A method for calculating the inter-provincial market segmentation index based on traffic flow big data

Abstract

The present invention discloses a method for calculating a cross-provincial market segmentation index based on traffic big data. This method breaks through the limitation of traditional reliance on price data or sales data of a few categories of goods to estimate the market segmentation index, and instead uses more general and representative commercial vehicle traffic big data for evaluation. Traffic data can more intuitively and comprehensively reflect the actual economic connection between regions. Therefore, compared with traditional methods, this method can more realistically show the actual situation of cross-provincial market segmentation. In addition, the present invention constructs a gravity model based on traffic big data, and calculates the cross-regional market segmentation index in combination with multiple factors such as geography and culture, providing a new cross-regional market segmentation evaluation method. This method can not only quantify the degree of market segmentation between different regions, but also provide scientific data support for formulating regional integration policies and monitoring the progress of unified market construction.

Description

Technical Field

[0001] The present invention relates to the field of economic indicator research using Internet of Vehicles big data, spatial positioning technology and econometric estimation methods, and specifically to a method for calculating a cross-provincial market segmentation index based on vehicle flow big data. Background Art

[0002] In regional economic development, the problem of cross-regional market segmentation is an important factor restricting the construction of a unified market. Cross-regional market segmentation refers to the isolation or segmentation of markets between different regions due to administrative boundaries, geographical barriers, cultural differences and other factors. This phenomenon restricts the free flow of resources, reduces market efficiency, hinders the development of regional economic integration, and has a negative impact on the overall economic synergy. In order to achieve the strategic goal of a unified market, it is particularly important to scientifically evaluate and dynamically monitor the degree of cross-regional market segmentation.

[0003] Traditionally, the measurement of market segmentation relies on price indices, sales data or trade data of a few commodities, but these methods have limitations. First, commodity prices and sales data cannot fully reflect the economic connections between regions and are only applicable to specific areas. Second, the collection of these data is difficult and has a strong time lag, making it difficult to meet the needs of real-time and comprehensiveness. Especially in the context of increasingly complex regional economic connections, traditional methods can no longer accurately reflect the true situation of cross-regional market segmentation.

[0004] With the rapid development of transportation and information technology, big data technology and Internet of Vehicles data provide new tools for the analysis of regional economic connections. Internet of Vehicles data, especially commercial vehicle traffic data, can comprehensively and real-time reflect the economic connections between regions. Road transportation is one of the most important channels for regional economic connections. Therefore, analyzing regional market segmentation based on commercial vehicle traffic big data can overcome the limitations of traditional methods and monitor changes in regional economic connections more comprehensively and dynamically.

[0005] In view of this, the present invention is proposed. Summary of the invention

[0006] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide a method for calculating the cross-provincial market segmentation index based on traffic big data, thereby solving the problems raised in the above-mentioned background technology.

[0007] In order to solve the above technical problems, the basic concept of the technical solution adopted by the present invention is:

[0008] A method for calculating the inter-provincial market segmentation index based on traffic flow big data includes the following steps:

[0009] S1, calculate the annual inter-city traffic flow matrix based on commercial vehicle flow, by obtaining the Internet of Vehicles data reported by commercial vehicles, filtering the vehicle travel data by year according to the reporting time, and performing feature data analysis, including identifying and locating the start and end points of the trip, extracting key event features, and summing them up according to the attribution of the start and end points to form the annual traffic flow matrix of the city pairs;

[0010] S2, constructing inter-provincial variables based on map data, judging whether any two cities are located in the same province through administrative division maps, and generating inter-provincial variables between cities for subsequent gravity model calculations;

[0011] S3, constructs geographic segmentation proxy variables based on the map API, and quantifies geographic segmentation by calculating the ratio of the straight-line distance between cities to the actual driving distance;

[0012] S4, constructing a proxy variable for cultural differences based on dialect data, and calculating the dialect distance between cities to reflect the impact of cultural differences on market segmentation;

[0013] S5, using external data to construct other variables that affect inter-provincial traffic flow, including the geographical distance between the starting city and the end city, GDP, population, and characteristic data of whether it is a resource-based city, to control the impact of other factors on market segmentation;

[0014] S6. Use the gravity model to calculate the inter-provincial market segmentation index, and combine it with the ordinary least squares econometric method to evaluate whether the traffic volume between cities across provinces is significantly lower than the traffic volume between cities in the same province, so as to determine the extent of the inter-provincial market segmentation phenomenon.

[0015] Optionally, in step one, when calculating the annual inter-city traffic flow matrix based on commercial vehicle flow, first, based on the obtained inter-regional commercial vehicle flow data, using the starting and ending administrative division information and start time information of each economic trip, all truck trips are summed up into a traffic flow information matrix of "starting city, ending city, year"; then, according to the year, the "starting city, ending city traffic flow" matrix of the corresponding time period is extracted to form panel data for model analysis.

[0016] Optionally, when constructing inter-provincial variables based on map data in step 2, first, use the geographic information system (GIS) to identify the administrative division information of any two cities, and then compare the provincial administrative units to which each pair of cities belongs. If the two cities do not belong to the same province, they are marked as inter-provincial; if they belong to the same province, they are marked as the same province. Then, based on the judgment result, construct inter-provincial variables between cities. If the two cities are inter-provincial, the inter-provincial variable is recorded as 1; if the two cities are in the same province, the inter-provincial variable is recorded as 0. Finally, the generated inter-provincial variables are input into the gravity model to evaluate the impact of inter-provincial factors on inter-city traffic flow, thereby analyzing the degree of inter-provincial market segmentation.

[0017] Optionally, in step 3, when constructing a geographic segmentation proxy variable based on the map API, first, define the center point of each city as the latitude and longitude coordinates of the city government seat as the starting point and end point for calculating the geographic distance and driving distance. Then, use the route planning function of the map API platform to calculate the shortest driving distance between any two city centers during non-peak hours. Subsequently, use the geodist command of the Stata software to calculate the straight-line distance on the earth's surface between the above city centers to reflect the shortest physical distance between the cities. Finally, for each pair of cities, calculate the ratio of their actual driving distance to the straight-line geographic distance and define it as a "geographic segmentation variable". The higher the ratio, the more significant the barrier between the two cities caused by natural geographical conditions, which is used to quantify the degree of geographic segmentation.

[0018] Optionally, the steps for determining the extent of cross-provincial market segmentation are:

[0019] Take the "traffic flow of starting city and ending city" generated in step 1 as the main table, and merge the city-pair level variables generated in steps 2 to 4 into the main table with "starting city" and "ending city" as keywords; match the city-level variables of the starting city and ending city generated in step 5 to the main table respectively;

[0020] Through the formula: ln(traffic flow i，j )=α 0 +α 1 ×Interprovince i，j +α 2 ×Dialect Distance i，j +α 3 × Geographical Segmentation Index i，j +α 4 ×ln(distance i,j )+δ×M+∈ i，j The econometric model is set, where ln(traffic volume) is the logarithm of the traffic volume between cities, interprovincial is the indicator variable of whether the starting city and the end city are across provinces, dialect distance is the cultural difference between cities, geographic segmentation index is the degree of geographical isolation, ln(distance) is the logarithm of the actual geographical distance between cities, and M is the control variable, including GDP, population and other characteristics;

[0021] Using the reghdfe command in Stata software, the ordinary least squares method was used to estimate the parameters in the model, focusing on the analysis of the cross-provincial variable α 1 The estimated value and its statistical significance, if α 1 The sign of is negative and statistically significant, which means that after controlling other variables, the traffic volume between cities across provinces is significantly lower than the traffic volume between cities within the same province, that is, there is a phenomenon of cross-provincial market segmentation; α 1 The larger the absolute value of , the higher the degree of cross-provincial market segmentation.

[0022] After adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art. Of course, any product implementing the present invention does not necessarily need to achieve all the advantages described below at the same time:

[0023] 1. This method breaks through the limitation of estimating the market segmentation index by relying on price data or sales data of a few categories of goods in the traditional way, and instead uses more general and representative commercial vehicle traffic big data for evaluation. Traffic flow data can more intuitively and comprehensively reflect the actual economic connection between regions. Therefore, compared with traditional methods, this method can more realistically show the actual situation of cross-provincial market segmentation.

[0024] 2. This invention combines big data and scientific analysis methods, not only to describe the current economic ties between cities, but also to further refine and summarize the social science laws and policy implications related to regional economic development. By quantitatively analyzing the cross-provincial market segmentation, this method provides a scientific basis for understanding and improving the economic integration between cities and regions.

[0025] 3. Based on the traffic flow big data, this paper constructs a gravity model, combines geography, culture and other factors to calculate the cross-regional market segmentation index, and provides a new cross-regional market segmentation evaluation method. This method can not only quantify the degree of market segmentation between different regions, but also provide scientific data support for formulating regional integration policies and monitoring the progress of unified market construction.

[0026] The specific implementation modes of the present invention are further described in detail below in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The drawings described below are only some embodiments. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0028] In the figure:

[0029] Figure 1 The figure is a flow chart of the market segmentation index calculation method.

[0030] It should be noted that these drawings and textual descriptions are not intended to limit the conceptual scope of the present invention in any way, but are intended to illustrate the concept of the present invention for those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0031] The present invention will now be described in further detail with reference to the accompanying drawings.

[0032] See also Figure 1 As shown, in this embodiment, a method for calculating the inter-provincial market segmentation index based on traffic big data is provided, comprising the following steps:

[0033] S1, calculates the annual inter-city traffic flow matrix based on commercial vehicle flow, obtains the Internet of Vehicles data reported by commercial vehicles, filters the vehicle travel data by year according to the reporting time, and performs feature data analysis, including identifying and locating the start and end points of the travel, extracting key event features, and summing them up according to the attribution of the start and end points to form an annual traffic flow matrix for cities; Internet of Vehicles data includes the time when the Internet of Vehicles data is reported, in order to obtain the vehicle's travel data and satellite positioning data for one year, and analyze the obtained satellite positioning data, mainly analyzing the satellite positioning time and satellite positioning longitude and latitude of the data points.

[0034] The Internet of Vehicles data also includes license plate number, trip number, start time, starting longitude and latitude, and end longitude and latitude.

[0035] Feature data analysis refers to the extraction of features that can represent the driving behavior of a commercial vehicle from at least one year of continuous satellite positioning data. The behavioral event features analyzed include: trip mark, economic trip starting point, economic trip end point and the characteristics of the located starting and ending cities.

[0036] When performing data screening, based on satellite positioning data, trips with "distance <= 500 meters" and other unnecessary indicators are deleted, and the starting and ending points of all economic trips are defined.

[0037] S2, constructing inter-provincial variables based on map data, judging whether any two cities are located in the same province through administrative division maps, and generating inter-provincial variables between cities for subsequent gravity model calculations;

[0038] Construct inter-provincial variables based on map data. Use administrative division maps to determine whether any two cities are located in the same province, construct inter-provincial variables between cities, and use them for subsequent gravity model calculations to assess whether the traffic between inter-provincial city pairs is significantly lower than that between city pairs in the same province after controlling other variables, so as to determine whether there is a phenomenon of inter-provincial market segmentation and its extent.

[0039] The characteristic data obtained are constructed by comparing maps, manually marking whether any two cities are located in the same province, and constructing cross-provincial variables;

[0040] S3, constructs geographic segmentation proxy variables based on the map API, quantifies geographic segmentation by calculating the ratio of the straight-line distance between cities to the actual driving distance; constructs geographic segmentation proxy variables based on the map API. It is used to calculate the cross-provincial market segmentation index using the gravity model, excluding the market segmentation caused by geographical factors, and increasing the accuracy of the index.

[0041] The behavioral event characteristics analyzed are the ratio of the actual distance increased by geographical barriers to the straight-line distance between any two cities. The characteristic data obtained are: the straight-line geographical distance between the city centers represented by the municipal governments of any two cities; and the shortest driving distance between the city centers represented by the municipal governments of two cities during non-congested hours.

[0042] S4, constructs a proxy variable for cultural differences based on dialect data, and reflects the impact of cultural differences on market segmentation by calculating the dialect distance between cities; constructs a proxy variable for cultural differences based on dialect data, and the external data obtained are: the dialect distance between any two cities; the original data for constructing the dialect distance are: the main dialect types at the district and county level, the value assigned to the dialect difference, and the population of any two districts and counties.

[0043] Drawing on the method of Xu Xianxiang's team from Sun Yat-sen University (Liu Yuyun, Xu Xianxiang, Xiao Zekai. Inverted U-shaped pattern of labor force mobility across dialects [J]. Economic Research, 2015, 50(10): 134-146), based on the "1986 Dialect Affiliation Data of Chinese Counties", the dialect type spoken in the most districts and counties in a prefecture-level city is defined as the dialect type of the city. The dialect distance between prefecture-level cities is then calculated by weighting the population size of each county in the 2000 census to represent the cultural differences between regions.

[0044] S5. Use external data to construct other variables that affect inter-provincial traffic flow, including the geographical distance, GDP, population, and whether the starting city and the end city are resource-based city characteristics data, which are used to control the impact of other factors on market segmentation; the characteristic data obtained are: the straight-line distance between the starting city and the end city; the external data obtained are: GDP, population, whether the starting city and the end city are resource-based city and other characteristic variable data.

[0045] Drawing on the method of Xu Xianxiang's team from Sun Yat-sen University (Liu Yuyun, Xu Xianxiang, Xiao Zekai. Inverted U-shaped pattern of labor force mobility across dialects [J]. Economic Research, 2015, 50(10): 134-146), based on the "1986 Dialect Affiliation Data of Chinese Counties", the dialect type spoken in the most districts and counties in a prefecture-level city is defined as the dialect type of the city. The dialect distance between prefecture-level cities is then calculated by weighting the population size of each county in the 2000 census to represent the cultural differences between regions.

[0046] S6. Use the gravity model to calculate the inter-provincial market segmentation index, and combine it with the ordinary least squares econometric method to evaluate whether the traffic volume between cities across provinces is significantly lower than the traffic volume between cities in the same province, so as to determine the extent of the inter-provincial market segmentation phenomenon.

[0047] In the present embodiment, in step 1, when calculating the annual inter-city traffic flow matrix based on commercial vehicle flow, first, based on the obtained inter-regional commercial vehicle flow data, using the starting and ending administrative division information and starting time information of each economic journey, all truck trips are summed up into a traffic flow information matrix of "starting city, ending city, year"; then, according to the year, the "starting city, ending city traffic flow" matrix of the corresponding time period is extracted to form panel data for model analysis.

[0048] In this embodiment, when constructing cross-provincial variables based on map data in step 2, first, use the geographic information system (GIS) to identify the administrative division information of any two cities, and then compare the provincial administrative units to which each pair of cities belongs. If the two cities do not belong to the same province, they are marked as cross-provincial, and if they belong to the same province, they are marked as the same province. Then, based on the judgment result, construct cross-provincial variables between cities. If the two cities are across provinces, the cross-provincial variable is recorded as 1; if the two cities are in the same province, the cross-provincial variable is recorded as 0. Finally, the generated cross-provincial variable is input into the gravity model to evaluate the impact of cross-provincial factors on inter-city traffic flow, thereby analyzing the degree of cross-provincial market segmentation.

[0049] In this embodiment, when constructing the geographic segmentation proxy variable based on the map API, step 3 first defines the center point of each city as the latitude and longitude coordinates of the city government seat, as the starting point and end point for calculating the geographic distance and driving distance, then uses the route planning function of the map API platform to calculate the shortest driving distance between any two city centers during non-peak hours, then uses the geodist command of the Stata software to calculate the straight-line distance on the earth's surface between the above city centers, reflecting the shortest physical distance between the cities, and finally, for each pair of cities, calculates the ratio of its actual driving distance to the straight-line geographic distance, and defines it as a "geographic segmentation variable". The higher the ratio, the more significant the barrier between the two cities caused by natural geographical conditions, which is used to quantify the degree of geographic segmentation.

[0050] In this embodiment, the steps for determining the extent of the cross-provincial market segmentation phenomenon are as follows:

[0051] Take the "traffic flow of starting city and ending city" generated in step 1 as the main table, and merge the city-pair level variables generated in steps 2 to 4 into the main table with "starting city" and "ending city" as keywords; match the city-level variables of the starting city and ending city generated in step 5 to the main table respectively;

[0052] Through the formula: ln(traffic flow i,j )=α 0 +α 1 ×Interprovince i，j +α 2 ×Dialect Distance i，j +α3 × Geographical Segmentation Index i，j +α 4 ×ln(distance i，j )+δ×M+∈ i，j The econometric model is set, where ln(traffic volume) is the logarithm of the traffic volume between cities, interprovincial is the indicator variable of whether the starting city and the end city are across provinces, dialect distance is the cultural difference between cities, geographic segmentation index is the degree of geographical isolation, ln(distance) is the logarithm of the actual geographical distance between cities, and M is the control variable, including GDP, population and other characteristics;

[0053] Using the reghdfe command in Stata software, the ordinary least squares method was used to estimate the parameters in the model, focusing on the analysis of the cross-provincial variable α 1 The estimated value and its statistical significance, if α 1 The sign of is negative and statistically significant, which means that after controlling other variables, the traffic volume between cities across provinces is significantly lower than the traffic volume between cities within the same province, that is, there is a phenomenon of cross-provincial market segmentation; α 1 The larger the absolute value of , the higher the degree of cross-provincial market segmentation.

[0054] Additional notes:

[0055] (1) If you need to calculate the cross-provincial market segmentation index in a specific region, such as the eastern, central, or western regions, you can select samples from the total table whose starting and ending points are located in the selected region, and repeat the measurement model with reference to the above steps. At this time, the estimated "cross-province" coefficient α in the regression result of the model is 1 It can represent the market segmentation index within the region.

[0056] (2) If you need to calculate the market segmentation index of a specific region as a whole, you can filter out samples whose starting or ending points are in the region from the total table, and repeat the measurement model with reference to the above steps. Here, "whether it is across regions (cross-region = 1 means that only one of the starting and ending cities is in the region, and cross-region = 0 means that both the starting and ending cities are in the region)" replaces the "whether it is in the same province" variable in the original model. At this time, the "whether it is across regions" coefficient α 1 It can be used as the overall market segmentation index of the region, indicating the difference between the economic connections between cities across the region and the average economic connections between cities within the region when other conditions are similar.

[0057] (3) The above-mentioned specific areas can also be provincial administrative regions. According to the steps in (2), the overall market segmentation index of any province (except municipalities) can be estimated for intuitive quantification and comparative analysis.

[0058] The present invention is not limited to the above-mentioned embodiments. Anyone should be aware that any structural changes made under the enlightenment of the present invention, and any technical solutions that are the same or similar to the present invention, fall within the protection scope of the present invention. The technology, shape, and structural parts not described in detail in the present invention are all well-known technologies.

Claims

1. A method for calculating the inter-provincial market segmentation index based on traffic big data, characterized in that: The following steps are involved: S1, calculate the annual inter-city traffic flow matrix based on commercial vehicle flow, by obtaining the Internet of Vehicles data reported by commercial vehicles, filtering the vehicle travel data by year according to the reporting time, and performing feature data analysis, including identifying and locating the start and end points of the trip, extracting key event features, and summing them up according to the attribution of the start and end points to form the annual traffic flow matrix of the city pairs; S2, constructing inter-provincial variables based on map data, judging whether any two cities are located in the same province through administrative division maps, and generating inter-provincial variables between cities for subsequent gravity model calculations; S3, constructs geographic segmentation proxy variables based on the map API, and quantifies geographic segmentation by calculating the ratio of the straight-line distance between cities to the actual driving distance; S4, constructing a proxy variable for cultural differences based on dialect data, and calculating the dialect distance between cities to reflect the impact of cultural differences on market segmentation; S5, using external data to construct other variables that affect inter-provincial traffic flow, including the geographical distance between the starting city and the end city, GDP, population, and characteristic data of whether it is a resource-based city, to control the impact of other factors on market segmentation; S6. Use the gravity model to calculate the inter-provincial market segmentation index, and combine it with the ordinary least squares econometric method to evaluate whether the traffic volume between cities across provinces is significantly lower than the traffic volume between cities in the same province, so as to determine the extent of the inter-provincial market segmentation phenomenon.

2. According to claim 1, a method for calculating the inter-provincial market segmentation index based on traffic flow big data is characterized in that: Step 1: When calculating the annual inter-city traffic flow matrix based on commercial vehicle flow, first, based on the obtained inter-regional commercial vehicle flow data, use the starting and ending administrative division information and start time information of each economic trip to add up all truck trips into the traffic flow information matrix of "starting city, ending city, year"; then, according to the year, extract the "starting city, ending city traffic flow" matrix of the corresponding time period to form panel data for model analysis.

3. The method for calculating the inter-provincial market segmentation index based on traffic flow big data according to claim 1 is characterized in that: When constructing inter-provincial variables based on map data in step 2, first, use the geographic information system (GIS) to identify the administrative division information of any two cities. Then, for each pair of cities, compare the provincial administrative units to which they belong. If the two cities do not belong to the same province, they are marked as inter-provincial. If they belong to the same province, they are marked as same-province. Then, based on the judgment results, construct inter-provincial variables between cities. If the two cities are inter-provincial, the inter-provincial variable is recorded as 1; if the two cities are in the same province, the inter-provincial variable is recorded as 0. Finally, the generated inter-provincial variables are input into the gravity model to evaluate the impact of inter-provincial factors on inter-city traffic flow, thereby analyzing the degree of inter-provincial market segmentation.

4. The method for calculating the inter-provincial market segmentation index based on traffic flow big data according to claim 1 is characterized in that: Step 3: When constructing geographic segmentation proxy variables based on the map API, first, define the center point of each city as the latitude and longitude coordinates of the city government seat as the starting point and end point for calculating geographic distance and driving distance. Next, use the route planning function of the map API platform to calculate the shortest driving distance between any two city centers during non-peak hours. Subsequently, use the geodist command of the Stata software to calculate the straight-line distance on the earth's surface between the above city centers, reflecting the shortest physical distance between the cities. Finally, for each pair of cities, calculate the ratio of their actual driving distance to the straight-line geographic distance, and define it as a "geographic segmentation variable". The higher the ratio, the more significant the barrier between the two cities caused by natural geographical conditions, which is used to quantify the degree of geographic segmentation.

5. The method for calculating the inter-provincial market segmentation index based on traffic flow big data according to claim 1 is characterized in that: The steps to determine the extent of cross-provincial market segmentation are: Take the "traffic flow of starting city and ending city" generated in step 1 as the main table, and merge the city-pair level variables generated in steps 2 to 4 into the main table with "starting city" and "ending city" as keywords; match the city-level variables of the starting city and ending city generated in step 5 to the main table respectively; By formula: ln(traffic flow i,j )=α0+α1×cross-province i，j +α2×dialect distance i，j +α3×geographic segmentation index i，j +α4×ln(distance i，j )+δ×M+∈ i，j The econometric model is set, where ln(traffic volume) is the logarithm of the traffic volume between cities, interprovincial is the indicator variable of whether the starting city and the end city are across provinces, dialect distance is the cultural difference between cities, geographic segmentation index is the degree of geographical isolation, ln(distance) is the logarithm of the actual geographical distance between cities, and M is the control variable, including GDP, population and other characteristics; The reghdfe command in Stata software was used to estimate the parameters in the model using the ordinary least squares method, focusing on analyzing the estimated value of the inter-provincial variable α1 and its statistical significance. If the sign of α1 is negative and statistically significant, it means that after controlling other variables, the traffic volume between cities across provinces is significantly lower than the traffic volume between cities in the same province, that is, there is an inter-provincial market segmentation phenomenon; the larger the absolute value of α1, the higher the degree of inter-provincial market segmentation.

Images