Method, storage medium and system for calibrating a new business district

Abstract

The present invention relates to the technical field of data processing, and particularly to a method, a storage medium, and a system for calibrating a new business district. The present invention provides a method for calibrating a new business district. By performing kernel density analysis on POI point data to identify high-density areas, an initial business district contour is formed based on the identified high-density areas to confirm the business district scope. Compared with the process of defining the existing business district contour that involves a large amount of manual intervention, the business district contour of the present invention is automatically realized, reducing the deficiencies of manually demarcating the business district. Moreover, the POI store technology is used in combination with passenger flow data to demarcate the new business district. The score of the POI points is combined with the passenger flow data of the business district shopping centers to form a business district score and grade the business district, reducing the deviation of calibrating the business district with a single POI data.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly relates to a method, a storage medium and a system for calibrating a new business district. Background Art

[0002] With the rapid development of information technology and the advent of the big data era, Point of Interest (POI) data of cities has gradually become an important tool for studying the characteristics of urban business districts and evaluating the vitality of business districts. POI data, as an important part of urban spatial information, not only records the locations and attributes of various facilities in the city, but also implies information in many aspects such as urban spatial structure, population distribution, and economic activities. By deeply analyzing POI data, the composition, characteristics and evolution laws of business districts can be grasped more accurately, so as to comprehensively evaluate the vitality, competitiveness and development potential of business districts. Although POI data has significant advantages in business district evaluation, it still faces some difficulties. For example, due to the complexity of data collection and processing, POI data may have errors and omissions, resulting in the analysis results deviating from the actual situation, especially in the calibration of new business districts. Some emerging business formats or small commercial facilities may not be included in the POI database in time, resulting in the underestimation of the diversity and vitality of business districts, and large deviations in the calibration of new business districts. Therefore, there is an urgent need for an optimized method for calibrating business districts using existing POI data. Summary of the Invention

[0003] The technical problem to be solved by the present invention is how to identify the core competitors in a business district.

[0004] To solve the above technical problem, the present invention provides a method for calibrating a new business district, including the following steps executed by a computer program:

[0005] S1. Call the multi-format POI point data of the target area, perform kernel density analysis on the obtained POI point data to identify high-density areas, and obtain the business district ranges of all the identified high-density areas respectively;

[0006] S2. Call the map data of the historical business district library and compare it with these business district ranges. If there is a business district range whose overlap degree with all the business districts in the historical business district library is lower than the preset degree, then use this business district range as a new target project;

[0007] S3. Obtain all the POI point data of the target project, identify the target stores from them, calculate the aggregation value of the identified target stores, and use the sum of the aggregation values of all the target stores as the POI total score;

[0008] S4. Obtain the passenger flow data within the business district range and the passenger flow data of multiple similar business districts, calculate the ratio of the passenger flow data within the current business district range to the passenger flow data of similar business districts, and multiply the ratio by 100 as the passenger flow total score;

[0009] S5. Respectively confirm the weights of the total POI score and the total passenger flow score, denoted as W_A and W_B, where W_A + W_B = 1. Calculate their weighted values according to the values of the total POI score and the total passenger flow score and the corresponding weights, and take the sum of all the weighted values as the business district score of the current target project. If the business district score reaches the preset threshold, then label this target project as a new business district.

[0010] Further, in the step S1, the specific operation of performing kernel density analysis on the obtained POI point data to identify high-density areas is as follows:

[0011] Identify the business types of the obtained POI point data, extract the coordinates of all POI points of the M business types with the largest number of POI points in the target area. In the unit area, calculate the regional density of each of these M business types respectively through the following kernel density analysis KDE formula

[0012] ;

[0013] f(x) represents the density estimate at position x, n is the number of POI points, h is the bandwidth parameter of kernel density analysis, K(·) is the Gaussian kernel function, and x i is the coordinate of the i-th POI point of this business type;

[0014] Calculate the sum of the regional densities of the M business types in the unit area. If the calculated sum of the regional densities reaches the preset saturation threshold, then consider the geographical area corresponding to this unit area as a high-density area. If multiple adjacent unit areas are all high-density areas, then combine these unit areas into one high-density area.

[0015] Further, in the step S1, the specific operation of obtaining the business district scope of each identified high-density area is as follows: Construct a heat map of the number of POI points for each high-density area to identify the business activity degree of each high-density area. Connect the areas with the business activity degree reaching the preset value at the outermost periphery of each high-density area to construct the largest outer contour, and take the area inside the largest outer contour as the business district scope of this high-density area.

[0016] Further, extract all POI points whose distance from the largest outer contour is within the preset range, calculate their regional influence according to the brand type and business value of each POI point, and if the regional influence reaches the preset degree, then include this POI point in the business district scope; the business value includes the brand popularity of the brand to which this POI point belongs in the target area and the popularity of the shopping center associated with this POI point.

[0017] Further, in the step S2, the degree of coincidence being lower than the preset degree means that: the degree of coordinate coincidence between the business district range and any business district in the historical business district library is lower than the coincidence threshold R1, or the sum of the degrees of coordinate coincidence between the business district range and multiple business districts in the historical business district library is lower than the coincidence threshold R2, and the coincidence threshold R2 is slightly greater than the coincidence threshold R1.

[0018] Further, in the step S3, the target stores include business district stores and street stores where the POI point heat reaches the preset degree; the POI point heat is the sum of the search volume statistical value, the social media interaction assignment, and the public opinion monitoring score.

[0019] Further, calculating the aggregation value of the identified target stores means: obtaining the brand indicators corresponding to the target stores, invoking the scores given by experts pre-stored for each brand indicator, and using the Analytic Hierarchy Process (AHP) to calculate the comprehensive score of the target stores on these brand indicators, which is denoted as the aggregation value;

[0020] The brand indicators include multiple of the following indicators: whether the POI brand is a national chain brand, whether the POI brand is a regional chain brand, whether the POI brand is a local chain brand, the total number of POI brand stores, the proportion of POI brand stores in the same city, the POI brand level, the social software heat value of the POI brand, and the weight of the business type to which the POI belongs.

[0021] Further, in the step S4, obtaining the passenger flow data within the business district range means: obtaining the population density of each hot spot area in the business district, and calculating the expected passenger flow data from each hot spot area in the business district in combination with the radiation probability of the business district to each hot spot area.

[0022] The present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps in the method for calibrating a new business district as described above are implemented.

[0023] The present invention also provides a business district calibration system, including a processor and the computer-readable storage medium as described above that are connected to each other.

[0024] The present invention has the following beneficial effects: In the method for calibrating a new business district, by performing kernel density analysis on POI point data to identify high-density areas, and forming an initial business district contour based on the identified high-density areas to confirm the business district range. Compared with the process of defining the existing business district contour that involves a large amount of manual intervention, the business district contour of the present invention is automatically realized, reducing the deficiencies of manually delineating the business district. Moreover, the POI store technology is used in combination with the passenger flow data to delineate a new business district, combining the score of the POI point with the passenger flow data of the business district shopping center to form a business district score and grade the business district, reducing the deviation of calibrating the business district with a single POI data. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is a block diagram of the calculation method for the business district score for calibrating a new business district. Specific implementation manner

[0026] The following further elaborates on the present invention in detail in conjunction with the specific implementation manner.

[0027] This embodiment provides a business district calibration system, which includes a processor and a computer-readable storage medium connected to each other. A computer program is stored on the computer-readable storage medium, and when the computer program is executed by the processor, it implements the method for calibrating a new business district as Figure 1 shown, and the method includes the following steps.

[0028] S1. Invoke the multi-format POI point data of the target area, perform kernel density analysis on the obtained POI point data to identify high-density areas, and obtain the business district ranges of all the identified high-density areas respectively.

[0029] The specific process of business district extraction is as follows: collect POI point data, import the data using GIS software (such as ArcGIS Pro), perform geocoding and spatial matching to ensure that each POI point is accurately corresponding to the geographical coordinates; subsequently, preprocess the data, including outlier detection, missing value filling, etc., to ensure the accuracy of the analysis; then, perform the statistics of POI points of each format in the urban grid - kernel density analysis - high-density area identification.

[0030] In step S1, the specific operation of performing kernel density analysis on the obtained POI point data to identify high-density areas is as follows:

[0031] Identify the format types of the obtained POI point data, extract the coordinates of all POI points of the M format types with the largest number of POI points in the target area, and within the unit area, calculate the regional density of each of these M format types respectively through the following kernel density analysis (Kernel Density Estimation) KDE formula,

[0032] ;

[0033] f(x) represents the density estimate at position x, n is the number of POI points, h is the bandwidth parameter of the kernel density analysis, K(·) is the Gaussian kernel function, and x i is the coordinate of the i-th POI point of this format type;

[0034] Calculate the sum of the regional densities of M business types within a unit area. If the calculated sum of the regional densities reaches a preset saturation threshold, the geographical area corresponding to this unit area is considered a high-density area. If multiple adjacent unit areas are all high-density areas, these unit areas are combined into one high-density area.

[0035] In this embodiment, the contributions of all kernel functions are superimposed through the kernel density analysis (KDE) formula to obtain an overall density estimate. This superimposing method enables each data point to contribute to the final density estimate, but the magnitude of the contribution depends on its distance from other data points and the shape of the kernel function. Therefore, KDE can generate a smoother and more accurate density curve.

[0036] In step S1, the specific method for obtaining the business circle range of each identified high-density area is as follows: construct a heat map of the number of POI points for each high-density area to identify the business activity degree of each high-density area, connect the areas with the business activity degree reaching the preset value at the outermost periphery of each high-density area to construct the largest outer contour, and use the area inside the largest outer contour as the business circle range of this high-density area. The automatic realization of the business circle contour in this embodiment reduces the deficiencies of manually demarcating the business circle.

[0037] In this embodiment, by collecting POI point data, it is possible to analyze the geographical distribution and density, thereby identifying business circle hotspots and heat maps to show the business activity degrees of different regions, and further identifying high-density areas. Moreover, by using the edge of the area with the activity degree reaching the preset range as the initial business circle contour, the business circle range can be accurately confirmed.

[0038] Furthermore, in this embodiment, all POI points within a preset range from the largest outer contour are extracted, and their regional influence is calculated according to the brand type and commercial value of each POI point. If the regional influence reaches the preset degree, the POI point is included in the business circle range; the commercial value includes the brand popularity of the brand to which the POI point belongs in the target area and the popularity of the shopping center associated with the POI point. Specifically, all POI points within the preset range from the edge of the obtained initial business circle contour are called, including but not limited to location coordinates, names, classification codes, etc. Brand attribute annotation: For each POI point, clarify the brand to which it belongs and mark whether the brand belongs to a well-known chain or a specific category (such as catering, retail, entertainment, etc.). At the same time, record whether the POI is located in a shopping center and all shopping center information in the target area. The popularity of the shopping center has been pre-recorded in the parameter table. Calculate the regional influence of each POI point according to its specific brand type and commercial value. If the regional influence reaches the preset degree, the POI point is included in the business circle range. The commercial value includes the brand popularity of the brand to which the POI point belongs in the current business circle location area and whether the POI point is in a shopping center.

[0039] S2. Call the map data of the historical business district library and compare it with these business district ranges. If there is a business district range whose coincidence degree with all business districts in the historical business district library is lower than the preset degree, then use this business district range as the new target project.

[0040] In step S2, the coincidence degree being lower than the preset degree means that: the coordinate coincidence degree of this business district range with any business district in the historical business district library is lower than the coincidence threshold R1, or the sum of the coordinate coincidence degrees of this business district range with multiple business districts in the historical business district library is lower than the coincidence threshold R2, and the coincidence threshold R2 is slightly larger than the coincidence threshold R1.

[0041] To solve the problem of low discrimination caused by similar brand densities in adjacent areas, this embodiment also performs refined grid, multi-scale analysis, and introduction of auxiliary variable processing on adjacent areas. Refined grid: Divide the research area into smaller grid units to improve spatial resolution. Multi-scale analysis: Draw heat maps at different scales to observe the spatial distribution characteristics of brand density. Introduction of auxiliary variables: Combine auxiliary variables such as population density, consumption level, and traffic conditions to comprehensively analyze brand density and improve the discrimination of adjacent areas.

[0042] S3. Obtain all POI point data of the target project, identify the target stores from them, calculate the aggregation value of the identified target stores, and use the sum of the aggregation values of all target stores as the POI total score.

[0043] In step S3, the target stores include business district stores and street stores where the POI point heat reaches the preset degree; the POI point heat is the sum of the search volume statistical value, social media interaction assignment, and public opinion monitoring score.

[0044] Among them, the search volume statistical value: Through the keyword planner tool provided by the search engine, count the number of searches for the brand term within a period of time (such as the past 30 days), and use the ratio of this number of searches to the average number of searches for the POI point as the search volume statistical value. Social media interaction: Monitor the discussion heat of the brand on major social platforms, including the number of likes, comments, reposts, etc., and use the ratio of this discussion heat to the average discussion heat of the POI point as the social media interaction assignment. Public opinion monitoring: Use a special brand heat index query tool, such as Sogou Brand Monitoring, to regularly track the mention frequency and sentiment tendency (positive and negative) of the brand on the Internet, use the ratio of the mention frequency to the average mention frequency of the POI point as the basic value of the public opinion monitoring score, and use the result of multiplying this basic value by the sentiment tendency weight as the public opinion monitoring score. The sentiment tendency weight is specifically: 110% for positive and 95% for negative.

[0045] Among them, the POI point heat reaching the preset degree means that the POI point heat is higher than the median of the POI point heat in this target area.

[0046] In step S3, calculating the aggregation value of the identified target store means: obtaining the brand metrics corresponding to the target store, invoking the scores of each brand metric pre-stored by experts, and using the Analytic Hierarchy Process (AHP) to calculate the comprehensive score of the target store on these brand metrics, denoted as the aggregation value.

[0047] The POI data in this embodiment is the effective brand POI matched with the standard brand library, that is, the brand dimension information after standard construction can be obtained, so that the scores of experts on each brand metric can be obtained, and the weights of brand metrics can be obtained through the Analytic Hierarchy Process (AHP), thereby calculating the comprehensive score of the brand. Specifically, the implementation process of calculating the brand score based on the Analytic Hierarchy Process (AHP) in this embodiment includes the following steps 1 to 3.

[0048] Step 1: Confirm brand metrics

[0049] The brand score in this embodiment is used to measure the influence and competitiveness of the brand in the market. The confirmed brand metrics include the following eight types: whether the POI brand is a national chain brand, whether the POI brand is a regional chain brand, whether the POI brand is a local chain brand, the total number of POI brand stores, the proportion of POI brand stores in the same city, the POI brand level, the popularity value of the POI brand on social software, and the weight of the business type to which the POI belongs.

[0050] Whether the POI brand is a national chain brand: This feature is of boolean type, indicating whether the POI brand has chain stores nationwide.

[0051] Whether the POI brand is a regional chain brand: This feature is also of boolean type, indicating whether the POI brand has chain stores in a specific region (such as a province or a city).

[0052] Whether the POI brand is a local chain brand: A boolean type feature, indicating whether the POI brand has chain stores only in a certain place (such as a county or district).

[0053] The total number of POI brand stores: A numerical type feature, representing the total number of stores the brand has nationwide.

[0054] The proportion of POI brand stores in the same city: A numerical type feature, representing the proportion of the number of stores of the brand in the same city to the total number of stores.

[0055] POI brand level: A numerical type or categorical type feature, indicating the level or grade of the brand.

[0056] The popularity value of the POI brand on social software: A numerical type feature, representing the attention and interaction popularity of the brand on social software.

[0057] Weight of the business type to which the POI belongs: A numerical feature indicating the importance or market share of the business type to which the POI belongs in the entire market.

[0058] Step 2: Construct the judgment matrix

[0059] Compare these eight brand indicators pairwise through expert scoring to determine their relative importance. Each comparison should follow a consistent rating scale from 1 to 9, where 1 indicates that the two elements are equally important, and 9 indicates that the former is extremely more important than the latter.

[0060] The judgment matrix of the eight brand indicators is as follows:

[0061]

[0062] Step 3: Calculate the weights and perform consistency check

[0063] Based on the above judgment matrix, solve its maximum eigenvalue and its corresponding eigenvector. Then normalize the eigenvector to obtain the weight values of each brand indicator. Also, perform a consistency check to ensure that the consistency of the judgment matrix meets the requirements, and the consistency ratio CR is less than 0.1. The above processing process is the general process of the Analytic Hierarchy Process (AHP) and will not be elaborated here.

[0064] After calculation, the following weight distribution is obtained:

[0065] National chain brand: 0.30

[0066] Regional chain brand: 0.20

[0067] Local chain brand: 0.10

[0068] Total number of stores: 0.15

[0069] Proportion in the same city: 0.05

[0070] Brand level: 0.05

[0071] Popularity value on social software: 0.10

[0072] Weight of the business type to which it belongs: 0.05.

[0073] Step 4: Calculate the brand score

[0074] According to the scores of each brand indicator evaluated by experts, calculate the comprehensive score of the brand by combining the above weights: the sum of the scores of each indicator multiplied by its corresponding weight. This is used as the aggregation value of the target store.

[0075] S4. Obtain the passenger flow data within the business district scope and the passenger flow data of multiple similar business districts, calculate the ratio of the current passenger flow data within the business district scope to the passenger flow data of similar business districts, and multiply the ratio by 100 as the total passenger flow score. The total passenger flow score can also be obtained by analyzing the passenger flow value of the project using the entropy weight method.

[0076] In step S4, obtaining the passenger flow data within the business district scope means: obtaining the population density of each hot spot area in the business district, and calculating the expected passenger flow data from each hot spot area in the business district by combining the radiation probability of the business district to each hot spot area.

[0077] Take the square root of the ratio of the number of connected paths between hot spot areas within the current business district scope to the average number of connected paths in this target area as the passenger flow weight. The radiation probability is the average passenger flow in this target area / the average population density multiplied by the passenger flow weight. Obtain the population density of each hot spot area in the business district, which includes the residential population density and the working population density, and then multiply the population density of a certain hot spot area in the business district by the radiation probability of the business district to this hot spot area to calculate the expected passenger flow data from this hot spot area.

[0078] Generate the theoretical heat map of the business district according to the expected passenger flow data of each hot spot area in the business district. Specifically, mark different depths of colors for each hot spot area in the business district according to the expected passenger flow data, so as to generate the theoretical heat map of the business district. The larger the expected passenger flow data of a certain hot spot area in the business district, the darker the color marked for this hot spot area.

[0079] S5. Respectively confirm the weights of the total POI score and the total passenger flow score, denoted as W_A and W_B, where W_A + W_B = 1. Calculate their weighted values according to the values of the total POI score and the total passenger flow score and the corresponding weights, and take the sum of all weighted values as the business district score of the current target project. If the business district score reaches the preset threshold, then label this target project as a new business district. In this embodiment, the business district levels corresponding to the business district scores are pre-stored: municipal / regional / community level. Find the corresponding business district level according to the business district score and output it. In this embodiment, by combining the scoring of POI points with the passenger flow data of shopping centers in the business district to form a business district score and classify the business district, the deviation of calibrating the business district with a single POI data is reduced.

[0080] The method for calibrating a new business district in this embodiment processes the multi-format POI point data by using kernel density analysis, automatically identifies high-density areas, and determines the business district scope. This process reduces the subjectivity and inconsistency that may occur in the traditional manual definition of business districts. In addition, by combining the calculation of the aggregation value of target stores and the analysis of passenger flow data for comprehensive scoring, it not only reflects the concentration of business activities through the total POI score, but also reflects the actual popularity of the flow of people through the total passenger flow score, thus realizing a comprehensive and objective evaluation of the business district. Therefore, compared with the traditional method that only relies on single-dimensional data, this composite evaluation strategy can more accurately reflect the true value and development potential of the business district, effectively reduce the risk of misjudgment, and provide solid data support for business decisions.

[0081] As described above, it is only the implementation manner of the invention-creation, and does not limit the scope of patent protection. Those skilled in the art make non-substantive changes or substitutions on the basis of the invention-creation, and still fall within the scope of patent protection.

Claims

1. A method for calibrating a new business district, characterized in that, Including the following steps executed by a computer program: S1. Invoke the multi-format POI point data of the target area, perform kernel density analysis on the obtained POI point data to identify high-density areas, and obtain the business district scope of each identified high-density area; S2. Invoke the map data of the historical business district library and compare it with these business district scopes. If there is a business district scope with a coincidence degree lower than the preset degree with all business districts in the historical business district library, then use this business district scope as a new target project; S3. Obtain all the POI point data of the target project, identify the target stores from it, calculate the aggregation value of the identified target stores, and use the sum of the aggregation values of all target stores as the POI total score; The calculation of the aggregation value of the identified target stores refers to: obtaining the brand indicators corresponding to the target stores, invoking the scores given by experts pre-stored for each brand indicator, and using the Analytic Hierarchy Process (AHP) to calculate the comprehensive score of the target stores on these brand indicators, denoted as the aggregation value; the brand indicators include multiple of the following indicators: whether the POI brand is a national chain brand, whether the POI brand is a regional chain brand, whether the POI brand is a local chain brand, the total number of POI brand stores, the proportion of POI brand stores in the same city, the POI brand level, the popularity value of the POI brand on social software, the weight of the POI business format; S4. Obtain the passenger flow data within the business district scope and the passenger flow data of multiple similar business districts, calculate the ratio of the current passenger flow data within the business district scope to the passenger flow data of similar business districts, and use the ratio multiplied by 100 as the passenger flow total score; S5. Confirm the weights of the POI total score and the passenger flow total score respectively, denoted as W_A and W_B, where W_A + W_B = 1. Calculate their weighted values according to the values of the POI total score and the passenger flow total score and the corresponding weights, and use the sum of all weighted values as the business district score of the current target project. If the business district score reaches the preset threshold, then label this target project as a new business district.

2. The method for calibrating a new business district according to claim 1, characterized in that, In step S1, the specific process of performing kernel density analysis on the obtained POI point data to identify high-density areas is as follows: Identify the business format types of the obtained POI point data, extract the coordinates of all POI points of the M business format types with the largest number of POI points in the target area, and within the unit area, calculate the regional density of each of these M business format types respectively through the following kernel density analysis (KDE) formula ； f(x) represents the density estimate at position x, n is the number of POI points, h is the bandwidth parameter of kernel density analysis, K(·) is the Gaussian kernel function, and x i is the coordinate of the i-th POI point of this business type; Calculate the sum of the regional densities of the M business format types within the unit area. If the calculated sum of the regional densities reaches the preset saturation threshold, then consider the geographical area corresponding to this unit area as a high-density area. If multiple adjacent unit areas are all high-density areas, then combine these unit areas into one high-density area.

3. The method for calibrating a new business district according to claim 1, wherein In the step S1, the specific method for obtaining the business circle range of each identified high-density area is as follows: construct a heat map of the number of POI points for each high-density area to identify the business activity degree of each high-density area, connect the areas with the business activity degree reaching the preset value at the outermost periphery of each high-density area to construct the maximum outer contour, and use the area inside the maximum outer contour as the business circle range of the high-density area.

4. The method for calibrating a new business district according to claim 3, characterized in that, Extract all POI points within a preset distance from the maximum outer contour, calculate their regional influence according to the brand type and business value of each POI point, and include the POI point in the business circle range if the regional influence reaches the preset degree; the business value includes the brand popularity of the brand to which the POI point belongs in the target area and the popularity of the shopping center associated with the POI point.

5. The method for calibrating a new business district according to claim 1, characterized in that, In the step S3, the target stores include business circle stores and street stores with the POI point heat reaching the preset degree; the POI point heat is the sum of the search volume statistical value, social media interaction assignment, and public opinion monitoring score.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps in the method for calibrating a new business circle as described in any one of claims 1 to 5.

7. Business district calibration system, characterized by It includes a processor and a computer-readable storage medium as described in claim 6 that are connected to each other.

Images