Multi-source living body data processing method and system for real estate value assessment

Abstract

The application discloses a multi-source living body data processing method and system for real estate value evaluation, and particularly relates to the technical field of data processing; the method comprises the following steps: acquiring multi-source living body data in a real estate evaluation area, and constructing a standard space-time data set; performing short-period time series analysis based on the standard space-time data set, and generating a crowd flow intensity index in the real estate evaluation area; performing event correlation analysis on the standard space-time data set, and extracting a quantitative response index of events in the real estate evaluation area; performing attribution analysis on the crowd flow intensity index and the quantitative response index of events, and respectively obtaining an influence weight and a contribution weight; constructing a real estate value dynamic calibration model according to the influence weight and the contribution weight; and using the real estate value dynamic calibration model to adjust a real estate evaluation result of the real estate evaluation area in real time; and the real-time performance and dynamic adaptability of real estate value evaluation are improved.

Description

Technical Field

[0001] This invention relates to the field of anomaly identification and early warning technology, and more specifically, to a method and system for processing multi-source liveness data for real estate valuation. Background Technology

[0002] Existing property valuation methods primarily rely on static data such as historical transaction data and publicly listed properties, making them suitable for periods of market stability. However, these methods suffer from a mismatch in time resolution between data sources and valuation techniques, resulting in valuations that fail to effectively reflect short-term fluctuations in the property market in real time. Consequently, property valuations struggle to accurately reflect the real-time market conditions and true value of properties within the assessed area. Summary of the Invention

[0003] In order to overcome the above-mentioned deficiencies of the prior art, embodiments of the present invention provide a multi-source live data processing method and system for real estate valuation to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A multi-source liveness data processing method for property valuation includes the following steps:

[0006] S1: Obtain multi-source liveness data within the property assessment area, and perform time alignment and spatial normalization on the multi-source liveness data to generate a standard spatiotemporal dataset;

[0007] S2: Perform short-period time series analysis on the standard spatiotemporal dataset to extract the changing trend of population flow intensity within the property assessment area and obtain the population flow intensity index within the property assessment area.

[0008] S3: Perform event correlation analysis on standard spatiotemporal datasets, and extract quantitative response indicators of events within the property valuation area by setting event windows;

[0009] S4: Conduct attribution analysis on the population flow intensity index and the quantitative response index of events within the property valuation area to determine the weight of the impact of population flow changes on short-term property prices and the weight of the contribution of events to the phased fluctuations of property prices.

[0010] S5: Establish a dynamic calibration model for real estate value based on the influence weight and contribution weight;

[0011] S6: Based on the property value dynamic calibration model, the property valuation of the property assessment area is calibrated and drift corrected, and the property valuation results of the property assessment area are adjusted in real time.

[0012] In a preferred embodiment, S1 specifically refers to:

[0013] The first dataset is obtained by collecting mobile terminal signaling data, real-time population heat map data, real-time listing and transaction data, and remote sensing data on changes in building activities within the real estate assessment area.

[0014] The first dataset is time-aligned using a unified timestamp, and a time-consistent dataset is formed by linear interpolation and filling in missing values.

[0015] The geographic coordinates in the time-consistent dataset are converted into a unified geographic reference system, and the spatial resolution of each data source is rasterized and normalized to generate a spatially consistent dataset.

[0016] Organize spatially consistent data sets according to time series order and spatial grid index, and output standard spatiotemporal datasets.

[0017] In a preferred embodiment, S2 specifically refers to:

[0018] Population heat map data and mobile terminal signaling data within the property assessment area are read from the standard spatiotemporal dataset, and short-period sequences are formed through resampling;

[0019] By performing sliding window averaging and differencing on short-period sequences, a population flow measurement sequence is obtained.

[0020] Applying Fast Fourier Transform to the population flow measurement sequence separates the daily and weekly periodic components, resulting in a periodic decomposition sequence.

[0021] The first-order difference of the periodic decomposition sequence is calculated and linear regression is performed to extract the trend curve of population flow intensity.

[0022] Based on the mean and coefficient of variation of the population flow intensity trend curve within a preset time window, a population flow intensity index is generated for the property assessment area.

[0023] In a preferred embodiment, S3 specifically refers to:

[0024] Establish event windows for policy release, regional planning adjustments, and supporting infrastructure construction, and determine the start time and duration of each event window;

[0025] Retrieve real-time listing transaction data and remote sensing data of building activity changes that fall into each event window from the standard spatiotemporal dataset to generate a subset of event window data;

[0026] Construct a price comparison sequence for a subset of event window data in the order of event occurrence, and calculate the interval average and the maximum deviation of the interval to form a vector of the impact of event prices;

[0027] The vector of the price impact of an event and the duration of the corresponding event window are normalized to output a quantitative response index for the event within the property valuation area.

[0028] In a preferred embodiment, S4 specifically refers to:

[0029] Extract short-period real estate price series that are consistent with the time of population flow intensity indicators, and construct population flow-price pairing series;

[0030] Based on the population flow-price pairing sequence, a mapping model between the population flow intensity index and short-term real estate prices is established using the multiple linear regression method, and the weight of the impact of population flow changes is calculated.

[0031] Extract the real estate price phase sequence that is consistent with the time of the event quantitative response indicator, and construct the event impact-price pairing sequence;

[0032] Based on the event impact-price pairing sequence, a mapping model between the quantitative response index of events and the phased fluctuations of real estate prices is established using the difference regression analysis method, and the contribution weight of event prices is calculated.

[0033] In a preferred embodiment, S5 specifically refers to:

[0034] Initialize the parameters of the property value dynamic calibration model, including the basic valuation deviation coefficient, the population flow adjustment coefficient, and the event contribution adjustment coefficient;

[0035] Read the benchmark valuation sequence, real-time property price observations within the property valuation area, the weight of the impact of population flow changes, and the weight of the price contribution of events to construct a calibration training dataset;

[0036] For the calibration training dataset, the least squares method is used to optimize the parameters of the property value dynamic calibration model;

[0037] Save the optimized parameters of the property value dynamic calibration model and output the property value dynamic calibration model.

[0038] In a preferred embodiment, S6 specifically refers to:

[0039] The benchmark valuation series is initially adjusted based on the basic valuation deviation coefficient to obtain the adjusted real estate valuation series.

[0040] The population flow adjustment value is calculated based on the population flow adjustment coefficient and the weight of the impact of population flow changes.

[0041] The event adjustment value is calculated based on the event contribution adjustment coefficient and the event price contribution weight.

[0042] The calibrated property valuation results are obtained by integrating the basic adjusted property valuation series, the population flow adjusted value, and the event adjusted value.

[0043] Update the property valuation results for the property valuation area based on the calibrated property valuation results.

[0044] On the other hand, the present invention provides a multi-source liveness data processing system for real estate valuation, comprising:

[0045] Data processing module: Acquires multi-source liveness data within the property valuation area, performs time alignment and spatial normalization on the multi-source liveness data, and generates a standard spatiotemporal dataset;

[0046] Population Trends Module: Performs short-period time-series analysis on standard spatiotemporal datasets to extract the changing trend of population flow intensity within the property assessment area and obtain population flow intensity indicators within the property assessment area;

[0047] Event Response Module: Performs event correlation analysis on standard spatiotemporal datasets and extracts quantitative response indicators for events within the property valuation area by setting event windows;

[0048] Weighted Attribution Module: Performs attribution analysis on the population flow intensity index and the quantitative response index of events within the property assessment area to determine the weight of the impact of population flow changes on short-term property prices and the contribution weight of events to the phased fluctuations of property prices.

[0049] Calibration Modeling Module: Establishes a dynamic calibration model for property value based on influence weights and contribution weights;

[0050] Valuation Adjustment Module: Based on the property value dynamic calibration model, the valuation of the property assessment area is calibrated and drift corrected, and the property valuation results of the property assessment area are adjusted in real time.

[0051] The technical effects and advantages of this invention regarding the multi-source liveness data processing method and system for real estate valuation are as follows:

[0052] By performing unified spatiotemporal preprocessing on multi-source liveness data within the property assessment area, the completeness and consistency of the multi-source liveness data are significantly improved. Short-cycle time-series analysis based on standard spatiotemporal datasets enables timely capture of population flow trends, providing accurate indicators for quantitative assessment. By setting event windows to extract quantitative response indicators, the impact of policy, planning, and supporting infrastructure events can be objectively quantified. Attribution analysis using population flow intensity indicators and event quantitative response indicators within the property assessment area determines the contribution of different factors to short-cycle price fluctuations. A dynamic calibration model for property value is constructed based on influence and contribution weights to ensure that the valuation algorithm is synchronized with market dynamics. Applying this dynamic calibration model calibrates and corrects drift in the property assessment area's valuation, adjusting the valuation results in real time. This ensures that the property valuation results accurately reflect the latest market supply and demand structure and short-term fluctuations, effectively improving the real-time performance, accuracy, and reliability of the valuation. Attached Figure Description

[0053] Figure 1 This is a schematic diagram of the multi-source liveness data processing method for real estate valuation of this invention;

[0054] Figure 2 This is a schematic diagram of the multi-source live data processing system for real estate valuation according to the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0056] Example 1

[0057] Figure 1 This invention presents a multi-source liveness data processing method for real estate valuation, which includes the following steps:

[0058] S1: Obtain multi-source liveness data within the property assessment area, and perform time alignment and spatial normalization on the multi-source liveness data to generate a standard spatiotemporal dataset;

[0059] S2: Perform short-period time series analysis on the standard spatiotemporal dataset to extract the changing trend of population flow intensity within the property assessment area and obtain the population flow intensity index within the property assessment area.

[0060] S3: Perform event correlation analysis on standard spatiotemporal datasets, and extract quantitative response indicators of events within the property valuation area by setting event windows;

[0061] S4: Conduct attribution analysis on the population flow intensity index and the quantitative response index of events within the property valuation area to determine the weight of the impact of population flow changes on short-term property prices and the weight of the contribution of events to the phased fluctuations of property prices.

[0062] S5: Establish a dynamic calibration model for real estate value based on the influence weight and contribution weight;

[0063] S6: Based on the property value dynamic calibration model, the property valuation of the property assessment area is calibrated and drift corrected, and the property valuation results of the property assessment area are adjusted in real time.

[0064] S1: Acquire multi-source liveness data within the property valuation area, and perform temporal alignment and spatial normalization on the multi-source liveness data to generate a standard spatiotemporal dataset, including:

[0065] The first dataset is obtained by collecting mobile terminal signaling data, real-time population heat map data, real-time listing and transaction data, and remote sensing data on changes in building activities within the real estate assessment area.

[0066] A property valuation area is a designated spatial region used for property valuation. The boundaries of the area can be determined based on administrative divisions, such as a specific area or street in a city, or based on natural geographical boundaries or artificial boundaries, such as rivers, roads, or pre-defined spatial boundary coordinates. For example, when valuing property in a specific urban area, the property valuation area is all the spatial range within the administrative boundaries of the urban area.

[0067] Mobile terminal signaling data refers to data records generated when users use communication terminal devices to establish wireless communication connections with mobile communication base stations. This includes information such as the base station area identifier of the user's terminal device location and the timestamp of connecting to or switching base stations. Mobile terminal signaling data can reflect the flow and distribution of population within a property assessment area. For example, if multiple communication base stations exist near a residential area, and a large number of user terminal devices move from one base station's service area to another, the mobile terminal signaling data can reflect the direction and intensity of user movement. Real-time population heatmap data is data obtained based on wireless communication networks and geographic information systems. It mainly represents the population density at specific locations within a property assessment area at a specific time. For example, on a geospatial map, color gradients are used to display the population density of each area. Real-time listing and transaction data is information obtained in real-time from real estate transaction management platforms or publicly available housing transaction platforms of real estate agencies, including but not limited to the location, area, price, unit type, building age, listing date, and transaction price of listed properties. Real-time listing and transaction data can reflect the current market's acceptance of housing prices and the market supply and demand relationship. Remote sensing data on changes in building activities refers to remote sensing imagery data that uses remote sensing satellites or drones to observe land use changes, construction progress, and new building facilities within a property assessment area. For example, in a certain area, continuous observation of satellite imagery data can reflect the construction progress of buildings from planning to basic construction to topping out, as well as the geographic spatial distribution of new infrastructure.

[0068] The first dataset is time-aligned using a unified timestamp, and a time-consistent dataset is formed by linear interpolation and filling in missing values.

[0069] A unified time stamping standard is established, adopting a time stamping accuracy of seconds, meaning data acquisition is marked in units of one second. For each data record in the first dataset, alignment is performed based on the data acquisition time and the unified time stamping standard. For example, if the time of mobile terminal signaling data records is not synchronized with the time of real-time population heatmap data, a linear interpolation method is used to interpolate the data between known sampling times. This involves calculating the data value at the interpolation time based on the data values ​​at the preceding and following known times. The linear interpolation method involves subtracting the data value at the preceding known time from the data value at the following known time to obtain the difference, and then dividing the difference by the difference between the preceding and following known times. The time interval is used to obtain the change per unit time. This change per unit time is then multiplied by the time difference between the inserted time and the previous known time, and added to the data value of the previous known time to obtain the estimated data value for the inserted time. For missing data, such as when there is no corresponding data record for a certain time, linear interpolation is used to calculate the data value for filling the missing time using existing data points before and after the interpolation, based on the above interpolation calculation method. After interpolation and filling, a time-consistent data set with continuous and consistent timestamps is finally generated. The time-consistent data set ensures the synchronization of the time dimension between various data sources and ensures that all data sampling records have the same timestamp.

[0070] The geographic coordinates in the time-consistent dataset are converted into a unified geographic reference system, and the spatial resolution of each data source is rasterized and normalized to generate a spatially consistent dataset.

[0071] Define a unified spatial coordinate standard, i.e., a unified geographic reference system, such as the WGS84 coordinate system as the base coordinate system. For spatial location information from different data sources in a time-consistent dataset, such as mobile terminal signaling data and real-time population heat map data, different geographic coordinate representations may be used. First, the geographic coordinates in all data are uniformly transformed to the WGS84 coordinate system according to spatial coordinate transformation rules. Then, spatial resolution rasterization and normalization processing is performed, that is, the property assessment area is divided into spatial grid units of uniform size. For example, the property assessment area is divided into spatial grid units with fixed length and width. The data in each grid unit is statistically summarized, such as calculating the average value or density value of the data in the grid unit, thereby forming a data record with a unified spatial resolution. After spatial coordinate transformation and gridding processing, a spatially consistent dataset is finally obtained. All data records in the spatially consistent dataset have unified spatial coordinates and a unified spatial resolution.

[0072] Organize spatially consistent data sets according to time series order and spatial grid index, and output standard spatiotemporal datasets;

[0073] The data records in the spatially consistent dataset are sorted in chronological order, and a spatial grid index number is assigned to each spatial grid cell. That is, a unified grid cell numbering method is adopted, and each spatial grid cell is assigned a unique spatial location number. The data is then organized in the order of spatial grid index and time series sorting. Each data record contains a unified timestamp, spatial grid index and corresponding data source identifier and value; finally, a standard spatiotemporal dataset is generated.

[0074] S2: Perform short-period time-series analysis on the standard spatiotemporal dataset to extract the changing trend of population flow intensity within the property assessment area, and obtain population flow intensity indicators within the property assessment area, including:

[0075] Population heat map data and mobile terminal signaling data within the property assessment area are read from the standard spatiotemporal dataset, and short-period sequences are formed through resampling;

[0076] To analyze the short-term trends in population movement within a property valuation area, it is necessary to read population heatmap data and mobile terminal signaling data within a specific time range from a standard spatiotemporal dataset. For example, this could involve selecting a period of several consecutive days or weeks and reading the population heatmap data and mobile terminal signaling data at a high frequency of minutes or seconds. Resampling involves resampling the original data at a new, uniform time interval. For instance, if the original sampling frequencies of the population heatmap data and mobile terminal signaling data are inconsistent—with the population heatmap data collected every minute and the mobile terminal signaling data every thirty seconds—then during resampling, a uniform sampling interval of one minute can be used. The mobile terminal signaling data within each minute can be statistically analyzed, such as calculating the average or cumulative value of the mobile terminal signaling data within each minute. This ensures consistency with the sampling frequency of the population heatmap data, forming a unified short-term sequence.

[0077] By performing sliding window averaging and differencing on short-period sequences, a population flow measurement sequence is obtained.

[0078] In a short-period series, a fixed-length time window is set. Each time, all data records within a time window are taken, the sum of all data records within the window is calculated, and this sum is divided by the total number of data records to obtain the average value within the window. The window is then moved forward by a fixed time interval, and the above averaging calculation is performed again to obtain a new average value. This process continues until all data in the short-period series has been processed, resulting in a series of smoothed data values ​​calculated using the sliding window averaging method. For example, if the window length is several minutes, and the window is moved by several minutes each time, each calculation yields the average number of people or mobile devices for the corresponding time period. The data sequence after sliding window averaging is then differentially processed, specifically by calculating the difference between data values ​​at two adjacent time points. All the calculated difference values ​​are combined to form a differentially processed data sequence, called the population flow measurement sequence. For example, differential calculation can visually reflect the changes in population flow over a continuous time period, showing the increase or decrease in the population within the property assessment area over various time periods.

[0079] Applying Fast Fourier Transform to the population flow measurement sequence separates the daily and weekly periodic components, resulting in a periodic decomposition sequence.

[0080] The Fast Fourier Transform (FFT) method is used to transform the population flow measurement sequence from the original time domain to the frequency domain. The original time-ordered population flow measurement sequence data is converted into a frequency-based data representation. In this frequency-based representation, data components corresponding to different frequencies can be obtained; for example, the daily cycle component represents the frequency components that repeat regularly each day, and the weekly cycle component represents the frequency components that repeat regularly each week. Frequency filtering is then applied to the frequency-domain transformed data sequence to extract the daily and weekly cycle components. An inverse FFT is then performed on these components, transforming them back from the frequency domain to the time domain, resulting in data sequences containing only daily and weekly cycle patterns, respectively. These two data sequences are then combined and output as a periodic decomposition sequence. For example, the daily cycle component reflects the periodicity of population activity within a fixed time period each day, while the weekly cycle component reflects the periodicity of population activity on specific days each week.

[0081] The first-order difference of the periodic decomposition sequence is calculated and linear regression is performed to extract the trend curve of population flow intensity.

[0082] The periodic decomposition sequence is subjected to first-order differencing, which involves calculating the difference between data values ​​at adjacent time points in the periodic decomposition sequence to obtain the differencing data sequence. Linear regression fitting is then applied to the differencing data sequence, using the least squares method to calculate a linear function that minimizes the sum of the squared differences between the predicted and actual data sequences. The linear function is calculated by squaring the error between each data value and the predicted value, summing all the squared errors, and selecting appropriate linear function parameters to minimize the total sum of squared errors, thus obtaining the optimal fitting function. The fitted linear function reflects the overall trend of the data sequence and is called the population flow intensity trend curve. For example, the fitted trend curve can show whether the population flow within the property assessment area is increasing, decreasing, or remaining stable overall.

[0083] Based on the mean and coefficient of variation of the population flow intensity trend curve within a preset time window, a population flow intensity index is generated for the property assessment area.

[0084] A preset time window is set on the trend curve of population flow intensity. The window length is set according to the needs of property valuation, such as several hours, one day, or several days. The average value and coefficient of variation of the trend curve within each time window are calculated. The average value is calculated by adding all data values ​​of the trend curve within each time window and dividing by the total number of data values ​​within the time window. The coefficient of variation is calculated by summing the squares of the differences between all data values ​​within the window and the average value within the time window, dividing the sum of squares by the total number of data values ​​to obtain the square root, and then dividing the square root by the average value within the window. The average value and coefficient of variation calculated within the window are used as statistical features to describe the intensity and fluctuation of population flow in the property valuation area, and the output is the population flow intensity index of the property valuation area. For example, when the population flow in the area tends to be stable and the change is small, the coefficient of variation value is low; when the population flow in the area changes drastically, the coefficient of variation value is high.

[0085] S3: Perform event correlation analysis on the standard spatiotemporal dataset. By setting an event window, extract quantitative response indicators for events within the property valuation area, including:

[0086] Establish event windows for policy release, regional planning adjustments, and supporting infrastructure construction, and determine the start time and duration of each event window;

[0087] The policy release event window is an analytical time interval set for the timing of policy releases affecting property prices within the property valuation area. For example, when a city in the property valuation area introduces real estate control policies such as purchase restrictions, sales restrictions, or tax adjustments, the start date can be set as the policy release date, and the duration can be selected from several days to several weeks to observe the short-term impact of real estate control policies on the property transaction market within the property valuation area. The regional planning adjustment event window is an analytical time interval set for urban planning adjustments or land use planning changes within the property valuation area. For example, when the land use plan within the property valuation area is changed from industrial land to residential land, or when new commercial facilities or schools are planned near residential areas, the start date of the planning adjustment announcement or the official public announcement of the planning scheme can be used as the event window start date. A reasonable period after the planning adjustment announcement can be selected as the window duration to observe the changing trends in the impact of planning changes on property prices. The supporting infrastructure construction event window is a time interval set for analyzing the construction or commissioning of infrastructure or supporting facilities within a property valuation area. For example, after a newly built rail transit line, hospital, school, or commercial complex is completed and put into use within the property valuation area, the date of completion or official operation of the supporting facility is selected as the starting time, and a certain time length is selected as the observation window to study the actual impact of the supporting facility on the real estate market. Therefore, the starting time of each event window is marked with the date of the event, and the duration is reasonably set according to the event type and actual analysis needs.

[0088] Retrieve real-time listing transaction data and remote sensing data of building activity changes that fall into each event window from the standard spatiotemporal dataset to generate a subset of event window data;

[0089] Based on the defined start time and duration of each event window, data records corresponding to the time range of the event window are extracted from the standard spatiotemporal dataset. For example, taking the policy release event window as an example, if the start time of the policy release event window is a specific date, then all real-time listing transaction data and remote sensing data on changes in construction activities within a certain duration after the specific date are extracted from the standard spatiotemporal dataset to form a data subset corresponding to the event window. Similarly, for regional planning adjustment event windows and supporting facility construction event windows, real-time listing transaction data and remote sensing data on changes in construction activities within the corresponding time range are extracted from the standard spatiotemporal dataset according to their respective start times and durations, and data subsets for different event windows are formed respectively. The event window data subsets contain the timestamps of data collection and spatial location index information to ensure that the data can be accurately traced and located.

[0090] Construct a price comparison sequence for a subset of event window data in the order of event occurrence, and calculate the interval average and the maximum deviation of the interval to form a vector of the impact of event prices;

[0091] The real-time listing transaction data within each event window's data subset are sorted according to the time sequence within the event window to form a price comparison sequence. For example, taking a policy release event window as an example, transaction price information is arranged in chronological order before and after the policy release, recording property price data at consecutive time points before and after the start time of the event window to form a sequence of data reflecting the time-varying trend of property prices. The interval average and maximum interval deviation of the sequence data are calculated based on the price comparison sequence. The interval average is calculated by summing the price data at all times within the event window and then dividing by the total number of price data points within the event window to obtain the average price. The maximum interval deviation is calculated by calculating the absolute value of the difference between the price data at each time point within the event window and the interval average, and the largest absolute value of the difference is taken as the maximum interval deviation. For example, after the announcement of a regional planning adjustment, the market prices of listed properties within the region are recorded at various times over several consecutive weeks. Through the above calculation methods, the average listing price and the maximum price change of listed properties within the event window can be obtained. The calculated interval average and the maximum deviation of the interval are combined to form a vector that reflects the characteristics of price changes. This vector is called the event price impact magnitude vector, which is used to quantitatively describe the degree of impact of events on real estate prices within the event window.

[0092] The vector of event price impact magnitude and the duration of the corresponding event window are normalized to output a quantitative response index for events within the property valuation area.

[0093] Each component of the event price impact magnitude vector is divided by the duration of the event window. For example, if the event window lasts for several days, each component of the event price impact magnitude vector is divided by the duration of the event window. The event price impact magnitude vector within each event window is adjusted to a uniform time scale to facilitate comparison across events. The normalized event price impact magnitude vector is the quantitative response index of events within the property valuation area, quantifying the degree of response of each event to changes in property prices within the property valuation area.

[0094] S4: Conduct attribution analysis on the population flow intensity index and the quantitative response index of events within the property valuation area to determine the weight of the impact of population flow changes on short-term property prices and the contribution weight of events to the phased fluctuations of property prices, including:

[0095] Extract short-period real estate price series that are consistent with the time of population flow intensity indicators, and construct population flow-price pairing series;

[0096] In a standard spatiotemporal dataset, data records of property prices at corresponding times are retrieved based on the timestamp of the population flow intensity index. For example, if the standard spatiotemporal dataset contains population flow intensity index data for a specific day or hour, the corresponding property listing or transaction price information is simultaneously retrieved from the dataset. Data records with the same timestamp are paired, and each pair includes a population flow intensity index and a property price at the corresponding time, forming a population flow-price pairing sequence. For instance, within a certain region, the population flow intensity index and property price are recorded daily for several consecutive days. The population flow intensity index for each day is paired with the property price for the same day, thus obtaining a population flow-price pairing sequence that reflects the relationship between changes in population flow and changes in property prices.

[0097] Based on the population flow-price pairing sequence, a mapping model between the population flow intensity index and short-term real estate prices is established using the multiple linear regression method, and the weight of the impact of population flow changes is calculated.

[0098] A regression model is constructed using the population flow intensity index as the influencing variable (independent variable) and the short-term housing price series as the affected variable (dependent variable). In the regression model, housing price is the predicted variable, and the population flow intensity index is the predictor variable explaining changes in housing price. The regression coefficients are calculated using the least squares method, that is, the error between the model's predicted housing price and the actual housing price is calculated for all paired data, the squares of all errors are summed, and a set of the most suitable coefficient parameters is selected to minimize the sum of squared errors, thus obtaining a linear regression mapping model. For example, if housing price increases with the increase of population flow intensity, the regression coefficients in the linear regression mapping model show a positive relationship. Finally, the influence weight of population flow changes is calculated based on the obtained mapping model, which is the regression coefficient in the linear regression mapping model. The regression coefficient reflects the magnitude of the housing price change caused by each unit change in the population flow intensity index. For example, the increase in housing price when the population flow intensity index increases is reflected by the regression coefficient and is called the influence weight of population flow changes.

[0099] Extract the real estate price phase sequence that is consistent with the time of the event quantitative response indicator, and construct the event impact-price pairing sequence;

[0100] Events include, but are not limited to, policy releases, regional planning adjustments, and the completion of supporting infrastructure construction. The property price phase sequence refers to the short-term changes in property listing prices or transaction prices over a period of time following the occurrence of the corresponding event. To construct an event impact-price pairing sequence, it is necessary to ensure that the time scale of the event's quantitative response indicators corresponds to the property price phase sequence, meaning that the data records have the same timestamp. For example, within a continuous timeframe following a policy release, the event's quantitative response indicators and the corresponding property price data for each moment are recorded, forming a data pair between each event's quantitative response indicator data record and the property price data record at the same moment. For instance, after a city announces a new purchase restriction policy, for several consecutive days or weeks, the quantitative response indicators of the policy release to property prices and the daily property price changes are recorded, and the daily data records are paired to form an event impact-price pairing sequence that reflects the relationship between the event's impact and property price changes.

[0101] Based on the event impact-price pairing sequence, a mapping model between the quantitative response index of events and the phased fluctuations of real estate prices is established using the difference regression analysis method, and the contribution weight of event prices is calculated.

[0102] The event-impact-price pairing sequence of real estate price data is differentially processed, that is, the difference between the real estate price at each time point and the previous time point is calculated to form a differential real estate price fluctuation sequence. Using the event's quantitative response index as the influencing variable (independent variable) and the differential real estate price fluctuation sequence as the affected variable (dependent variable), a regression analysis model is established. The regression coefficient parameters are optimized using the least squares method to minimize the sum of squared errors between the predicted and observed real estate price fluctuations. Specifically, the regression analysis model calculates by squaring the difference between the predicted and actual values, summing all squared values, and selecting appropriate regression coefficient parameters to minimize the total sum of squared errors. Finally, the regression coefficients obtained from the optimized regression analysis model reflect the degree of influence of the event on the phased fluctuations of real estate prices. The regression coefficients represent the contribution weights of the event to the price, reflecting the degree of influence of the event's changes on real estate price fluctuations. For example, when a regional planning adjustment event leads to the construction of new public facilities in the region, the weight coefficient of the regional planning adjustment event on the impact on real estate prices is relatively large, indicating that the regional planning adjustment event has a significant contribution to changes in real estate prices.

[0103] S5: Based on the influence weight and contribution weight, establish a dynamic calibration model for real estate value, including:

[0104] Initialize the parameters of the property value dynamic calibration model, including the basic valuation deviation coefficient, the population flow adjustment coefficient, and the event contribution adjustment coefficient;

[0105] The dynamic calibration model for property value is used to correct and calibrate the market valuation of properties within a property assessment area in real time, ensuring that the valuation results fully reflect the impact of actual market conditions and event changes on property prices. To ensure the effective operation of the dynamic calibration model, it is first necessary to set and initialize its core parameters. The basic valuation deviation coefficient describes the average deviation between the market valuation result and the actual market price under normal conditions, without the influence of population movement or event shocks. For example, within the property assessment area, the average value reflecting the normalized valuation deviation is obtained by analyzing the difference between property valuation and market transaction prices based on long-term historical data, serving as the initial parameter setting for the basic valuation deviation coefficient. The population flow adjustment coefficient represents the degree of impact of changes in population flow intensity on property market prices within the property assessment area. For example, within the property assessment area, when the resident population increases significantly and housing demand rises rapidly, the property market typically experiences a supply shortage, leading to rising property prices. In this case, the population flow adjustment coefficient is set to reflect a significant positive impact of population flow on property market price changes. The event contribution adjustment coefficient describes the degree of impact of events on property price changes. When initializing the basic valuation deviation coefficient, the population flow adjustment coefficient, and the event contribution adjustment coefficient, it is necessary to base them on reasonable statistical data analysis and determine the initial values ​​based on historical data analysis or market expert experience.

[0106] Read the benchmark valuation sequence, real-time property price observations within the property valuation area, the weight of the impact of population flow changes, and the weight of the price contribution of events to construct a calibration training dataset;

[0107] The benchmark valuation series is a basic valuation prediction series of property prices within a property valuation area, based on a dynamic calibration model of property value without considering the impact of population movement or events. It is typically estimated based on long-term historical data and market trends; for example, it consists of a continuous time series composed of the historical average or trend prediction prices of long-term market prices within the property valuation area. Real-time property price observations, on the other hand, are data series formed by recording and observing actual listing or transaction prices of properties within the property valuation area in real time. For example, listing or transaction prices are obtained daily or at specific times through real estate transaction platforms and recorded as real-time observations. The population movement change impact weight is a data parameter used to describe the degree to which changes in population movement within the area affect property prices. The event price contribution weight reflects the degree to which an event affects the periodic fluctuations in property prices. To construct a calibration training dataset, the baseline valuation sequence, real-time property price observations, population flow change impact weights, and event price contribution weights within the same time interval are sequentially read from the standard spatiotemporal dataset. Data records with the same timestamp are paired to form training data records. For example, taking a month's worth of market data for a property valuation area, the daily baseline valuation, the corresponding daily real-time property price observations, the daily population flow change impact weights, and the daily event price contribution weights are read and matched according to the corresponding time. Each set of data with the same timestamp forms a calibration training data record. All data records for a consecutive month are organized to form a calibration training dataset that can be used for model parameter optimization.

[0108] For the calibration training dataset, the least squares method is used to optimize the parameters of the property value dynamic calibration model;

[0109] Using the calibration training dataset as input data, a least squares optimization method is employed to jointly optimize the baseline valuation deviation coefficient, the crowd flow adjustment coefficient, and the event contribution adjustment coefficient. Based on the baseline valuation sequence, the weight of crowd flow changes, and the weight of event price contribution in each set of calibration training data records, the predicted calibration property price is calculated using these coefficients. For example, the baseline valuation deviation coefficient is multiplied by the baseline valuation, the crowd flow adjustment coefficient is multiplied by the weight of crowd flow changes, and the event contribution adjustment coefficient is multiplied by the weight of event price contribution, and these products are then summed to obtain the predicted calibration property price. The predicted calibration property price is compared with the observed real-time property price to calculate the difference, and then the difference is squared to obtain the squared error value. The squared error values ​​of all calibration training data records are summed to obtain the overall sum of squared errors. The least squares optimization method iteratively adjusts the baseline valuation deviation coefficient, the crowd flow adjustment coefficient, and the event contribution adjustment coefficient to minimize the overall sum of squared errors, thereby obtaining the optimal set of model parameters.

[0110] Save the optimized parameters of the property value dynamic calibration model and output the property value dynamic calibration model.

[0111] The optimized baseline valuation deviation coefficient, population flow adjustment coefficient, and event contribution adjustment coefficient are saved as formal parameters of the property value dynamic calibration model. The property value dynamic calibration model with the optimized parameters is used to calibrate and correct the property market valuation in the property valuation area in real time. For example, when the property price observation value, the influence weight of population flow changes, and the event price contribution weight are obtained in real time, the property market valuation can be dynamically corrected and calibrated in real time through the property value dynamic calibration model with the optimized parameters.

[0112] S6: Based on the dynamic calibration model for property value, the valuation of the property assessment area is calibrated and drift corrected, and the property valuation results of the property assessment area are adjusted in real time, including:

[0113] The benchmark valuation series is initially adjusted based on the basic valuation deviation coefficient to obtain the adjusted real estate valuation series.

[0114] In property valuation, the benchmark valuation series is a set of property price data obtained from long-term historical data or market average price trend predictions. It reflects the predicted market price for the property valuation area without considering population movement and event impacts. The basic valuation deviation coefficient is an optimized model parameter used to describe the long-term stable deviation between the benchmark valuation series and the actual observed prices in the property market. For example, if the long-term historical data of the property valuation area shows that the benchmark valuation series of the property valuation system is consistently higher or lower than the actual market transaction price, the basic valuation deviation coefficient is used to correct this valuation deviation. The adjustment method for the basic valuation deviation coefficient is to multiply the data value at each moment in the benchmark valuation series by the basic valuation deviation coefficient, thus obtaining the adjusted property valuation series. For example, within the property valuation area, if the basic valuation deviation coefficient indicates that the benchmark valuation needs to be reduced by a certain proportion, then by multiplying each data value in the benchmark valuation series at each moment by the basic valuation deviation coefficient, a pre-calibrated adjusted property valuation series can be obtained, which better reflects the long-term observed price change trend in the property market.

[0115] The population flow adjustment value is calculated based on the population flow adjustment coefficient and the weight of the impact of population flow changes.

[0116] The impact weight of population flow changes is a data parameter obtained through historical data analysis and statistical calculation, which quantitatively describes the degree of impact of population flow on property prices within a property valuation area. The population flow adjustment coefficient is an optimized parameter used to adjust and calibrate the impact of population flow in the dynamic calibration model of property value. The calculation method for the population flow adjustment value is as follows: multiply the population flow adjustment coefficient by the impact weight of population flow changes at each time step to form a continuous sequence of population flow adjustment values. For example, in a property valuation area, when the impact weight of population flow changes shows that the intensity of population flow in the property valuation area is continuously increasing, and the population flow adjustment coefficient shows that population flow has a high positive impact on property prices, then by multiplying the impact weight of population flow changes by the population flow adjustment coefficient at each time step, a sequence of population flow adjustment values ​​reflecting the positive driving effect of population flow on property valuation within the area can be obtained.

[0117] The event adjustment value is calculated based on the event contribution adjustment coefficient and the event price contribution weight.

[0118] The event price contribution weight reflects the degree of impact of events such as policy releases, regional planning adjustments, and infrastructure construction within the property valuation area on the phased fluctuations of property prices. The event contribution adjustment coefficient is optimized and used to adjust the strength of the event's influence in the model calibration calculation. For example, when a newly planned subway line in an urban area officially opens, property prices rise. The event price contribution weight reflects the actual price-driving force of such events on the property market, while the event contribution adjustment coefficient reflects the influence weight of the event in the overall property value dynamic calibration model during model optimization. The event adjustment value is calculated by multiplying the event price contribution weight by the event contribution adjustment coefficient at each time step to obtain a sequence of event adjustment values ​​that reflect the influence of the event on property valuation.

[0119] The calibrated property valuation results are obtained by integrating the basic adjusted property valuation series, the population flow adjusted value, and the event adjusted value.

[0120] The adjusted property valuation series, the population flow adjustment series, and the event adjustment series are added together at their corresponding points in time to obtain a continuous, calibrated property valuation result series. This fusion calculation process comprehensively integrates basic market trends, the impact of population flow, and the impact of events to form a final property valuation result that reflects the true market conditions within the property valuation area. For example, in a specific area, if the adjusted property valuation shows that property market prices are stabilizing, the population flow adjustment shows rapid population growth, and the event adjustment reflects the price-driving effect of newly constructed urban transportation facilities, then by merging and calculating these three factors moment by moment, a calibrated property valuation result that fully reflects the actual property market prices within the area can be obtained.

[0121] Update the property valuation results for the property valuation area based on the calibrated property valuation results;

[0122] The calculated and calibrated property valuation results are used as real-time property market valuation results, and these are used to update and cover the property valuation information for the corresponding property valuation area in the property valuation system. For example, in the real-time property valuation system database or data platform, the latest calculated and calibrated property valuation results are recorded as the latest property market valuation within the property valuation area, and the updated property valuation results are pushed to relevant real estate transaction platforms or market participants in real time to guide the decision-making activities of the property transaction market and its participants. Through this real-time update, it is ensured that the property valuation within the property valuation area always accurately and effectively reflects the latest market conditions and price change trends, and promptly reflects the actual dynamic impact of population movement and event changes in the area on the real estate market.

[0123] Example 2

[0124] The difference between Embodiment 2 and Embodiment 1 is that this embodiment introduces a multi-source live data processing system for real estate valuation.

[0125] Figure 2 A schematic diagram of the multi-source liveness data processing system for real estate valuation according to the present invention is provided. The multi-source liveness data processing system for real estate valuation includes:

[0126] Data processing module: Acquires multi-source liveness data within the property valuation area, performs time alignment and spatial normalization on the multi-source liveness data, and generates a standard spatiotemporal dataset;

[0127] Population Trends Module: Performs short-period time-series analysis on standard spatiotemporal datasets to extract the changing trend of population flow intensity within the property assessment area and obtain population flow intensity indicators within the property assessment area;

[0128] Event Response Module: Performs event correlation analysis on standard spatiotemporal datasets and extracts quantitative response indicators for events within the property valuation area by setting event windows;

[0129] Weighted Attribution Module: Performs attribution analysis on the population flow intensity index and the quantitative response index of events within the property assessment area to determine the weight of the impact of population flow changes on short-term property prices and the contribution weight of events to the phased fluctuations of property prices.

[0130] Calibration Modeling Module: Establishes a dynamic calibration model for property value based on influence weights and contribution weights;

[0131] Valuation Adjustment Module: Based on the property value dynamic calibration model, the valuation of the property assessment area is calibrated and drift corrected, and the property valuation results of the property assessment area are adjusted in real time.

[0132] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0133] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0134] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0135] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0136] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0137] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0138] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0139] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of protection of the claims. Finally: The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for processing multi-source liveness data for real estate valuation, characterized in that, Includes the following steps: S1: Acquire multi-source liveness data within the property valuation area, and perform temporal alignment and spatial normalization on the multi-source liveness data to generate a standard spatiotemporal dataset, specifically: The first dataset is obtained by collecting mobile terminal signaling data, real-time population heat map data, real-time listing and transaction data, and remote sensing data on changes in building activities within the real estate assessment area. The first dataset is time-aligned using a unified timestamp, and a time-consistent dataset is formed by linear interpolation and filling in missing values. The geographic coordinates in the time-consistent dataset are converted into a unified geographic reference system, and the spatial resolution of each data source is rasterized and normalized to generate a spatially consistent dataset. Organize spatially consistent data sets according to time series order and spatial grid index, and output standard spatiotemporal datasets; S2: Perform short-period time-series analysis on the standard spatiotemporal dataset to extract the changing trend of population flow intensity within the property assessment area, and obtain the population flow intensity index within the property assessment area, specifically: Population heat map data and mobile terminal signaling data within the property assessment area are read from the standard spatiotemporal dataset, and short-period sequences are formed through resampling; By performing sliding window averaging and differencing on short-period sequences, a population flow measurement sequence is obtained. Applying Fast Fourier Transform to the population flow measurement sequence separates the daily and weekly periodic components, resulting in a periodic decomposition sequence. The first-order difference of the periodic decomposition sequence is calculated and linear regression is performed to extract the trend curve of population flow intensity. Based on the mean and coefficient of variation of the population flow intensity trend curve within a preset time window, a population flow intensity index is generated for the property assessment area. S3: Perform event correlation analysis on standard spatiotemporal datasets, and extract quantitative response indicators of events within the property valuation area by setting event windows; S4: Conduct attribution analysis on the population flow intensity index and the quantitative response index of events within the property valuation area to determine the weight of the impact of population flow changes on short-term property prices and the weight of the contribution of events to the phased fluctuations of property prices. S5: Establish a dynamic calibration model for real estate value based on the influence weight and contribution weight; S6: Based on the property value dynamic calibration model, the property valuation of the property assessment area is calibrated and drift corrected, and the property valuation results of the property assessment area are adjusted in real time.

2. The multi-source liveness data processing method for real estate valuation according to claim 1, characterized in that, S1, specifically: The first dataset is obtained by collecting mobile terminal signaling data, real-time population heat map data, real-time listing and transaction data, and remote sensing data on changes in building activities within the real estate assessment area. The first dataset is time-aligned using a unified timestamp, and a time-consistent dataset is formed by linear interpolation and filling in missing values. The geographic coordinates in the time-consistent dataset are converted into a unified geographic reference system, and the spatial resolution of each data source is rasterized and normalized to generate a spatially consistent dataset. Organize spatially consistent data sets according to time series order and spatial grid index, and output standard spatiotemporal datasets.

3. The multi-source liveness data processing method for real estate valuation according to claim 2, characterized in that, S2, specifically: Population heat map data and mobile terminal signaling data within the property assessment area are read from the standard spatiotemporal dataset, and short-period sequences are formed through resampling; By performing sliding window averaging and differencing on short-period sequences, a population flow measurement sequence is obtained. Applying Fast Fourier Transform to the population flow measurement sequence separates the daily and weekly periodic components, resulting in a periodic decomposition sequence. The first-order difference of the periodic decomposition sequence is calculated and linear regression is performed to extract the trend curve of population flow intensity. Based on the mean and coefficient of variation of the population flow intensity trend curve within a preset time window, a population flow intensity index for the property assessment area is generated.

4. The multi-source liveness data processing method for real estate valuation according to claim 3, characterized in that, S3, specifically: Establish event windows for policy release, regional planning adjustments, and supporting infrastructure construction, and determine the start time and duration of each event window; Retrieve real-time listing transaction data and remote sensing data of building activity changes that fall into each event window from the standard spatiotemporal dataset to generate a subset of event window data; A price comparison sequence is constructed for a subset of event window data in the order of event occurrence, and the interval average and the maximum deviation of the interval are calculated to form a vector of the impact of event prices; The vector of the price impact of an event and the duration of the corresponding event window are normalized to output a quantitative response index for the event within the property valuation area.

5. The multi-source liveness data processing method for real estate valuation according to claim 4, characterized in that, S4, specifically: Extract short-period real estate price series that are consistent with the time of population flow intensity indicators, and construct population flow-price pairing series; Based on the population flow-price pairing sequence, a mapping model between the population flow intensity index and short-term real estate prices is established using the multiple linear regression method, and the weight of the impact of population flow changes is calculated. Extract the real estate price phase sequence that is consistent with the time of the event quantitative response indicator, and construct the event impact-price pairing sequence; Based on the event impact-price pairing sequence, a mapping model between the quantitative response index of events and the phased fluctuations of real estate prices is established using the difference regression analysis method, and the contribution weight of event prices is calculated.

6. The multi-source liveness data processing method for real estate valuation according to claim 5, characterized in that, S5, specifically: Initialize the parameters of the property value dynamic calibration model, including the basic valuation deviation coefficient, the population flow adjustment coefficient, and the event contribution adjustment coefficient; Read the benchmark valuation sequence, real-time property price observations within the property valuation area, the weight of the impact of population flow changes, and the weight of the price contribution of events to construct a calibration training dataset; For the calibration training dataset, the least squares method is used to optimize the parameters of the property value dynamic calibration model; Save the optimized parameters of the property value dynamic calibration model and output the property value dynamic calibration model.

7. The multi-source liveness data processing method for real estate valuation according to claim 6, characterized in that, S6, specifically: The benchmark valuation series is initially adjusted based on the basic valuation deviation coefficient to obtain the basic adjusted real estate valuation series. The population flow adjustment value is calculated based on the population flow adjustment coefficient and the weight of the impact of population flow changes. The event adjustment value is calculated based on the event contribution adjustment coefficient and the event price contribution weight. The calibrated property valuation results are obtained by integrating the basic adjusted property valuation series, the population flow adjusted value, and the event adjusted value. Update the property valuation results for the property valuation area based on the calibrated property valuation results.

8. A multi-source liveness data processing system for real estate valuation, used to implement the multi-source liveness data processing method for real estate valuation as described in any one of claims 1-7, characterized in that, include: Data processing module: Acquires multi-source liveness data within the property valuation area, performs time alignment and spatial normalization on the multi-source liveness data, and generates a standard spatiotemporal dataset; Population Trends Module: Performs short-period time-series analysis on standard spatiotemporal datasets to extract the changing trend of population flow intensity within the property assessment area and obtain population flow intensity indicators within the property assessment area; Event Response Module: Performs event correlation analysis on standard spatiotemporal datasets and extracts quantitative response indicators for events within the property valuation area by setting event windows; Weighted Attribution Module: Performs attribution analysis on the population flow intensity index and the quantitative response index of events within the property assessment area to determine the weight of the impact of population flow changes on short-term property prices and the contribution weight of events to the phased fluctuations of property prices. Calibration Modeling Module: Establishes a dynamic calibration model for property value based on influence weights and contribution weights; Valuation Adjustment Module: Based on the property value dynamic calibration model, the valuation of the property assessment area is calibrated and drift corrected, and the property valuation results of the property assessment area are adjusted in real time.

Images