registration data 303, and open data 304.

[0020] First, either or both of the lifeline data 302 and the resident registry data 303 are used to detect that a person has moved into a property, i.e., that the sale of that property has been completed. Even if only one of the lifeline data 302 or the resident registry data 303 is available, the completion of the sale can still be detected. Contractor information / customer information may also be used.

[0021] Next, using real estate advertising data 301 obtained from a real estate brokerage company, we estimate the transaction price and advertising period of the property. Real estate advertising data 301 is data from real estate sales advertisements submitted at the user's request. This real estate sales advertising data includes a history of the sales prices at which the user's property has been advertised. By looking at this real estate sales advertising data, we can find out the sales price that was advertised immediately before a new tenant moved in. We estimate that sales price as the transaction price.

[0022] As shown in Graph 311, suppose that for a property identified by address, the water supply was turned on or the water usage increased. Looking at the property advertisement data 301 immediately preceding that water usage increase, we can see that the advertised price was lowered. This suggests that lowering the advertised selling price made it easier to close a deal, which was then closed, and a new tenant moved in. The "closing price" can be obtained as an explanatory variable, and from "the period from when the advertisement was placed at that closing price until the deal was closed," "whether or not a deal was closed for each period" can be obtained as the dependent variable.

[0023] Simply removing a property advertisement doesn't necessarily mean it was actually sold, but by combining it with utility data, the transaction price can be estimated with high accuracy. The total period during which the property was unoccupied (vacancy period) contributes as an explanatory variable in estimating the probability of sale, with whether or not a sale was completed being the dependent variable. Open data 304 includes distance from the station and time to the Shinkansen station. The age of the person who actually lived there and the number of household members, which can be obtained from the Basic Resident Register 303, also serve as explanatory variables.

[0024] Next, the model learns whether or not a property was sold within a specified period, using the asking price of that property as training data. For example, suppose a property was put on the market for 30 million yen and sold after 6 months. In this case, when learning to generate the 3-month sales probability model, the explanatory variables of that property (selling price, location, age of the building, whether or not it has been renovated, time elapsed since renovation, building structure, total floor area, owner's attributes, etc.) are learned as data for properties that were not sold (0 data). On the other hand, when learning to generate the 6-month, 9-month, and 12-month sales probability models, the same explanatory variables of that property are learned as sales price data for properties that were sold (1 data).

[0025] In short, it detects real estate transactions and learns the attributes and transaction amounts of the properties that have been sold. As an example of utility data, it estimates the time of a transaction by looking at changes in water usage and when water taps are turned on, and learns either 0 data points or 1 data point depending on the period from the start of advertising to the time of the transaction.

[0026] For real estate advertising data 301, historical sales price data such as real estate advertisements published every month should be used. For lifeline data 302, changes in usage fees and start dates for utilities such as electricity and gas, not limited to water, may also be used. In addition, information about residents, such as registration and property tax payment status, may be used, not limited to the resident registry 303.

[0027] The open data 304 included in the explanatory variables contains not only the address of the property, but also location information such as the distance to the nearest station, the distance from the Shinkansen station, the surrounding population, and the rental rate.

[0028] Owner attributes included in the explanatory variables include, but are not limited to, age, gender, family structure, annual income, and assets, which are derived from sources such as the Basic Resident Register 303.

[0029] Through the learning process described above, four models are simultaneously generated as an example: model 305 for 3-month predictions, model 306 for 6-month predictions, model 307 for 9-month predictions, and model 308 for 12-month predictions. Here, we have generated contracted probability estimation models for each of these four periods, but the present invention is not limited to this. The periods may be divided into even smaller periods, or models for making predictions over shorter periods such as one month, or models for making predictions over longer periods such as 36 months may be generated.

[0030] By inputting various property attributes into this sales probability estimation model, users can be presented with graphs 309 and 310 showing how the sales probability changes with respect to price over different time periods. Users can input the attributes of the property they wish to sell and their desired selling price to understand how long it will take and what the probability of a sale is, enabling them to accurately revise their desired selling price. For example, a house that a user wants to sell quickly can lower its price to increase the probability of a sale. In this case, if it is found that the sales probability does not change significantly even if the selling price is lowered below a certain amount, that price is considered an appropriate selling price that will yield a high probability of a sale.

[0031] On the other hand, if profit is the priority, it becomes possible to set the desired selling price at the highest price (e.g., 60 million yen) within a predetermined range (e.g., medium) that has a relatively long probability of closing within a certain period (e.g., 12 months). In other words, it can be used as a sales strategy and a tool to promote the circulation of vacant houses. If such a system is introduced to a local government, users living in that local government will be able to easily develop a sales strategy for the property they want to sell and more accurately grasp an appropriate desired selling price. Ultimately, this is expected to increase the likelihood of real estate sales being completed within a certain period in that local government, and reduce the number of vacant houses. When making predictions, if age is entered, the accuracy will differ between people in their 20s and those in their 50s, and this system can also be used to select advertising targets.

[0032] Figure 4 shows the configuration of the information processing device 400 according to this embodiment. The information processing device 400 includes a model generation unit 401, a contract probability calculation unit 402, and a display data generation unit 403.

[0033] The model generation unit 401 generates transaction probability estimation models 305 to 308 that estimate the probability of a transaction by learning from past real estate transaction records 410 as training data.

[0034] The transaction probability calculation unit 402 inputs information 420 about the property to be sold as input data into the transaction probability estimation models 305-308 and calculates the transaction probability for each property selling price.

[0035] The display data generation unit 403 generates display data for displaying an image showing the change in the probability of closing a deal due to differences in real estate sales prices 430.

[0036] The sales probability estimation models 305-308 are prepared separately for each of the different sales periods, and the sales probability calculation unit 402 calculates the sales probability for each real estate sales price for each different sales period.

[0037] The display data generation unit 403 generates display data for displaying an image showing the change in the probability of closing a deal due to differences in real estate sales prices for different sales periods.

[0038] The model generation unit 401 learns whether or not a sale has been made for each property and for each period from when the sale price is announced until someone moves in, using this as the target variable, and generates sale probability estimation models 305 to 308. In other words, it detects properties for which a sale is estimated to be made and their estimated sale price, and learns whether or not a sale has been made during a predetermined period (in this case, a period of 3 to 12 months) from when the sale price is announced.

[0039] The model generation unit 401 may generate transaction probability estimation models 305 to 308 by learning the period of time a property is vacant and the transaction price as target variables for each property. The model generation unit 401 may generate transaction probability estimation models 305 to 308 by learning at least the transaction price and the period from when the transaction price is offered until someone moves in for each property as target variables.

[0040] Here, the model generation unit 401 determines the timing of a person moving into a property using at least one of the lifeline data 302 and the resident registry 303.

[0041] The model generation unit 401 further learns the estimated transaction price and transaction timing from the real estate advertising data 301 (price trends offered by real estate sales agents) to generate transaction probability estimation models 305 to 308.

[0042] Lifeline data 302 refers to data on the opening / closing or meter reading of water, electricity, or gas lines.

[0043] The model generation unit 401 may further use the resident registry 303 or the like to learn at least one of the purchaser's age, gender, family structure, annual income, and assets as explanatory variables to generate sales probability estimation models 305 to 308.

[0044] The model generation unit 401 may further use registration information, open data, etc., to learn any of the following as explanatory variables: location of the property, age of construction, building structure, area, number of floors, presence or absence of road access, and presence or absence of renovations, and generate sales probability estimation models 305 to 308.

[0045] The sales probability calculation unit 402 uses these sales probability estimation models 305-308 to calculate what price the product will be sold for, what the probability of it being sold, and by when.

[0046] As shown in Figure 4B, the model generation unit 401 may store data content 412, data to be used 413, and input data 414 to be used as explanatory variables for the learning model, associated with the name of the municipal data 411. The input data 414 to be used as explanatory variables may include, but are not limited to, water usage from water usage information, household size (household composition), number of days since moving in, resident age including the age of the head of household from the resident register, number of days since moving in from the fixed asset tax register, year of construction, and year of construction from real estate registration.

[0047] Figure 5 is a flowchart showing the processing flow performed by the information processing device 400. First, in step S501, the system comprehensively detects occupancy of real estate using lifeline data 302 and the resident registry 303. Next, in step S502, various data related to the real estate for which occupancy has been detected are input as training data. Specifically, real estate information 501, resident information 502, and estimated sales price 503 are input as explanatory variables. Real estate information 501 includes address, year built, building structure, area, number of floors, road access, whether or not renovations have been done, and the period since renovations. Resident information 502 includes the number of household members, family structure, and age of the residents who were living there. In step S503, the system constructs the contract probability estimation models 305-308 using the input training data. Furthermore, in step S504, the prediction accuracy is calculated.

[0048] Next, moving to the transaction probability estimation phase, step S511 estimates the transaction probability for each period and price using the generated transaction probability estimation models 305-308. At this time, the input data used is real estate information 551 and residential information 552 related to the property that the user wants to sell. Real estate information 551 includes address, year built, building structure, area, number of floors, road access, whether or not it has been renovated, and the period since renovation. Residential information 552 includes the period the property has been vacant. The desired selling price may also be used as input data.

[0049] Step S512 calculates the probability of closing a deal for each price range over different time periods. At this point, you may also calculate the price ranges that result in high, medium, and low closing probabilities.

[0050] Step S513 displays the change in the probability of closing a deal based on price over a period of time. In other words, it estimates and shows how much the probability of closing a deal increases when the price is lowered by a certain amount.

[0051] Figure 6 shows the calculated transaction probability distribution 601. Figure 6 shows the transaction probability for each property price and time to sale during the training and validation of the transaction probability model. From Figure 6, it can be seen that properties with low property prices have longer sales periods and take longer to reach the market. On the other hand, it can be seen that properties with medium property prices can be expected to reach the market sooner by improving their attractiveness through renovations, etc.

[0052] Figure 7 shows a prediction accuracy of 701. Of the prepared training data, 30% was not used for training but was used as test data to verify the model's performance. As shown in Figure 7, the performance verification using the test data resulted in the ability to predict conversions and non-conversions with an accuracy of 93.0%.

[0053] Figure 8 shows the GUI of the data input screen 801 that is displayed when a user who wishes to sell real estate accesses the information processing device 400 from their user terminal. On the map 801, the user specifies the location of the property they wish to sell, and then inputs the desired selling price 802, postal code 803, year built 804, building structure 805, whether or not it has been renovated 806, and the time elapsed since the last renovation 807. Once these are entered and the confirmation button 808 is selected, the system transitions to the estimated transaction probability display screen 901, as shown in Figure 9.

[0054] [Third Embodiment] Next, an information processing apparatus according to a third embodiment of the present invention will be described. Compared to the second embodiment described above, the information processing apparatus according to this embodiment differs in that it maps the sales price (estimated used property price) that achieves a predetermined probability of closing onto a map. The other configurations and operations are the same as in the second embodiment, so the same reference numerals are used for the same configurations and operations and their detailed descriptions are omitted.

[0055] The sales probability calculation unit 1002 uses various models to determine an estimated selling price for each property on the map that will achieve a predetermined sales probability (e.g., 80%) over a predetermined period (e.g., 6 months). The display data generation unit 1003 then uses the map data and the estimated selling price (used property price) to generate a mapping image of the selling prices that will achieve the predetermined sales probability.

[0056] According to this embodiment, estimated prices of used properties can be mapped on a map, providing useful information to users who are considering purchasing a used property, for example.

[0057] [Fourth Embodiment] Next, an information processing apparatus according to the fourth embodiment of the present invention will be described with reference to Figure 11. The information processing apparatus according to this embodiment differs from the second embodiment in that it generates models based on age and family structure. The other configurations and operations are the same as those of the second embodiment, so the same reference numerals are used for the same configurations and operations, and their detailed descriptions are omitted.

[0058] The model generation unit 1101 obtains the age and family structure of people (buyers, residents) who are presumed to have completed a transaction from the resident registry and contract information. Then it generates age-specific transaction probability estimation models 1121 to 1124.

[0059] The sales probability calculation unit 1102 calculates the sales probability 1130 using the sales probability estimation model 1121 for people in their 20s if the person wishing to sell the property in the sales property information 420 is in their 20s. Here, the models are divided by age, but models based on family structure may also be prepared. For example, a sales probability estimation model for single people, a sales probability estimation model for married couples without children, a sales probability estimation model for married couples with one child, a sales probability estimation model for married couples with two children, a sales probability estimation model for married couples with three children, a sales probability estimation model for multi-generational households, and so on.

[0060] According to this embodiment, it is possible to calculate the optimal selling price and property information (such as whether or not renovations have been done) that will lead to re-occupancy, based on age and family structure. For example, for a family in their 30s with children, there is a 20% chance that re-occupancy will occur within 3 months at a price of 20 million yen, and for a working couple in their 40s, there is a 10% chance that re-occupancy will occur within 3 months at the same price. Thus, the probability of closing a deal can be calculated with high accuracy according to the age of the seller.

[0061] [Other embodiments] Although the present invention has been described above with reference to embodiments, the present invention is not limited to the above embodiments. Various modifications to the structure and details of the present invention can be made, as can be understood by those skilled in the art within the technical scope of the present invention. Furthermore, any system or apparatus that combines the separate features included in each embodiment is also within the technical scope of the present invention.

[0062] Furthermore, the present invention may be applied to a system composed of multiple devices or to a single device. Moreover, the present invention is also applicable when an information processing program that realizes the functions of the embodiment is supplied to a system or device and executed by a built-in processor. To realize the functions of the present invention on a computer, a program installed on a computer, a medium storing the program, a server that downloads the program, and a processor that executes the program are all included in the technical scope of the present invention. In particular, at least a non-transitory computer-readable medium storing a program that causes a computer to execute the processing steps included in the embodiments described above is included in the technical scope of the present invention.

Claims

1. A calculation unit that inputs real estate information as input data to a sales probability estimation model, which is generated by learning from past real estate transaction data as training data to estimate the probability of a transaction, calculates the probability of a transaction for each real estate sales price, A generation unit that generates display data for displaying an image showing the change in the probability of closing a deal due to the difference in real estate sales prices, Equipped with an information processing device.

2. The aforementioned contract probability estimation model is prepared separately for each different sales period, and the calculation unit calculates the contract probability for each real estate sales price for each different sales period. The information processing apparatus according to claim 1, wherein the generation unit generates display data for displaying an image showing the change in the probability of closing a deal due to differences in real estate sales prices for different sales periods.

3. The information processing device according to claim 1, wherein the contract probability estimation model is generated by learning whether or not a contract has been concluded for each property and for each period from when the contract price is offered until a person moves in, with the outcome of the contract being determined as the target variable.

4. The information processing device according to claim 1, wherein the transaction probability estimation model is generated by learning, at least for each property, the transaction price and the period from when the transaction price is offered until a person moves in as target variables.

5. The information processing device according to claim 3, which determines the timing of the person's occupancy using at least one of resident registration data and utility data.

6. In generating the aforementioned contract probability estimation model, The information processing device according to claim 3, which determines the timing of occupancy of the aforementioned real estate using at least one of the resident registration data and the utility data, and learns the price trends offered by real estate sales agents and the estimated closing price from the timing of occupancy.

7. The information processing device according to claim 5 or 6, wherein the lifeline data is data on the opening / closing or meter reading of water, electricity, or gas.

8. The information processing device according to claim 1, wherein the transaction probability estimation model is generated by learning at least one of the buyer's age, gender, family structure, annual income, and assets as explanatory variables.

9. The information processing device according to claim 1, wherein the transaction probability estimation model is generated by learning any of the following as explanatory variables: location of the property, age of the building, building structure, area, number of floors, presence or absence of road access, and presence or absence of renovations.

10. The calculation unit generates a sales probability estimation model that estimates the probability of a sale by learning from past real estate transaction data as training data. This model is then input with real estate information as input data, and the calculation step involves calculating the probability of a sale for each real estate sales price. The generation unit generates display data for displaying an image showing the change in the probability of closing a deal due to the difference in real estate sales prices, Information processing methods including

11. The calculation step involves inputting real estate information as input data into a sales probability estimation model, which is generated by learning from past real estate transaction data as training data, to calculate the probability of closing a sale for each real estate sales price. A generation step to generate display data for displaying an image showing the change in the probability of closing a deal due to the difference in real estate sales prices, An information processing program that causes a computer to execute something.

Images