JP5284548B2 - Profiling system using regional characteristics - Google Patents

Description

  The present invention relates to a system for analyzing regional characteristics of areas such as administrative boundaries, and more particularly to a profiling system for scoring areas according to regional characteristics and creating a report of regional characteristics in relation to bases.

[Conventional technology]
Some conventional trade area analysis systems analyze sales and profits at bases such as stores.
As related prior art documents, Japanese Patent Application Laid-Open No. 2003-167880 “Statistical Information Aggregation Device” (Kao Corporation) [Patent Document 1], Japanese Patent Application Laid-Open No. 2003-345996 “Store Information Collection and Supply Device, Store Information” Collection and providing method and program describing it "(NC System Technology Co., Ltd.) [Patent Document 2].

  Patent Document 1 is a device that designates a store on a map and displays data such as the sales amount of the store together with map information, acquires data held by the POS system, creates a display file, It is described that when data to be displayed in a file is designated, it is displayed on a map to support analysis work.

  Patent Document 2 is a device that evaluates the location of a store and provides an analysis result, receives store location information from an information provider, transmits store information to the user, and receives response information from the user. It is described that an evaluation of a user corresponding to a store position is transmitted to a person who wants to obtain information, and the user's reaction behavior is digitized and aggregated based on response information from the user, and the evaluation of the user with respect to the store position is calculated. .

JP 2003-167880 A JP 2003-345932 A

  However, although the above conventional analysis system analyzes individual stores, etc., it generates a report that makes it possible to score and evaluate according to regional characteristics and grasp the regional characteristics of a specific area including the base. There was a problem of not having.

The present invention has been made in view of the above circumstances, using the profiling data of the regional characteristics can properly analyze the local characteristics, the profiling system that can easily understand the relevance of the response data to the local characteristics The purpose is to provide.

The present invention for solving the problems of the above-mentioned conventional example is a profiling system for analyzing regional characteristics, wherein statistical data is stored in administrative boundaries or specific mesh units in a storage unit of a computer, and a control unit is programmed As a processing means to operate and execute, it is composed of multiple administrative boundaries or multiple meshes, calculates correlation between variables using statistical data or aggregated data that aggregates statistical data as variable candidates, and generates a correlation table between variables Variable determination processing means for determining the variable, and principal component analysis processing is performed for the determined variable, the factor is identified from the principal component, the principal component score of the factor is calculated, and the correlation between the variable and the principal component score is calculated. A principal component analysis processing means that calculates and generates a principal component load matrix having a correlation coefficient as a principal component load, and performs cluster analysis with factors to determine administrative boundaries or specific messages. A table of the subclass cluster analysis is generated by calculating the distance between the subclass clusters using the factor of the unit, and the average value is calculated for each factor of the subclass cluster and the average value is used. Calculate the distance between the clusters in the middle class, group the clusters according to the calculated distance value between the clusters, classify the clusters, combine the groups based on the calculated distance, Cluster analysis processing means that generates an average table for each cluster in the middle class through optimization, and for each cluster obtained by the cluster analysis processing means, based on response data including numerical values that the customer responds to store behavior And a report generation means for generating a report by aggregating numerical values that the customer has responded to .

In the profiling system according to the present invention, in the profiling system, the report generation unit divides the first value obtained by dividing the number of responses of each cluster by the number of valid households of each cluster, and the number of responses of all clusters divided by the number of valid households of all clusters. An unneutralized index divided by the second value is generated.

In the profiling system according to the present invention, the report generation means calculates the number of neutralization responses obtained by processing the number of responses from the analysis site and the location data of the customer, and the number of neutralization responses for each cluster is calculated as the number of effective households in each cluster. An index with neutralization is generated by dividing the first divided value by the second value obtained by dividing the number of neutralization responses of all clusters by the number of effective households of all clusters.

In the profiling system, the report generation means displays the administrative boundaries or meshes in different colors on the map displayed on the display unit, and uses the customer location information as response data to display the customer on the map. The position data is displayed with dots.

The present invention is a profiling program that stores statistical data in administrative boundaries or specific mesh units in a storage unit of a computer, operates on the computer, and analyzes regional characteristics, and includes a plurality of administrative boundaries or a plurality of administrative boundaries. Variable decision processing means consisting of meshes, calculating correlations between variables using statistical data or aggregated data obtained by collecting statistical data as variable candidates, generating a correlation table between variables, and confirming variables Performs principal component analysis, identifies the factors from the principal components, calculates the principal component scores of the factors, calculates the correlation between the variables and the principal component scores, and uses the correlation coefficient as the principal component loading. Principal component analysis processing means to generate a matrix and cluster analysis with factors, and calculate the distance between clusters of small classifications using factors in administrative boundaries or specific mesh units A sub-class cluster analysis table, calculate the average value for each factor of each sub-class cluster, and calculate the distance between the middle-class clusters using the average value. Clusters are grouped and classified according to the distance value between the clusters, and group recombination is optimized by combining the groups based on the calculated distance to generate an average table for each of the middle-class clusters. and cluster analysis processing means, for each cluster obtained by the cluster analysis processing means, reports to generate a report by aggregating numerical values the customer reaction based on the response data including the numerical values customer response behavior of the store generated Means.

In the profiling program according to the present invention, in the above profiling program, the report generation unit divides the first value obtained by dividing the number of responses of each cluster by the number of valid households of each cluster, and the number of responses of all clusters divided by the number of valid households of all clusters. An unneutralized index divided by the second value is generated.

In the profiling program according to the present invention, the report generation means calculates the number of neutralization responses obtained by processing the number of responses from the analysis site and the location data of the customer, and the number of neutralization responses for each cluster is calculated as the number of effective households in each cluster. An index with neutralization is generated by dividing the first divided value by the second value obtained by dividing the number of neutralization responses of all clusters by the number of effective households of all clusters.

In the profiling program according to the present invention, in the profiling program, the report generation unit displays the administrative boundaries or meshes by color on the cluster displayed on the display unit, and uses the customer location information as response data to display the customer on the map. The position data is displayed with dots.

According to the present invention, statistical data is stored in the administrative boundary unit or specific mesh unit in the storage unit of the computer, and the control unit includes a plurality of administrative boundaries or a plurality of meshes as the processing means that operates and executes the program. , Statistical data or statistical data or aggregated data of the statistical data is used as a variable candidate to calculate the correlation between variables, generate a correlation table between variables and determine the variables, and principal component analysis of the determined variables Processes, identifies the factors from the principal components, calculates the principal component scores of the factors, calculates the correlation between the variables and the principal component scores, and generates a principal component loading matrix with the correlation coefficient as the principal component loading Principal component analysis processing means to perform cluster analysis with factors, and calculate the distance between sub-class clusters using factors in administrative boundaries or specific mesh units, sub-class cluster analysis Generate a table, calculate the average value for each factor of each small classification cluster, calculate the distance between the middle classification clusters using the average value, and calculate the distance value between the calculated clusters. Cluster analysis processing means for grouping and classifying clusters accordingly, combining groups based on the calculated distance, optimizing cluster recombination, and generating an average table for each medium-class cluster, and cluster For each cluster obtained by the analysis processing means, as a profiling system having a report generation means for generating a report by aggregating the numerical values that the customer has reacted based on the response data including the numerical values that the customer has reacted to the behavior of the store Therefore, it is possible to properly analyze regional characteristics using regional characteristic profiling data, and to respond to regional characteristics. There is easily grasped can effect the relevance of the data.

1 is a configuration block diagram of a profiling system according to an embodiment of the present invention. It is a figure which shows original data. It is a figure which shows the correlation table between variables. It is a figure which shows the specific correlation table between variables. It is a figure which shows factor number determination. It is a figure which shows a principal component load matrix. It is a figure which shows a principal component interpretation table. It is a figure which shows the relationship table of an administrative boundary and a factor. It is a figure which shows the table | surface of a cluster analysis (small classification). It is a figure which shows the table | surface of a cluster analysis (medium classification). It is a dendrogram in hierarchical cluster analysis. It is a figure which shows the frequency of CL according to area. It is a figure which shows the variable average table according to CL. It is an analysis graph which shows a factor characteristic. It is a figure which shows the cluster characteristic using a response. It is a cross analysis graph of cluster characteristics.

Embodiments of the present invention will be described with reference to the drawings.
[Outline of the embodiment]
The profiling system according to the embodiment of the present invention uses statistical data and the like as variable candidates, creates a correlation table between variables, performs variable determination processing to determine variables, performs principal component analysis, identifies factors from principal components, The principal component score of the factor is calculated, the correlation between the variable and the principal component score is calculated, the principal component analysis is performed to generate the principal component load matrix with the correlation coefficient as the principal component load, and the cluster analysis is performed. , Calculate the distance between sub-class clusters and generate a sub-cluster analysis table, calculate the average value for each factor of the sub-class cluster, and use the average value between the middle-class clusters , And group and classify the clusters according to the calculated distance between the clusters, combine the groups based on the calculated distance, and optimize the cluster recombination, During ~ There row cluster analysis process of generating the average table another cluster of the kind, for each cluster, is intended to generate a report compiled on the basis of the response data about the customer, be carried out in the proper analysis treatment of regional characteristics can a shall easily understand the relevance of the response data to the local characteristics.

[This system: Fig. 1]
A profiling system according to an embodiment of the present invention will be described with reference to FIG. FIG. 1 is a configuration block diagram of a profiling system according to an embodiment of the present invention.
As shown in FIG. 1, the profiling system (this system) according to the embodiment of the present invention is realized by a computer including a control unit 11, a storage unit 12, and an interface unit 13. The display unit 14 and the input unit 15 are connected.

The storage unit 12 stores a processing program, and each processing is realized by reading the processing program from the storage unit 12 and operating it.
Further, a network system of a client server that operates a processing program on a server connected to a network, inputs an instruction from a client connected to the server via the network, and displays and outputs a processing result on the client may be used. A web system in which a server in this network system is a web server and a client is a computer that accesses the web server may be used.

[Processing content]
The processing executed by the control unit 11 of the system according to the processing program can be divided into “variable determination processing”, “principal component analysis processing”, “hierarchical / non-hierarchical cluster analysis processing”, and “report generation processing”.
Each processing means for realizing these processes by a processing program is “variable determination processing means”, “principal component analysis processing means”, “cluster analysis processing means”, and “report generation processing means”.
Each process will be specifically described below.

[1. Variable determination process]
(Variable candidate determination: Fig. 2)
An area including a plurality of administrative boundaries (for example, a specific area, the Tokyo metropolitan area, the whole country, etc.) is specified, and variable candidates are determined. Statistical data may be determined as variable candidates as they are, or statistical data may be aggregated and determined as variable candidates.

  As candidates for variables, for example, nuclear family household rate (calculation formula: number of nuclear family households / number of households), youth population rate (calculation formula: (15-19 year old population + 20-24 year old population) / total population), detached house There is a household rate (calculation formula: number of detached houses / number of households).

As the variable candidates, statistical data to be used by the operator who performs profiling for the area or the aggregated data thereof is determined.
An example of the original data is shown in FIG. FIG. 2 is a diagram showing original data.
As shown in FIG. 2, for example, numerical values for variables 1 to 4 are calculated for each administrative boundary (for example, town chome).

In addition, although the said area was made into the administrative boundary, it may be a mesh area of 1 km × 1 km, 500 m × 500 m, 250 m × 250 m provided by the national census. In addition, from the national census, statistical data is provided for each mesh area.
In addition, when assuming an arbitrary mesh other than the above, for example, an area of 300 m × 300 m, it is possible to apportion statistical data using the area apportionment of the mesh and apply it to an area of an arbitrary mesh. is there.
In the following, the part described as “administrative boundary” can be read as the above “mesh”.

Furthermore, the variable candidates are not limited to static data, but may be data that varies with time, day, month, season, and year. As the data that fluctuates, there is data obtained by counting the position data of mobile phones, smartphones, etc., and the data of non-contact type IC cards of transportation facilities in the above period.
Even abstract data may be a candidate for a variable if it can be evaluated and converted into a numerical value.

(Normalization processing)
It is desirable to use original data that can be distributed as normal data.
Therefore, the original data is indexed, proportioned, reciprocalized, and logarithmized into a normal distribution.
Further, the skewness and kurtosis of the original data are calculated and compared with a preset reference value (for example, within ± 1.5) to determine whether normal distribution is appropriate. If normal distribution is inappropriate, the variable should not be used in principle. However, even original data that cannot be normalized is used dare if it is determined to be necessary.
Further, the quantile of the original data is extracted, it is determined whether there is a protruding portion, and if there is a protruding data portion, a process for removing the data portion is performed.

(Inter-variable correlation table creation processing: FIGS. 3 and 4)
The correlation with all variables is calculated for the normalized variable, and an inter-variable correlation table is created.
The variable correlation table will be described with reference to FIG. FIG. 3 shows a correlation table between variables.
As shown in FIG. 3, variable candidates determined and normalized are subjected to correlation calculation with brute force, and the calculated correlation values are set in the correlation table between variables.
In FIG. 3, candidates for the same variable are arranged vertically and horizontally, and correlation values are set. The correlation between the same variables is “1.00”, and if it is “0.50” or more, the correlation is high.

A specific inter-variable correlation table is shown in FIG. FIG. 4 is a diagram showing a specific inter-variable correlation table.
As shown in Fig. 4, the variables are "Young population rate", "Permanent employment rate", "Full-time employment rate", "Rate of households with relatives over 65", "Annual income over 15 million yen", "Number of general households" ... etc. are used to calculate the correlation between all variables.

Here, the above variables are defined as follows.
“Youth population rate” (“Youth rate” in FIG. 4) = Population 14 to 24 years old / Total population “Permanent employment rate” = Number of permanent employees / Working population (employed population + total unemployed population)
“Full-employed population rate” (“Full-employed population 2” in FIG. 4) = Completely-employed population / Labor force “Percentage of households with relatives over 65 years old” = Number of households with relatives over 65 years old / Total number of households “Annual income 15 million "Yen or higher household rate"("15million" in Fig. 4) = Annual income of 15 million yen or more / Total number of households "Number of general households"("05 general households" in Fig. 4) = Total number of households

(Variable confirmation processing)
In the correlation table between variables, if there is a variable with 0.5 or more with reference to the allowable range ± R (correlation value) = 0.5 as a standard, select the combination of variables with a low correlation coefficient It is desirable to do.

[2. Principal component analysis processing]
(Principal component analysis)
Principal component analysis, when there are many quantitative variables, considers the correlation structure between them and converts them into low-dimensional synthetic variables (principal components), which are interpreted from the structure of the data It is an analysis to make it easier to do.
There are two types of principal component analysis: “correlation coefficient matrix” and “dispersion covariance matrix”. Here, “correlation coefficient matrix” is used.

As the correlation coefficient matrix R, the first principal component z ₁ is obtained from the eigenvector corresponding to the first eigenvalue (maximum eigenvalue) λ ₁ of the correlation coefficient matrix R, and then the second eigenvalue λ ₂ of the correlation coefficient matrix R is obtained. the second principal component z ₂ from the eigenvector corresponding to the determined. Similarly, the k-th principal component is obtained.
The eigenvalue represents how many pieces of information the original component has, and the total eigenvalue is equal to the number of variables.
Then, the contribution rate and cumulative contribution rate of each main component are calculated. The contribution ratio of the principal component indicates how much the principal component is in the entire information (number of variables).

ここで、ｘバーは平均を、ｓは標準偏差を示している。 Specifically, variables x _1, x _2, ..., if there is x _p, standardized variables, normalized values u _1, u ₂ of the following formula (1), ..., calculated u _p To do.

Here, x bar represents an average, and s represents a standard deviation.

ここで、ａは、固有ベクトルである。 Then, the first principal component z ₁ is calculated as shown in the following formula (2). The second principal component z _{2 and} subsequent values are calculated in the same manner.

Here, a is an eigenvector.

Furthermore, a correlation coefficient matrix R is generated as shown in the following equation (3).

Next, the contribution ratio of the k-th principal component is calculated by the following equation (4). Here, λ ₁ is the eigenvalue of the first principal component, λ ₂ is the eigenvalue of the second principal component, λ _k is the eigenvalue of the k-th principal component, and λ _p is the eigenvalue of the p-th principal component.

The cumulative contribution rate up to the k-th principal component is shown in the following formula (5).

(Principal component number determination processing: FIG. 5)
For the eigenvalue and the cumulative contribution calculated in the principal component analysis process, the main component is selected based on the eigenvalue of “1 or more” or the cumulative contribution of “over 80%”. Here, the number of selected principal components is determined.
FIG. 5 shows a case where the principal component is selected. FIG. 5 is a diagram showing determination of the number of factors. As shown in FIG. 5, the eigenvalues of the principal components are calculated and rearranged in descending order, the principal components having an eigenvalue of “1.0” or more are selected as “factors”, and the number of factors is determined.

(Principal component score calculation processing)
When the number of factors is determined, principal component scores (factor scores) are calculated for the determined factors (for example, factors 1 to 9).

(Principal component load matrix creation processing: FIG. 6)
The correlation between the principal component score of the factor calculated in the principal component score calculation process and the original variable is calculated, the correlation coefficient is set as the principal component load amount, and the matrix is generated as the principal component load amount matrix.

The principal component load amount of the first principal component is calculated by the following equation (6). Similarly, the principal component load amount is calculated for the second and subsequent principal components.

Here, the principal component load amount matrix will be described with reference to FIG. FIG. 6 is a diagram illustrating a principal component load matrix.
As shown in FIG. 6, the determined principal component is a factor (Factor 1 to Factor 9), and the correlation between the factor principal component score and the original variable is calculated to obtain a correlation coefficient, which is created as a matrix. is there.

(Factor interpretation table: Fig. 7)
For factors 1 to 9, a factor interpretation table is created from the characteristics of the principal components with reference to the principal component loading matrix. In other words, it is a definition for the factor number.
The factor interpretation table will be described with reference to FIG. FIG. 7 is a diagram showing a principal component interpretation table.
In FIG. 7, the factor number “1” is the name of the “city rich area and youth singles mixed area”, “the population composition ratio in the 20s to 30s is high, and there are many singles. An area with a lot of apartments and a large working population. Although there is no significant feature in the composition of annual income and savings, it is a densely populated area with high savings per square meter. ..., the factor number “9” is the name of “high asset class with a lot of company residences” and gives an interpretation of “a group with a lot of company residences and a high savings level”.

  Although not shown, in another factor interpretation table, for example, the factor number “1” is the name of “New Family Factor (city center / suburb)”, “Population 25-35 years old, 0-9 years old” "The population is large and the elderly population is small." ..., the factor number “9” is the name of “Public Corporation Residential Factor” and may give an interpretation of “High unemployment rate in areas where public corporations and public corporations live in a high ratio”.

[3. Hierarchical / non-hierarchical cluster analysis processing]
Perform hierarchical or non-hierarchical cluster analysis processing.
Specifically, the hierarchical cluster analysis process is first performed, and if the convergence does not occur or it takes too much time, the non-hierarchical cluster analysis process is performed.

In cluster analysis, two steps are iteratively calculated to classify into a specified cluster. First, n points called cluster seeds are selected. The seed is the first estimate of the cluster average.
Each measurement is then assigned to the nearest seed to form a temporary cluster. The seed is then replaced with a new cluster average and points are reassigned to that average. This process is repeated until there is no change in the cluster and the cluster analysis is completed.

As a method for calculating the distance between clusters for determining the hierarchical structure in the hierarchical cluster analysis, there are a group average method, a centroid method, a Ward method, a shortest distance method, a longest distance method, and the like.
In this system, either method may be used, but the Ward method is used.
In the Ward method, the distance between two clusters is calculated as the sum of ANOVA squares between the clusters for all variables. Clusters are merged so that the sum of squares within the cluster is minimized.
In the Ward method, clusters are combined so that the likelihood becomes maximum under the assumption of multivariate normal mixture distribution, spherical covariance matrix, and equal extraction probability.

(Hierarchical cluster analysis processing: FIGS. 8 to 10)
Next, hierarchical cluster analysis processing will be described with reference to FIGS. 8 is a diagram showing a relationship table between administrative boundaries and factors, FIG. 9 is a diagram showing a table of cluster analysis (small classification), and FIG. 10 is a diagram showing a table of cluster analysis (medium classification). It is.
First, as shown in FIG. 8, for each administrative boundary, for example, factor scores of factors 1 to 3 are calculated to generate a relational table. In FIG. 8, “every administrative boundary” is shown, but “every mesh” may be used.
The calculation of the factor score for each administrative boundary is obtained when the principal component load matrix of FIG. 6 is created.

  Next, cluster analysis processing is performed based on the factor scores of factors 1 to 3 for each administrative boundary, the distance is calculated using factors 1 to 3 for each administrative boundary, and the calculated distance is shown in FIG. Thus, a cluster is specified for each administrative boundary, a cluster number (small classification cluster number) is assigned, and a cluster analysis (small classification) table is generated.

  Then, the average of the factor scores for each factor is calculated for each cluster number, and as shown in FIG. 10, the average value of the factor scores for factors 1 to 3 is set for each cluster number, and further the cluster analysis process is performed. For each classification cluster number, the distance is calculated using the average value of the factor scores of factors 1 to 3, and based on the calculated distance, the middle classification cluster is identified for each small classification cluster, and the cluster number (medium classification A cluster analysis (medium classification) table (average table) is generated.

There are two methods for creating a middle classification by cluster analysis: a method for obtaining an average factor score for each cluster number and further performing cluster analysis; and a method for determining the number of middle classifications based on a tree diagram described later.
Although it is described in FIG. 10 that the average value is set, actually, a plus / minus value (a value obtained by converting ± to the average value) is set with respect to the average value aggregated for each factor.

In the case of the small classification and the middle classification, the cluster is always interpreted at each stage to determine the appropriate number of classifications.
In the cluster analysis process described above, cluster analysis is performed using only factors. However, cluster analysis may be performed using variables.
In this way, the cluster number is converged in the hierarchical cluster analysis.

(Tree diagram of hierarchical cluster analysis: Fig. 11)
Next, a tree diagram of hierarchical cluster analysis will be described with reference to FIG. FIG. 11 is a tree diagram in the hierarchical cluster analysis.
In FIG. 11, the analysis method employs a principal factor method, rotation uses orthogonal rotation, and the number of clusters (CL) is set to “50 CL” with respect to the number of factors of 9.
The tree diagram is generated based on the distance value calculated by the above-described calculation method for determining the distance between clusters.

(Cluster number determination process)
Then, cluster recombination is optimized by combining the clusters, and the number of clusters is determined.
For example, although the number of clusters is “50CL”, it can be rearranged to “20CL” by conjugation.

(CL frequency creation processing by area: Fig. 12)
CL frequency creation processing is performed for each area.
Specifically, as shown in FIG. 12, a list is generated by calculating the number of town chomes corresponding to factors 1 to 9 for each cluster by area (by prefecture). FIG. 12 is a diagram illustrating the frequency of CL for each area.
Referring to FIG. 12, clusters can be interpreted by area-specific trends for factors.

(Variable average table creation processing by CL: FIG. 13)
Processing for creating a variable average table for each CL will be described with reference to FIG. FIG. 13 is a diagram showing a variable average table for each CL.
In FIG. 13, for each cluster, the “number of town-letter cases” for each factor, the population ratio by age, the ratio of the number of households, etc., for example, a score is calculated with a Z score, and an average table is generated.
Referring to FIG. 13, the cluster can be interpreted based on the tendency of variables with respect to factors.
Further, a variable average table based on Z score scores may be generated as variables based on the ratio of workers in the industry, annual income, and the like.

[4. Report generation process]
Based on the cluster analysis process, a map is displayed, and the administrative boundaries on the map are displayed in different colors for each cluster to generate a report. As a result, the cluster distribution on the map can be easily recognized.
Based on the above cluster analysis process, the population and composition ratio, the number of households and composition ratio, the number of areas and composition ratio for each CL are displayed in a list, and the cluster characteristics in a specific area such as the Tokyo metropolitan area are reported. Can be displayed as

Further, for a specific area, for example, the Tokyo metropolitan area, the distribution may be displayed by taking the average age on the horizontal axis and the average annual income on the vertical axis, and reflecting the size of the number of households by the diameter of the circle for each CL. In this case, the color of the circle is changed for each CL, or CLs with similar principal component scores (CLs with a short distance) are displayed in the same color as a group.
Thereby, the relationship between the average age of a specific area and the average household income can be recognized for each CL.

  In addition, when using a store as a base, set one or more trade areas by radius (distance such as 1 km, 2 km, 3 km, etc.) or travel time (access time such as 5 minutes, 10 minutes, 30 minutes, etc.) You may specify.

(Factor characteristics: Fig. 14)
The factor characteristics shown in FIG. 14 can be displayed from the relation table of administrative boundaries and factors in FIG. 8 obtained by principal component analysis. FIG. 14 is an analysis graph showing factor characteristics.
As shown in FIG. 14, based on the relationship table between administrative boundaries and factors, the analysis area trade area, national average factor, metropolitan area average factor, urban area average factor, local area average factor for the item (factor name) corresponding to the factor Generate an analysis graph using and generate a report.
Specifically, from the relationship table between administrative boundaries and factors, administrative boundaries included in the analysis area (for example, trade area, urban area, local area, etc.) are extracted, and the average score (average factor) for each factor is calculated. To do.

Here, the trade area of the analysis area is the average score (average factor) of the factor scores in the specific area to be analyzed, the national average factor is the average factor of prefectures nationwide, It is the average factor of Kanto (Tokyo, 3 prefectures: Tokyo, Saitama, Chiba, Kanagawa), and the metropolitan average factor is Kanto, Tokai (3 prefectures: Aichi, Gifu, Mie), Kinki (2 prefectures and 4 prefectures: Osaka prefecture, Kyoto prefecture, Shiga prefecture, Nara prefecture, Wakayama prefecture, Hyogo prefecture). The regional average factor is the average factor of prefectures other than Kanto, Tokai and Kinki. is there.
In FIG. 14, the factor characteristics of the trade area of a specific area (analysis area A) can be compared with the average factors (average score) of Tokyo, the whole country, the metropolitan area, the urban area, the local area, and the like.

  In FIG. 14, the regional characteristics of the business area of the base A and the areas such as Tokyo are compared. For example, the regional characteristics of the business areas of the base B and the base C may be plotted in one figure. Thereby, there is an effect that the regional characteristics can be compared for the trade areas of a plurality of bases.

  In FIG. No. 1 is “New Family (City Center / Nearby)”. 2 is “My Home Nuclear Family”, No. 3 is “Three-generation large family”, No.3. No. 4 is “elderly rural village”. No. 5 is “New Family (City / Suburb)”. 6 is “manufacturability”. 7 is “suburban family”. No. 8 is “Company Residential”. 9 adopts the factor name of “public housing”.

For example, no. “New family characteristics (city center / suburbs)” is “25-39 years old population, 0-9 year old population is large. Conversely, the elderly population is extremely low. There are many apartment houses and the labor force population is very high. And high score in densely populated areas such as cities designated by government ordinances and prefectural offices.
No. “New family characteristics (city center / suburb)” of 5 is “30-44 years old population, 0-14 years old population is large, and the composition ratio of baby boomers is low. There are many apartment residents, factor 1 and distribution area overlap. However, these factors are also scattered in the suburbs. "

(Cluster characteristics using response: Fig. 15)
A trade area is determined by identifying an area (an area determined by a radius or a drive time) by taking a store as a base.
It is also possible to display an indicator of regional characteristics for a base such as a store by superimposing a response such as a store on the cluster characteristics.
The response is the number of visitors to the direct from the store or an answer to the questionnaire from the store or the like.
The regional characteristic index indicates to which cluster number the store belongs.

  In addition, the administrative boundaries in the trade area are displayed on a map, the administrative boundaries or meshes are displayed in different colors for each cluster, and the customer's location data is displayed on the map using the location information such as the customer's address as response data. Can be displayed. As a result, the distribution status of the response data (customer position data) can be easily grasped with respect to the regional characteristics on the map, and the relevance of the response data to the regional characteristics can be easily grasped.

The cluster characteristics using the response will be described with reference to FIG. FIG. 15 is a diagram illustrating cluster characteristics using a response.
As shown in FIG. 15, subclasses and names are assigned to each CL, and for each CL, the number of responses and composition ratio, the number of neutralization responses and composition ratio, the number of households and composition ratio, the number of effective households and composition ratio, and the penetration The rate, index (without neutralization), and index (with neutralization) are shown.

Here, the number of responses is the sum of the number of cluster responses when the number of households is 0 or the streets whose response rate exceeds 100% are excluded. The composition ratio is a composition ratio for each cluster in the trade area.
The number of neutralization responses is the number of responses multiplied by the square of travel time from the analysis site, which was canceled based on the idea that “the customer decays in inverse proportion to the square of distance”. Is.

The number of effective households is the sum of the number of cluster households excluding the number of responses, the number of households being 0, or the town chome where the response rate exceeded 100%.
The penetration rate is calculated by the number of responses / number of effective households.

The index (no neutralization) is calculated by (number of each cluster response / number of effective households in each cluster) / (number of total cluster responses / number of effective households in all clusters).
The index (with neutralization) is calculated as (number of neutralization responses for each cluster / number of effective households for each cluster) / (number of neutralization responses for all clusters / number of effective households for all clusters).

[Cross analysis of cluster characteristics: Fig. 16]
In FIG. 15, the small classifications are those of the “Metropolitan area 01-30” classification, but as other small classifications, the “national 01-30” classification version, the “urban area 01-30” classification version, the “local area” 01-30 "classification version.
FIG. 16 shows the appearance rate of clusters with respect to variables of population density and household annual income in the subcategory “National 01-30”. FIG. 16 is a cross analysis graph of the cluster appearance rate and population density / household annual income.

The size of the circle in FIG. 16 represents the size of the appearance rate.
Here, the cluster shown on the right side of FIG. 16 is a middle class, and is “year resident”, “city resident”, “elderly countryside”, or the like. The middle class cluster contains a minor class cluster.
In FIG. 16, the reason why a plurality of circles of the same cluster exist is because a small classification cluster is displayed. Therefore, in FIG. 16, it is preferable to display the cluster number of the small classification in each circle so that the relationship with the cluster of the small classification becomes clear.

[Effect of the embodiment]
According to this system, regional characteristics can be analyzed using regional characteristics profiling data, and a report can be generated showing what kind of regional characteristics a specific area that includes a base has. It is possible to easily grasp the regional characteristics of a specific area.

In addition, it is possible to display the regional characteristic index of the base by superimposing the response data of the base on the report, and it is possible to easily grasp the relevance of the response data with respect to the local characteristic of the specific area including the base.
In addition, in the case of a store, a report using regional characteristics can be used as a reference for product assortment, shelf allocation, etc., and can also be used for store evaluation, store scrap and build, and investment decisions. .

The present invention utilizes the profiling data of the regional characteristics can properly analyze the regional characteristics, it is suitable for profiling system that can easily understand the relevance of the response data to the local characteristics.

  DESCRIPTION OF SYMBOLS 11 ... Control part, 12 ... Memory | storage part, 13 ... Interface part, 14 ... Display part, 15 ... Input part

Claims (8)

A profiling system that analyzes regional characteristics,
As a processing means for storing statistical data in administrative boundaries or specific mesh units in the storage unit of the computer, and for the control unit to operate and execute the program,
Variable decision processing means that consists of multiple administrative boundaries or multiple meshes, calculates the correlation between variables using statistical data or aggregated data obtained by collecting statistical data as variable candidates, generates a correlation table between variables, and determines the variables When,
Principal component analysis is performed on the determined variable, the factor is identified from the principal component, the principal component score of the factor is calculated, the correlation between the variable and the principal component score is calculated, and the correlation coefficient is calculated as the principal component A principal component analysis processing means for generating a principal component load matrix as a load,
A cluster analysis is performed using the factors, and a sub-cluster analysis table is generated by calculating a distance between sub-class clusters using the administrative boundary unit or the specific mesh unit factor. The average value is calculated for each of the other factors, the distance between the middle class clusters is calculated using the average value, and the clusters are grouped and classified according to the calculated distance value between the clusters. Cluster analysis processing means for combining groups based on the calculated distance and optimizing cluster rearrangement to generate an average table for each of the middle class clusters;
For each cluster obtained by the cluster analysis processing means, there is a report generation means for generating a report by aggregating the numerical values reacted by the customer based on response data including numerical values reacted by the customer with respect to the store behavior Profiling system characterized by The report generation means neutralizes the first value obtained by dividing the number of responses for each cluster by the number of effective households for each cluster by the second value obtained by dividing the number of responses for all clusters by the number of effective households for all clusters. The profiling system according to claim 1, wherein an index of none is generated. The report generation means calculates the number of neutralization responses obtained by processing the number of responses from the analysis site and customer location data, and calculates the first value obtained by dividing the number of neutralization responses for each cluster by the number of effective households for each cluster. The profiling system according to claim 1, wherein an index with neutralization is generated by dividing the number of neutralization responses of clusters by a second value obtained by dividing the number of effective households of all clusters. The report generation means displays the administrative boundaries or meshes by color on the cluster displayed on the display unit, and displays the customer location data as dots on the map using the customer location information as response data. The profiling system according to claim 1, wherein the profiling system is a profiling system. A profiling program that stores statistical data in administrative boundaries or specific mesh units in a computer storage unit, operates on the computer, and analyzes regional characteristics,
Variable decision processing means that consists of multiple administrative boundaries or multiple meshes, calculates the correlation between variables using statistical data or aggregated data obtained by collecting statistical data as variable candidates, generates a correlation table between variables, and determines the variables When,
Principal component analysis is performed on the determined variable, the factor is identified from the principal component, the principal component score of the factor is calculated, the correlation between the variable and the principal component score is calculated, and the correlation coefficient is calculated as the principal component A principal component analysis processing means for generating a principal component load matrix as a load,
A cluster analysis is performed using the factors, and a sub-cluster analysis table is generated by calculating a distance between sub-class clusters using the administrative boundary unit or the specific mesh unit factor. The average value is calculated for each of the other factors, the distance between the middle class clusters is calculated using the average value, and the clusters are grouped and classified according to the calculated distance value between the clusters. Cluster analysis processing means for combining groups based on the calculated distance and optimizing cluster rearrangement to generate an average table for each of the middle class clusters;
For each cluster obtained by the cluster analysis processing means, there is a report generation means for generating a report by aggregating the numerical values reacted by the customer based on response data including numerical values reacted by the customer with respect to the store behavior Profiling program characterized by
The report generation means neutralizes the first value obtained by dividing the number of responses for each cluster by the number of effective households for each cluster by the second value obtained by dividing the number of responses for all clusters by the number of effective households for all clusters. 6. The profiling program according to claim 5, wherein an index of none is generated. The report generation means calculates the number of neutralization responses obtained by processing the number of responses from the analysis site and customer location data, and calculates the first value obtained by dividing the number of neutralization responses for each cluster by the number of effective households for each cluster. 6. The profiling program according to claim 5, wherein an index with neutralization is generated by dividing the number of neutralization responses of clusters by a second value obtained by dividing the number of effective households of all clusters. The report generation means displays the administrative boundaries or meshes by color on the cluster displayed on the display unit, and displays the customer location data as dots on the map using the customer location information as response data. 8. The profiling program according to claim 5, wherein the profiling program is any one of claims 5 to 7.

Images