Device and method for predicting rental price variation using artificial intelligence-based lease market analysis

Abstract

An apparatus for analyzing a rental market using artificial intelligence to predict a rate of change in rent is disclosed. The apparatus includes a rent information collection unit configured to collect rent information for rental contracts in which actual transactions have taken place; a converted rent calculation unit configured to calculate a converted rent for each rental contract based on the collected rent information and to calculate and store a rate of change in rent for each segment defined by a predetermined time interval; and a rent change rate estimation unit configured to estimate a rate of change in rent in a predetermined future segment based on policy information and non-policy information of multiple past segments including the current segment.

Description

Technology Field

[0001] Embodiments of the present invention relate to an apparatus for analyzing the rental market and predicting the rate of change in rent using artificial intelligence, and a method of operating the same. Background Technology

[0002] Generally, in the real estate rental market, the monthly rent of a property is determined through negotiation between the parties, and market prices can be identified through actual transaction data. However, for properties where actual transaction data is unavailable, there is a problem in objectively calculating an appropriate monthly rent. In particular, in areas with excellent transportation accessibility, such as those near subway stations, rents vary significantly depending on location conditions; therefore, reliable price predictions are difficult using methods that rely on experience or intuition. Furthermore, most current rental information services provide data only on an individual property basis and have limitations in providing map-based information that allows for a visual and intuitive understanding of rent levels at the regional level.

[0003] In addition, participants in the rental market intend to make decisions based on forecasting not only the rental rates of the target area but also the rate of change in rental rates of the overall rental market. Therefore, technology is required to predict the direction of rental rates in the overall rental market. The problem to be solved

[0004] The objective of the present invention, which aims to solve the aforementioned problems, is to provide a device and a method of operation thereof for analyzing the correlation between various factors affecting the rental market and the rate of change in rent (change rate) using artificial intelligence, and for predicting the rate of change in rent in the future based on this analysis. means of solving the problem

[0005] A rent map generating device for generating a rent map of a station area according to embodiments of the present invention comprises: a rent information collecting unit configured to collect rent information for lease contracts in which actual transactions have taken place; a converted rent calculation unit configured to calculate a converted rent for each lease contract based on the collected rent information and to calculate and store a rent change rate for each segment defined at predetermined time intervals; and a rent change rate estimating unit configured to estimate a rent change rate in a future predetermined segment based on policy information and non-policy information of a plurality of past segments including the current segment. Effects of the invention

[0006] According to embodiments of the present invention, the correlation between various factors affecting the rental market and the rate of change in rent (rate of change) is analyzed using artificial intelligence, and based on this, there is an effect of predicting the rate of change in rent in the future. Brief explanation of the drawing

[0007] FIG. 1 shows a rental map generating device according to embodiments of the present invention. FIG. 2 shows rental fee information collected according to embodiments of the present invention. FIG. 3 illustrates the process of calculating the converted rent according to embodiments of the present invention. FIGS. 4 and FIGS. 5 are drawings for illustrating a rental map according to embodiments of the present invention. FIG. 6 is a diagram illustrating policy information and non-policy information according to embodiments of the present invention. FIG. 7 is a drawing for explaining a rent change rate estimation unit according to embodiments of the present invention. FIG. 8 is a diagram illustrating the hardware configuration of an operating server according to embodiments of the present invention. Specific details for implementing the invention

[0008] The present invention is susceptible to various modifications and may have various embodiments; specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the invention to specific embodiments, and it should be understood that the invention includes all modifications, equivalents, and substitutions that fall within the spirit and scope of the invention. Similar reference numerals have been used for similar components in the description of each drawing.

[0009] Terms such as first, second, A, B, etc., may be used to describe various components, but said components should not be limited by said terms. These terms are used solely for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be named the second component, and similarly, the second component may be named the first component. The term "and / or" includes a combination of a plurality of related described items or any of a plurality of related described items.

[0010] When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. On the other hand, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between.

[0011] The terms used in this application are used merely to describe specific embodiments and are not intended to limit the invention. The singular expression includes the plural expression unless the context clearly indicates otherwise. In this application, terms such as "comprising" or "having" are intended to specify the presence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0012] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the present invention pertains. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this application.

[0013] The server referred to in the present invention may be constructed as a server performing at least one of the roles of a web server, a database server, and a mobile server; for example, it may display processed results on a webpage via an online network or receive necessary input data through a webpage. Here, a webpage should be understood as a page that includes text, images, sound, and video, as well as a page where software for performing specific tasks, such as a web application, is loaded. Furthermore, the server may perform at least one of the functions of a web application server, a web server, a mobile server, and a database server on a single physical server, or it may be composed of and operated by multiple physically separated servers. However, it is not limited thereto, and the type of server can be varied to a level obvious to a person skilled in the art.

[0015] Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the attached drawings.

[0016] FIG. 1 illustrates a rent map generating device according to embodiments of the present invention. Referring to FIG. 1, the rent map generating device (100) can generate a rent map showing the average rent of each region. At this time, the average rent is based on the rent of residential buildings (e.g., officetels or apartments, etc.) within the region, and may be the average rent of residential buildings, but is not limited thereto.

[0017] A rent map generating device (100) according to embodiments of the present invention can divide a predetermined area (e.g., Seoul) into station areas adjacent to at least one subway station, calculate the average rent of each station area, and generate a rent map displaying the station area and the average rent. In particular, the rent map generating device (100) can generate a rent map based on rent information of lease contracts actually traded for buildings within each station area (i.e., actual transaction price information), and also has the effect of estimating the appropriate average rent of a station area based on actual transaction price information of buildings within surrounding station areas, even for station areas where the latest actual transaction price information is unavailable.

[0018] Meanwhile, in the present specification, the term "station area" refers to an area within a predetermined distance from at least one subway station, and, for example, may refer to an area located within 500m from at least one subway station. In addition, such station areas may be designated and partitioned in advance, and the rent map generating device (100) can identify each station area by utilizing such station area partitioning information.

[0019] The rental fee map generating device (100) is a device having a computational processing function, and may be, for example, a computing device (e.g., a server) including a processor and memory, but is not limited thereto.

[0020] The rent map generating device (100) includes a rent information collection unit (110), a converted rent calculation unit (120), a rent map generating unit (130), and a rent change rate estimation unit (140).

[0021] The rent information collection unit (110) collects rent information from lease contract information of buildings (officetels or apartments) located in a station area within a designated region. According to embodiments, the rent information collection unit (110) may collect actual transaction price data from the Ministry of Land, Infrastructure and Transport of the Republic of Korea. The actual transaction price data includes information regarding the address, landlord, tenant, lease contract period, rent, etc., of the buildings where actual lease contracts have been concluded.

[0022] According to embodiments, the rent information collection unit (110) may collect rent information including information on the transaction date, address, detailed address, building type, exclusive area, deposit, and monthly rent as illustrated in FIG. 2. Here, building type refers to information indicating whether the building is an officetel or an apartment.

[0023] For example, the rent information collection unit (110) can collect more than 400,000 actual transaction price data around 642 stations in the metropolitan area.

[0024] According to the embodiments, the rent information collection unit (110) can collect actual transaction price data for a predetermined period (e.g., 2 years).

[0025] The converted rent calculation unit (120) can calculate the converted rent for each lease contract using rent information collected from the rent information collection unit (110).

[0026] Lease contracts are broadly classified into monthly rent contracts, where only monthly rent is paid, and monthly rent contracts with a security deposit, where monthly rent is paid along with a specified deposit. In the case of monthly rent contracts with a security deposit, the security deposit may vary for each lease contract, and accordingly, the monthly rent paid together may also vary. Therefore, it is required to calculate a monthly rent based on a specified common standard, that is, a converted rent.

[0027] The converted rent calculation unit (120) calculates the converted rent by processing the actual transaction deposit and actual transaction rent of the lease contracts in the actual transaction price data according to a predetermined method, allocating a fixed deposit (i.e., converted deposit) according to the area and type of each building, and converting the remaining deposit difference into a monthly rent according to a predetermined conversion rate and adding it to the actual rent. Through this, various lease contracts can be compared under the same conditions.

[0028] For example, since it is difficult to compare a lease agreement with a security deposit of 30 million won and a monthly rent of 1 million won with a lease agreement with a security deposit of 60 million won and a monthly rent of 800,000 won, it means that the security deposits for the two lease agreements are unified (e.g., 20 million won, etc.), and the remaining difference in security deposits is converted into a monthly rent through a predetermined conversion rate and added to the existing monthly rent.

[0029] According to embodiments, the converted rent calculation unit (120) calculates the full rent by applying a predetermined conversion rate to each actual transaction deposit and summing the accompanying actual transaction rents. Additionally, the converted rent calculation unit (120) can calculate the converted rent by subtracting the monthly rent, calculated by applying the conversion rate to the converted deposit corresponding to the building subject to the lease agreement, from the calculated full rent. This converted rent is used to generate a rent map.

[0030] For example, the converted rent calculation unit (120) can calculate the full rent according to the following mathematical formula 1 and calculate the converted rent according to the following mathematical formula 2.

[0031]

[0032]

[0033] Here, R f is the full rent, and R ct is the actual transaction rent, and D ctis the actual transaction deposit, t is a predefined deposit-to-rent conversion rate, and R con is converted rent, D con is the converted deposit. In this case, the conversion rate t and the converted deposit D con It can be defined in advance according to the building type (whether it is an officetel or an apartment) and the exclusive area of ​​each building.

[0034] Finally, the converted deposit and converted rent according to the embodiments of the present invention can be determined as shown in FIG. 3. As shown in FIG. 3, the converted deposit and converted rent may vary depending on the building type (apartment / officetel) and the building area. Referring to FIG. 3, the rent map generating device (100) according to the embodiments of the present invention calculates the converted rent by unifying the deposit (converted deposit) when calculating the rent of a building in each station area, and converting and summing the remaining difference into monthly rent. Through this, there is an effect of being able to compare rents based on a more common standard.

[0035] For the following, unless otherwise stated, it is assumed that the rents shown on the rent map are converted rents.

[0036] Meanwhile, the converted rent calculation unit (120) can calculate the rate of change in rent for each segment defined by a predetermined time interval (e.g., one year, quarter, or month). At this time, the rate of change in rent represents the rate of change of the current segment rent relative to the previous segment rent, and can be calculated as a general ratio or as a compound annual growth rate, but is not limited thereto.

[0037] The rent map generation unit (130) can generate a rent map showing converted rents by exclusive area for each station area by using the converted rents of each actual transaction price data. For example, the rent map generation unit (130) generates a rent map showing converted rents for officetels and apartments of various exclusive areas.

[0038] The rent map generation unit (130) can generate rent map data that displays a station area map and converted rents for each station area. For example, the rent map generation unit (130) can generate rent map data that displays rents for officetels and apartments of various exclusive areas for each station area.

[0039] According to embodiments, the rent map generation unit (130) can generate rent map data based on the converted rent by exclusive area of ​​each building type (officetel / apartment) for each station area during a predetermined period. This will be described later.

[0040] The rent change rate estimation unit (140) can estimate the rent change rate from the present time to a predetermined future time. For example, the rent change rate estimation unit (140) can estimate the rent change rate after 5 years, but is not limited thereto.

[0041] The rent change rate estimation unit (140) of the embodiments of the present invention can estimate the rent change rate for a predetermined period in the future by utilizing the relationship between non-policy information related to the rental market or demographic information and policy information related to government policies regarding the rental market (or real estate market) and the rent change rate.

[0042] Non-policy information refers to information related to the rental market, such as the number of lease contracts, market interest rates, and the conversion rate from Jeonse to monthly rent, as well as demographic information, such as population size, household income, and the number of single-person households. Policy information refers to information regarding government policies concerning the rental market.

[0043] According to embodiments of the present invention, it is possible to estimate the rate of change in rent over a predetermined period by considering both policy information and non-policy information. Here, the rate of change in rent refers to the rate of change in rent of a predetermined area and may represent the average rate of change in rent of the entire Korean rental market, but is not limited thereto.

[0044] FIGS. 4 and 5 are drawings for illustrating a rent map according to embodiments of the present invention. The rent map of FIGS. 4 and 5 may be shown based on rent map data generated by a rent map generation unit (130). For example, a certain electronic device may receive rent map data and display a rent map based on the rent map data through a display of the electronic device.

[0045] Referring to Figure 4, the rent map shows a station area map (MAP) and average rent information (RMI).

[0046] A station area map (MAP) indicates the geographical location of a station area, and for example, as shown in Fig. 4, a station area map (MAP) indicating the locations of Sinsa Station, Nonhyeon Station, Hakdong Station, etc. in the Gangnam area of ​​Seoul is displayed on the rent map.

[0047] The average rent information (RMI) represents information on the average of the converted rent by exclusive area of ​​each station area. For example, as shown in Fig. 4, the average rent information (RMI) for the average rents of each station area (e.g., Sinsa Station, Nonhyeon Station, Hakdong Station, etc.) is displayed on the rent map.

[0048] Referring to Fig. 5, the average rent information displayed on the rent map is explained in detail. Referring to Fig. 5, information on the average rent by exclusive area and period for each station area can be displayed as a table. At this time, the information on the average rent can be displayed separately by building type. At this time, for the user's visual identification, an identifier 'A' can be displayed for apartments, and an identifier 'O' can be displayed for officetels.

[0049] According to embodiments of the present invention, the rent map generation unit (130) can determine the average rent by exclusive area and by period for each of the station areas based on the converted rent calculated by exclusive area for each of the officetels and apartments within the station area, and generate average rent information (RMI). For example, the rent map generation unit (130) can generate average rent information (RMI) including the average rent for each exclusive area of ​​the 'apartment' at 'Seonjeongneung Station' during the years 2023 to 2024. At this time, information regarding the average rent may be stored in a structure that includes attribute fields of station area name, building type, period, exclusive area, and rent within a predetermined database.

[0050] According to embodiments, the rent map generating unit (130) may generate a rent map such that the average rent for each exclusive area is displayed in a different color for the user's visual identification. For example, an exclusive area of ​​20m² 3 Average rents of less than 20m² are indicated in red, and exclusive area 20m² 3 more than 50m 3 Average rents of less than 50m² are displayed in green, and exclusive area 50m² 3 Over 84m 3 Average rents for less than 84m² are indicated in blue, and the exclusive area is 84m². 3 More than 135m 3 Average rents below a certain level may be displayed in black.

[0051] Accordingly, the user can use the rent map generated by the rent map generation unit (130) to check the average rent by building type and exclusive area in each station area during a specified period.

[0052] FIG. 6 is a diagram illustrating policy information and non-policy information according to embodiments of the present invention. Referring to FIG. 6, non-policy information (NPI) and policy information (PI) are shown. Meanwhile, in embodiments of the present invention, non-policy information (NPI) and policy information (PI) may exist in intervals defined by a predetermined time interval (e.g., one year, a quarter, or a month). For example, non-policy information (NPI) and policy information (PI) for each interval corresponding to one month may be calculated.

[0053] Non-policy information (NPI) is unrelated to policies regarding the housing or rental market and can represent, for example, market interest rates, the number of rental contracts, average household income, population size, etc. That is, an arbitrary interval (T i Market interest rate (r) for every i ), number of lease agreements (C i ), average household income(I i ) and population (P i ) may exist.

[0054] Policy Information (PI) quantifies the impact of policies on the housing or rental market within an arbitrary range, and represents, for example, the supply index representing supply-side impact, the demand index representing demand-side influence, the tax index representing tax-side influence, the financial index representing financial-side influence, and the rental index representing rental-side influence.

[0055] The supply index is an indicator representing the extent to which housing supply is affected by policy; for example, if there is a policy encouraging the supply of new housing, the supply index rises, while if there are regulations restricting reconstruction, the supply index may fall.

[0056] The demand index is an indicator representing the extent to which housing demand is affected by policy. For example, if there is a policy related to the transaction permit system, the demand index may decrease, while if there is an incentive policy for housing subscriptions, the demand index may increase.

[0057] The tax index is an indicator representing the extent to which taxes related to housing or rental transactions are affected by policy; for example, if there is a policy that increases the relevant taxes, the tax index may decrease.

[0058] The financial index is an indicator representing the extent to which financial conditions for housing or rental transactions are affected by policy; for example, if there is a policy regulating lending, the financial index may fall.

[0059] The rental index is an indicator representing the extent to which the rental market is influenced by policy, regardless of housing transactions; it is assumed that policies favorable to tenants have a positive effect. This is because, in reality, tenants have a significant influence on the rental market. For instance, if a policy strengthens the protection of rental deposits, the rental index rises.

[0060] Meanwhile, the supply, demand, tax, financial, and rental indices for each policy can be determined in advance. For example, they can be calculated by undergoing a quantitative evaluation by real estate or rental market experts.

[0061] Meanwhile, policies do not exist only at a specific time but can exist for a considerable period and become effective. Accordingly, one or more policies may coexist in a specific period. Accordingly, in the embodiments of the present invention, each period (T i When calculating policy information in ), the corresponding interval (T i It considers all policy information of policies that are valid (i.e., effectively effective) in ). That is, for each interval (T i The supply index, demand index, tax index, financial index, and rental index of ) are the corresponding interval (T i It can be calculated based on the supply index, demand index, tax index, financial index, and rental index of policies effectively in effect in ), respectively. For example, each interval (T i The supply index, demand index, tax index, financial index, and rental index of ) are the corresponding interval (T i It can be calculated as the sum of the supply index, demand index, tax index, financial index, and rental index of policies that are effectively in effect in ), but is not limited thereto.

[0062] FIG. 7 is a diagram illustrating a rent change rate estimation unit according to embodiments of the present invention. Referring to FIG. 7, the rent change rate estimation unit (140) inputs an input vector (VIN) into a neural network (141) that has been supervised learning in advance, and can obtain an output vector (VOUT) representing the rent change rate over a predetermined period as the output of the neural network. To this end, the neural network (141) can be supervised learning in advance.

[0063] According to embodiments of the present invention, the rent change rate estimation unit (140) can calculate the rent change rate N intervals after the current interval based on non-policy information and policy information of a total of K intervals including the current interval to which the current time belongs and (K-1) intervals past from the current interval. To this end, the neural network (141) can be trained using an input vector (VIN) calculated based on non-policy information and policy information in K intervals past, including the t-th interval among the data observed in the past, and an output vector (VOUT) corresponding to the rent change rate N intervals after the t-th interval (i.e., the t+N-th interval; where N is a natural number) as training data.

[0064] Here, the training data may be actual data collected (or observed) during each predetermined period. For example, the rate of change in rent for each interval corresponding to the output vector (VOUT) may be utilized from data calculated and stored by the converted rent calculation unit (120).

[0065] The neural network (141) is trained in advance with such training data, and when an input vector (VIN) calculated according to policy information and non-policy information in K intervals including the current time is input, it generates a corresponding output vector (VOUT) as an output.

[0066] According to embodiments of the present invention, an input vector (VIN) is determined based on policy information and non-policy information during a predetermined interval. For example, the input vector (VIN) can be defined according to Equation 3 below.

[0067]

[0068] Here, X i represents a non-policy state vector having non-policy information from the i-th interval as each component, and Y i represents a policy state vector having policy information from the i-th interval as each component, and w iis Y i It is the time decay coefficient for. For example, the components of the non-policy state vector can be market interest rates, number of contracts, household income, and population size, and the components of the policy state vector can be supply index, demand index, tax index, financial index, and rental index.

[0069] Consequently, the input vector (VIN) is determined based on the policy state vector and non-policy state vector of the past K consecutive intervals including the t-th interval, and the neural network (141) determines the output vector (VOUT) representing the rate of change in rent in the (t+N)-th interval from the input vector (VIN) thus determined. At this time, K and N can be fixed as constants and can be determined in advance and set in the neural network (141). In this case, various training data can be obtained by processing data from multiple intervals in a sliding window manner for the training of the neural network (141).

[0070] Meanwhile, Y i The time decay coefficient w for i It can be defined according to the following mathematical formula 4.

[0071]

[0072] Here, a is a positive constant. According to this, the smaller i is, for the past interval, w i is set small. Accordingly, Y corresponding to the policy of the past interval i Since the time decay factor for is set small, the policy effect in the past interval is reflected with greater reduction. This reflects the characteristics of the policy.

[0073] Alternatively, Y i The time decay coefficient w for i may also be defined according to the mathematical formula 5 below. The intent is the same, but a more precise reflection is possible.

[0074]

[0075] Here, a is a positive constant. According to this, the smaller i is, for the past interval, w i is set small. Accordingly, Y corresponding to the policy of the past interval i Since the time decay factor for is set small, the policy effect in the past interval is reflected with greater reduction. This reflects the characteristics of the policy.

[0076] According to embodiments of the present invention, it is possible to predict the rate of increase in rent at a specific future point in time by considering policy information and non-policy information, which are factors affecting the rent of each time segment divided into predetermined units in advance. In particular, according to embodiments of the present invention, it is possible to predict the rate of increase in rent more accurately by considering the attenuation effect over time together with non-policy information that also affects the sentiment of rent market participants. To this end, by selectively reflecting a attenuation coefficient in the input vector input to the neural network (141), it is possible to calculate a more accurate output vector (VOUT).

[0077] The neural network (141) may include an input layer (141a), a hidden layer (141b), and an output layer (141c). The input layer (141a) may be composed of a number of input nodes equal to the number of components (i.e., dimensions) of the input vector (VIN).

[0078] The input layer (141a) can transmit an input vector (VIN) to the hidden layer (141b). According to embodiments, the input layer (141a) may generate an intermediate vector by applying one or more connection strength values ​​corresponding to each of the input nodes and transmit the generated intermediate vector to the hidden layer (141b). One or more connection strength values ​​may be set to arbitrary initial values ​​and then continuously updated through supervised learning.

[0079] The hidden layer (141b) learns the characteristics of the input vector (VIN) (or intermediate vector) transmitted from the input layer (141a) and, as a result of the learning, can output a characteristic vector corresponding to the input vector (VIN). For example, the characteristic vector may be a value that reflects the relationship between the input vectors (VIN).

[0080] According to embodiments, the hidden layer (141b) includes a plurality of hidden nodes and can transmit a feature vector generated by applying one or more connection strengths corresponding to each of the hidden nodes to an input transmitted to the input layer (141a) to the output layer (141c). At this time, the initial values ​​of the one or more connection strength values ​​corresponding to each of the hidden nodes included in the hidden layer (141b) are set to arbitrary values ​​and then can be updated as training data is continuously supervised.

[0081] The output layer (141c) determines an output vector (VOUT) representing the rate of change in rent for a predetermined future period by applying an activation function to a feature vector received from the hidden layer (141b). At this time, the artificial neural network (141) is trained by comparing the output vector (VOUT) obtained from the output layer (141c) according to the input result of the training input data with the training output data that forms a pair. For example, the artificial neural network (141) can be supervised learning by continuously updating the connection strength values ​​of each node based on a loss function so that the loss function value between the output vector (VOUT), which is the predicted value of the training model, and the training output data is minimized.

[0082] In this way, the rent change rate estimation unit (140) can calculate the rent change rate in a predetermined future interval (time) after the present based on the output vector (VOUT) obtained as the output of the artificial neural network (141).

[0083] FIG. 8 is a diagram illustrating the hardware configuration of an operating server according to embodiments of the present invention. The electronic device (300) of FIG. 8 represents the rent map generating device (100) described with reference to FIG. 1 to 7.

[0084] Referring to FIG. 8, the electronic device (300) may include at least one processor (310) and a memory (320) that stores instructions that instruct the at least one processor (310) to perform at least one operation.

[0085] The above at least one operation is interpreted to include at least one of the operations of the aforementioned electronic device (300) or the operations of the functional part, and a specific description is omitted to prevent redundant explanation.

[0086] Here, at least one processor (310) may mean a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor on which methods according to embodiments of the present invention are performed.

[0087] The memory (320) may be composed of at least one of a volatile storage medium and a non-volatile storage medium. For example, the memory (320) may be composed of at least one of a read-only memory (ROM) and a random access memory (RAM).

[0088] Additionally, the electronic device (300) may include a transceiver (330) that performs communication via a wireless network. Additionally, the electronic device (300) may further include an input interface device (340), an output interface device (350), a storage device (360, which may be referred to interchangeably with internal storage), etc. Each component included in the electronic device (300) may be connected by a bus (370) to communicate with one another.

[0089] The methods according to the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., either alone or in combination. The program instructions recorded on the computer-readable medium may be those specifically designed and configured for the present invention, or they may be those known and available to those skilled in the art of computer software.

[0090] Examples of computer-readable media may include hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions may include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. The aforementioned hardware devices may be configured to operate as at least one software module to perform the operation of the present invention, and vice versa.

[0091] In addition, the above-described method or device may be implemented by combining all or part of its configuration or function, or by implementing it separately.

[0092] Although the present invention has been described above with reference to preferred embodiments, those skilled in the art will understand that various modifications and changes can be made to the invention without departing from the spirit and scope of the invention as described in the following claims.

Claims

Claim 1 A device for analyzing the rental market using artificial intelligence to predict the rate of change in rent, comprising: a rent information collection unit configured to collect rent information for rental contracts in which actual transactions have taken place; and a converted rent calculation unit configured to calculate a converted rent for each rental contract based on the collected rent information, and to calculate and store a rate of change in rent for each interval defined by a predetermined time interval. The method includes a rent change rate estimation unit configured to estimate the rent change rate in a predetermined future interval based on policy information and non-policy information of multiple past intervals including the current interval, wherein the non-policy information includes information on market interest rates, the number of lease contracts, average household income, and population size, and the policy information includes information on a supply index representing the supply-side influence of government policy, a demand index representing the demand-side influence, a tax index representing the tax-side influence, a financial index representing the financial-side influence, and a rental index representing the rental-side influence. The rent change rate estimation unit estimates the rent change rate using a supervised neural network that outputs an output vector corresponding to the rent change rate after N intervals (where N is a natural number) from the current interval as an output to an input vector calculated based on non-policy information and policy information of a total of K intervals (where K is a natural number) including the current t-th interval and (K-1) past intervals from the current t-th interval, wherein the input vector is defined according to the following Equation 1, [Equation 1] (In the above mathematical formula 1, X i represents a non-policy state vector having non-policy information from the i-th interval as each component, and Y i represents a policy state vector having policy information from the i-th interval as each component, and w i is Y i It is a time decay coefficient for, wherein the components of the non-policy state vector are the market interest rate, the number of lease contracts, the average household income, and the population, and the components of the policy state vector are the supply index, the demand index, the tax index, the financial index, and the rental index) The time decay coefficient is defined according to the following Equation 2, [Equation 2] (In the above mathematical formula 2, a is a positive constant, and as i becomes smaller, w i A device for predicting the rate of change in rent (set to small). Claim 2 delete Claim 3 delete

Images