JP2014219785A - Market research and analysis system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable a highly accurate market research/analysis to be performed.SOLUTION: A market research/analysis system 1 includes: waveform generation means for corrugating client data accumulated in a database; means for calculating a difference of a point interval between at least two or more waveforms generated by the waveform generation means; means for calculating a difference between at least two or more segment angles of the waveforms generated by the waveform generation means; means for determining whether or not a difference between point intervals falls within a threshold value; means for determining whether or not a difference between the segment angles falls within the threshold value; means for calculating a ratio by which a difference between the point intervals falls within the threshold value; means for calculating a ratio by which a difference between the segment angles falls within the threshold value; and means for calculating an approximation degree of at least two or more waveforms from a ratio by which a difference between the point intervals falls within the threshold value and a ratio by which a difference between the segment angles falls within the threshold value.

Description

  The present invention relates to a market research / analysis system, and more particularly, to calculate data for providing an advertisement in accordance with a customer's hobbies and intentions.

  Encourage customers to purchase more products and services, such as point cards, credit card subscribers, sales network subscribers, and customer registrants who can obtain customer information and provide advertisements to customers. For this purpose, customer-specific advertisements such as issuing coupons at POS terminals, sending e-mails with advertisements, displaying customized advertisements on the member login screen on websites, direct mailing to customers, etc. Advertising is promoted using a so-called recommendation system to be presented. Previously, all types of coupons and advertisements were provided to customers, but only for customers with hobbies and orientations suitable for specific advertisements from the viewpoint of efficient advertising and cost reduction. Since providing advertisements can provide efficient advertisements with a large advertising effect, market research and analysis systems and recommendation systems have been devised for extracting target customers to provide specific advertisements. For example, a recommendation system that extracts advertisements to be presented based on customer attributes and purchase history (Patent Documents 1, 3, and 4), and a service coupon issue system that extracts service coupons to be presented based on customer attributes and purchase history (Patent Document 2). These market research and analysis systems and recommendation systems use member customer attributes and purchasing history as advertisement extraction profiles, and combine them with external factors such as weather and traffic information for specific advertisements. Member customers are extracted, and the criteria for extraction are fixedly set.

JP 2010-211687 A JP 2009-163533 A JP 2012-247926 A JP 2004-70504 A JP 2012-190061 A

  In the conventional market research and analysis system and recommendation system described above, it is possible to anticipate the customer's hobbies and intentions and to provide advertisements according to the customer's hobbies and intentions. However, companies have a variety of corporate activities such as product development, store planning, advertisement planning, selection of advertisement targets and advertisement delivery methods, and consumer hobbies and preferences as well as characteristics of stores, products, services, and areas. There is a demand to analyze the information accurately and use it for corporate activities. Accordingly, the present invention provides a system that enables reliable market research and analysis, and can provide the market research / analysis results in an easy-to-understand format to alliance companies that are users of the market research / analysis system. Objective.

  The market research / analysis system of the present invention includes a waveform generation means for converting customer data accumulated in a database into a waveform, and a means for calculating a difference between at least two waveform points generated by the waveform generation means. Means for calculating a difference between line segment angles of at least two or more waveforms generated by the waveform generating means; means for determining whether or not the difference between the points falls within a threshold; and Means for determining whether or not the difference falls within a threshold; means for calculating a ratio at which the difference between the points falls within the threshold; and means for calculating a ratio at which the difference between the line segment angles falls within the threshold; There is provided means for calculating an approximation degree of at least two or more waveforms from a ratio at which the difference between the points falls within a threshold and a ratio at which the difference between the line segment angles falls within the threshold.

  According to another aspect of the present invention, the waveform generation means generates a line graph.

  The waveform generation means has means for calculating averaged customer data.

  A means for selecting one or more destination data and a means for selecting one or more destination data are provided. The waveform generation means generates a waveform of the source data and a waveform of the destination data, and calculates the degree of approximation of the waveform. The means for calculating calculates the degree of approximation of the waveform of the destination data with respect to the source data.

  It has a table creation means for displaying a list of approximations of destination data and destination data.

  The destination data and the destination data can be selected from at least one category of customers, companies, stores, products, services, areas, and the like.

  It has a prediction processing means for calculating predicted data based on actual data stored in a database.

  The present invention is characterized in that the actual data is discretized and predicted data is calculated by the prediction processing means based on the discretized actual data.

  The forecast data calculated by the forecast processing means is customer data.

  It is characterized in that customer data is created by complementing missing items in actual data with predicted data calculated by a prediction processing means.

  The prediction data is characterized by comprising a probability value or a prediction value obtained by threshold determination of the probability value.

  It has a table creation means for displaying a list of customer data for a customer ID.

  It has a recommendation means for associating a recommendation medium of a company, a store, a product / service or an area for which an arbitrary degree of approximation is calculated for an arbitrary customer ID.

  It has a recommendation means for associating a customer ID having an arbitrary degree of approximation with a recommendation medium corresponding to an arbitrary company ID, store ID, product / service ID, or area ID.

  The headquarters terminal has destination data and destination data selection means and market research / analysis result browsing means.

  According to the market research / analysis system of the present invention, market research / analysis can be performed with high accuracy.

It is a system outline figure of the present invention. It is a flowchart of the system of this invention. It is a flowchart of the system of this invention. It is an example of the data table produced in the system of this invention. It is an example of the data table produced in the system of this invention. It is an example of the data table produced in the system of this invention. It is an example of the data table produced in the system of this invention. It is a flowchart of the system of this invention. It is a flowchart of the system of this invention. It is an example of the graph produced | generated in the system of this invention. It is an example of the graph produced | generated in the system of this invention. It is an example of the graph produced | generated in the system of this invention. It is a figure explaining the calculation process of the approximation degree in the system of this invention. It is a figure explaining the calculation process of the approximation degree in the system of this invention. It is an example of the data table which shows the usage example of the approximation degree in the system of this invention. It is a flowchart of the system of this invention.

  As shown in FIG. 1, the market research / analysis system 1 of the present invention is a company that is connected to an operation company system 2 and an operation company system 2 consisting of a server and a database group under the jurisdiction of the operation company. Store terminal 3 such as a POS terminal installed in the store, store terminal 3 such as a POS terminal connected to the operating company system 2 through the network 4 so as to be able to transmit information, a customer mobile terminal such as a mobile phone held by a customer, a smartphone, etc. 5, customer computer terminal 6 (hereinafter simply referred to as customer terminals 5 and 6), headquarters terminal 7 etc. installed in the company.

  As shown in the flowchart of FIG. 2, the market research / analysis system 1 of the present invention includes an actual data collection / accumulation unit that collects actual data as a previous stage of analysis, an estimation processing unit that calculates expected data from actual data, and actual data. And / or a waveform generating means for generating a waveform from customer data comprising predicted data, a synchronization rate calculating means for calculating a degree of approximation (hereinafter referred to as a synchronization rate) from the generated waveform, and a recommendation means added as necessary. Etc. are comprised.

(Actual data collection and storage means)
The actual data collection and storage means will be described with reference to FIGS. This means is a means for collecting actual customer data and storing the collected actual data in a database as a pre-stage of market research and analysis.

  The actual data collection and storage means constructed separately from the market research / analysis system of the present invention or constructed in a service point management system (not shown) forming part of the market research / analysis system of the present invention For each identifier (hereinafter referred to as customer ID) given to the customer when the customer registers the system, the actual data is collected and stored in the database. The customer ID is generally a number or symbol having an arbitrary number of digits different for each customer. As an example, the customer ID is recorded on a service point card held by the customer, and the customer ID can be read and input by the store terminal 3 and the customer terminals 5 and 6.

  An example of actual data collection is shown. When the customer registers the use of the system, the basic attribute data of the customer is transmitted to the operating company system 2 from the store terminal 3 or the customer terminals 5 and 6. The management company system 2 receives this, associates actual data for each customer ID, and stores them in the basic attribute database. Examples of basic attribute data items include gender, age, address, residence characteristics, commuting destination area characteristics, and the like.

  In addition, whenever a customer makes a purchase at a store or an EC site of an alliance company participating in the market research / analysis system 1, the store terminal 3 or the headquarter terminal 7 sends the history system data to the operating company system 2 together with the customer ID. The management company system 2 receives this, and stores the actual data related to each item of the history data for each customer ID in the history database. Examples of the history data items include a user company, a purchased product or a purchase service, a visit time or a visit time zone, a use store, and the like.

  In addition, it is possible to collect research data collected based on responses from customers. The customer transmits the research data together with the customer ID from the customer terminals 5 and 6 to the server of the operating company system 2, and the server receives the information and associates the actual data regarding each item of the research data with the customer ID. In the research database. Examples of research data items include member questionnaire items and orientation flags. Examples of member questionnaire items include whether a customer who is a member of the system is married or unmarried, who is collected by answering the questionnaire, the presence of children, the state of residence, annual income, the presence of a driver's license, and the like. In addition, as the orientation flag item, traditional orientation (individual preference) that classifies the degree of customer awareness, innovation orientation (innovation preference), luxury orientation (high quality) (Preferred orientation) and the like, and examples include setting a staged flag according to the degree of each orientation.

  Furthermore, as an example of an item that can be added, there is net behavior data indicating a usage behavior state of a customer to a communication network such as the Internet. Items of the net behavior data include access time to a communication network, use medium, use site, and the like. Examples of other items that can be added include service points given according to the purchase amount. Service points given when a customer makes a purchase are transmitted to the operating company system 2 from the store terminal 3 or the headquarters terminal 7 or the like, and the server receives this, and is accumulated by being associated with the customer ID. The service points acquired are added to the service points, and the service points are accumulated in the service point database as needed. Note that the above items are examples, and the actual data collected and accumulated is not limited to these items.

(Expected processing means)
Next, prediction processing means will be described with reference to FIGS. Various data such as basic attribute data, history data, research data, net behavior data, service point data are stored in the basic attribute database, history database, research database, and other databases by the actual data collection and storage means. It is also updated from time to time. However, collection and collection of actual data is not performed for all items. For example, when the actual data table as shown in FIG. 4 is created by the table creating means, the actual data is displayed in a state linked to the customer ID. Some items, such as the history data item of the store that has not been done or the research data item that was not answered to the questionnaire, are in a state where actual data is missing for various reasons. In addition, some actual data may include prominent data that is not suitable for the basis of market research and analysis for some reason such as customer or store input mistakes. Therefore, the forecast data is calculated from the actual data in the forecast process, and customer data as a basis for market research and analysis is created by the customer data creation process.

  When the operating company system 2 operates the prediction processing means, the prediction processing means uses the probabilistic reasoning algorithm from actual data stored in a basic attribute database, a research database, a history database, a point database, and other databases. The forecast process for forecasting forecast data for each item is performed. Examples of probabilistic reasoning algorithms include, but are not limited to, various methods such as Bayesian networks, neural networks, and random forests.

  Referring to FIG. 5, a first embodiment of the prediction process is shown. FIG. 5 shows a complete table in which the prediction processing means performs the probability inference, and calculates the prediction data including the prediction value obtained by determining the probability value of the inference result with the threshold value. The server of the management company system 2 calculates the probability of data for each item of the target customer from the actual data of the other customer item having actual data close to the target customer in the actual data of the basic item. Predictive processing is performed based on a probabilistic reasoning algorithm. For example, when the marriage probability of customer A (person with customer ID 100001) is predicted, the server of operation company system 2 uses actual data of other items (for example, gender data, age data, address data, history data, etc.) From the actual data of the item “Separate Married / Unmarried” of another customer who has actual data close to customer A, the marriage probability value of customer A (for example, the marriage probability value 70%) is calculated. Similarly, when predicting the probability of purchasing the product “Daisy” of Company A in the history data of customer A (for example, the person with ID100001), the server of operating company system 2 uses actual data (for example, gender) of other items. Data, age data, address data, history data) Probability value for purchasing the product “Saiten” from the actual data of the item “Saiden” of the product of other customer A who has actual data close to Customer A ( For example, a 10% probability of purchasing the product “garden” is calculated. Furthermore, due to the nature of the item, for items that should be represented as YES or NO, or 100% or 0% (eg, married or unmarried), YES if the probability value is greater than or equal to the threshold, If there is NO, similarly, a prediction process for determining an expected value of 100% if it is equal to or greater than the threshold and 0% if it is equal to or less than the threshold may be added. This threshold value is variable. For example, when the threshold value is set to 50% with respect to the inference result of the marriage probability value (for example, the marriage probability value 70%) of the customer A, the expected value of “married” is calculated for the customer A. . In the example of the complete data table shown in FIG. 5, the prediction process is performed only for the item for which the actual data is missing, or the actual data is missing after performing the prediction process for all the items. By associating the forecast data only with the item, the forecast data consisting of the actual data and the forecast value is used as the customer data. However, the forecast processing is performed for all items and the forecast data is linked to all the items so that only the forecast data is obtained. You may make it calculate the customer data which consists of.

  Also, at the stage of collecting and accumulating actual data, discretization processing is performed by various methods such as rounding or dividing the actual data numerical value to eliminate the outstanding data, and create discretized actual data Then, the discretized actual data may be predicted. That is, the actual data used as the basis of the prediction process may be collected raw actual data or discrete actual data.

  Referring to FIG. 6, a second embodiment of the prediction process is shown. FIG. 6 shows a complete table in which the prediction processing means performs probability inference, the probability value of the inference result is directly used as prediction data, and the missing portion of the actual data is complemented with the prediction data including the probability value. In this case, customer data composed of prediction data composed of actual data and probability values is calculated.

  Referring to FIG. 7, a third embodiment of the prediction process is shown. FIG. 7 shows a complete table in which the prediction processing means performs probability inference, the probability value of the inference result is directly calculated as prediction data, and all items are filled with prediction data including probability values. In this case, customer data composed of prediction data in which all items are made up of probability values is calculated.

  Further, as shown in FIGS. 5 to 7, the table creation means provided as necessary creates a complete table displayed in a state where customer data is linked to each item of the list. . This makes it possible to visually provide a complete table in which customer data is linked to a list and has no blank space.

(Waveform generation means)
The waveform generation means will be described with reference to FIGS. The waveform generation means is means for plotting customer data on a graph and converting it into a waveform.

  When the operating company system 2 operates the waveform generation means, the waveform generation means performs plot processing for plotting customer data of each item corresponding to the category taken along the vertical axis (S1), and waveform generation processing to form a line graph (S2) is performed. For example, when generating the waveform of customer A, as shown in FIG. 10, the horizontal axis represents items, the vertical axis represents customer data (%), and each item of customer A (for example, annual income level, smoking rate, car) Plot customer data against (holding ratio) and waveform it. Thereby, the property of customer A is represented by the waveform. Similarly, the waveform of customer B can also be represented.

  In another form of waveform generation means, the horizontal axis is the item, the vertical axis is the customer data (%), the average value of the customer data of each item is calculated, and the plot processing for plotting the averaged customer data is performed. And a waveform generation process for forming a line graph is performed. For example, as shown in FIG. 11, when generating a waveform of a product G, each item of customers having historical data on the purchase of the product G, where the horizontal axis is the item and the vertical axis is the customer data (%) An average value of other customer data is calculated, and averaged customer data for each item of the product G (for example, annual income level, smoking rate, car ownership rate) is plotted, and this is converted into a waveform. For example, the plot processing of the item “smoking rate” of the product G calculates the average value of the smoking rate of customer data of all purchasers of the product G or specific purchasers belonging to a specific group such as gender, age group, This average value for “smoking rate” is plotted as customer data. Such processing is performed for all items, and a waveform is generated. As a result, the purchasers of the product G are represented by waveforms. As shown in FIG. 12, the plot processing of the item “luxury-oriented” in store A store and store B store the customer data of all users of store A store or specific users belonging to a specific layer such as gender and age group. An average value of the degree of luxury orientation is calculated, and this average value is plotted as customer data for the item “luxury orientation”. Such processing is performed for all items, and a waveform is generated. Note that such waveform generation processing can be performed on analysis targets of various categories such as companies, areas, and ages as well as customers, products, and stores. 10 to 12 show an example in which two waveforms are generated, one waveform may be generated, or three or more waveforms may be generated. Although the line graph is suitable for calculating the synchronization rate described later, other types of graphs such as a curve graph, a function graph, a scatter diagram, an area graph, and a radar chart can be generated.

  As shown in FIG. 9, in order to compare a plurality of data, destination data and destination data can be selected in advance. Here, the original data means the average value of the member customer group created based on the attribute, orientation, purchasing tendency, etc., which is a model for calculating the degree of approximation, that is, the synchronization rate, and the destination data is the original data and The average value of a target person or a group of target persons created based on attributes, orientation, purchasing tendency, etc. for determining the synchronization rate. For example, as shown in FIG. 9 and FIG. 10, in order to compare the waveforms of customer data of customer A and customer B, customer A (customer ID) is selected as the destination data and customer B (customer ID) is selected as the destination data. By doing so (S1), the server of the operating company system 2 extracts customer data corresponding to the customer ID of customer A and the customer ID of customer B (S2), and customer data of customer A and customer B (for example, Then, plot processing (S3) for plotting annual income level rate, smoking rate, car ownership rate) and waveform generation processing (S4) for plotting each plot into a line graph are performed.

  In addition, as shown in FIG. 11, when customer A is selected as the destination data and product G is selected as the destination data, the server of the operating company system 2 stores the customer data of customer A and the customer who has purchased the product G. The customer data averaged is calculated, and waveforms of customer A and product G are generated.

  In addition, as shown in FIG. 12, when store A is selected as the destination data and store B is selected as the destination data, the server of the operating company system 2 stores the customer data averaged for the customers with store A's store history, The averaged customer data of customers with store B visit history is extracted, and the waveforms of store A and store B are generated. Select destination data and destination data from the same category, such as customer-to-customer, company-to-business, product-to-product, and area-to-area characteristics, such as customer-to-company, customer-to-product, and store-to-area characteristics. It is possible to select from different categories, and various categories such as customers, companies, stores, products / services, and area characteristics can be selected, and combinations of these categories are also arbitrary. By generating the waveform in this way, it is possible to visually confirm the feature to be compared.

(Synchronization rate calculation means)
With reference to FIG. 13 and FIG. 14, the synchronization rate calculation means as means for calculating the degree of approximation will be described. The synchronization rate refers to the degree of approximation of at least two or more waveforms generated by the waveform generation means. The synchronization rate is calculated by calculating a difference between at least two or more waveform points generated by the waveform generating unit and a line segment angle difference between at least two or more waveforms generated by the waveform generating unit. Means for determining whether or not the difference between the points falls within a threshold; means for determining whether or not the difference in line segment angle falls within the threshold; and if the difference between the points falls within the threshold Means for calculating a ratio that falls within the threshold, means for calculating a ratio that the difference between the line segment angles falls within the threshold, and a ratio that the difference between the points falls within the threshold and the difference between the line segment angles falls within the threshold. It is obtained from the ratio.

  As shown in FIGS. 8, 9, 13, and 14, when the operating company system 2 operates the synchronization rate calculation means, the synchronization rate calculation means causes each of the two or more waveforms generated by the waveform generation means described above. A difference in distance (probability difference) between points on the plot is calculated (S4). Further, the synchronization rate calculating means calculates the difference between each line of the line graph and the angle between the lines (difference between the line segment angles) (S5). Then, a threshold determination is made as to whether or not the calculated probability difference and line segment angle difference fall within the threshold (S6), and the ratio of variables falling within the threshold relative to the whole, that is, the point match rate and the waveform The coincidence rate is calculated, and the sync rate is calculated by dividing the result obtained by adding the point coincidence rate and the waveform coincidence rate by 2 (S7).

Here, the probability difference is a value indicating the distance between the corresponding plotted points by the difference of the probability values taken on the vertical axis, and referring to FIG. The probability of item X ₁ (for example, annual income level) of the item is X ₁ A% (for example, 75%), and the probability of item X ₁ (annual income level) of the item in destination data B is X ₁ B% (for example, 73%) Is plotted at the position of X ₁ A% (75%)-X ₁ B% (73%) = probability difference Y ₁ % (2%), this calculation is X ₁ ~ X _n calculating a repetition performed probability difference Y ₁ to Y _n for the points corresponding to all of the items. Next, when the value of the probability difference falls within a threshold (for example, ± 5%), it is determined that the points match, and when the probability difference does not fall within the threshold, it is determined that the points do not match. Further, the point coincidence rate is calculated for all the calculation results.

Further, the difference in line segment angle is the difference in line angle of the line graph. With reference to FIG. 13, the difference in line segment angle θ = θ ₁ −θ ₂ = tan ₋₁ D / C−tan _{− 1} D ′ / C is calculated, and this calculation is repeated for the lines corresponding to all items between X _{1 and} X _n . Next, when the difference value of the line segment angle falls within a threshold value (for example, ± 3 °), it is determined that the waveform matches, and when it does not fall within the threshold value, it is determined that the waveform does not match. Furthermore, the coincidence rate of the waveforms with respect to all the calculation results is calculated.

  Then, based on the point coincidence rate and the waveform coincidence rate obtained by the above-described calculation, the sync rate as the waveform approximation is calculated by the calculation of sync rate = (point coincidence rate + waveform coincidence rate) / 2. . The threshold value described above can be a variable condition.

  For graphs other than line graphs, such as curve graphs and function graphs, the difference between the probability difference and the line segment angle is calculated from the distance between arbitrary points on the curve, the distance between arbitrary points on the tangent, and the angle of the tangent. It is also possible to calculate the difference between the probability difference and the line segment angle by differentiating the waveform. As described above, the difference between the probability difference and the line segment angle can be obtained by various calculation methods corresponding to each graph.

  As shown in FIG. 15, the operating company system 2 may add a means for creating a list of synchronization rates of destination data with respect to destination data. In the example of the list shown in FIG. 15, the synchronization rate of the store data B to the store L of the destination data is represented with respect to the store A of the source data.

  In this way, through the forecast processing means, waveform generation means, and synchronization rate calculation means, the synchronization rate between the categories such as customer-to-customer, company-to-business, store-to-store, etc. By calculating the synchronization rate between different categories such as store, customer-to-product, customer-to-area, company-to-product, store-to-product, store-to-area, etc. In addition to being able to be used for recommending work to customers, this system can perform survey analysis for various purposes such as analysis of product and store assortments and analysis of store openings.

(Recommendation means)
Recommendation means as means that can be additionally provided in the market research and analysis system 1 of the present invention will be described. The recommendation means is a means for providing a recommendation medium to a customer registered in the market research and analysis system 1. The recommendation medium is displayed on a coupon ticket issued from the store terminal 3 or the like, an e-mail containing advertisements transmitted from the headquarters terminal 7 or the server 2 to the customer terminals 5 and 6, a member login screen on the website, etc. Examples include customized advertisements and direct mail sent to customers.

  As shown in FIG. 16, when the server of the operating company system 2 operates the recommendation means, the recommendation means assigns the customer ID of the customer corresponding to an arbitrary synchronization rate to the destination data (for example, store A). Extract. Next, the recommendation means provides the recommended medium of the store A corresponding to the original data to the extracted customers by various output means. The recommendation medium output means issues a coupon for store A to the store terminal 3 corresponding to the extracted customer's customer ID, and advertises store A to the email address corresponding to the extracted customer's customer ID. Send the content e-mail, display the customized advertisement of store A on the member login screen corresponding to the extracted customer's customer ID, or store A directly the address corresponding to the extracted customer's customer ID There are various output means such as address printing on mail.

  As another example of the recommendation means, the management company system 2 corresponds to the company ID of the company corresponding to the arbitrary synchronization rate and the arbitrary synchronization rate for the selected source data (for example, customer A). The store ID of the store or the product / service ID corresponding to any sync rate is extracted. Next, the recommendation means provides a recommendation medium corresponding to the extracted company, store, or product / service to the customer ID of the customer A by various output means. The output means of the recommendation medium is an e-mail address corresponding to the customer ID of the customer A who issues the coupon of the company, the store, or the product / service extracted to the store terminal 3 or the like corresponding to the customer ID of the customer A. Send an e-mail containing the advertisement of the extracted company, store, or product / service. Customized advertisement of the extracted company, store, or product / service to the member login screen corresponding to the customer ID of customer A Various output means such as displaying or printing the address on the direct mail of the company, the store, or the product / service from which the address corresponding to the customer ID of the customer A is extracted.

  In order to use the market research / analysis system of the present invention, an alliance company is provided with a browser for issuing a login ID and password accessible to the operation company system 2 and displaying the market research / analysis results. From the headquarters terminal 7 of the company participating in the system, the server of the operating company system 2 is accessed with the login ID and the password, and the address and destination data can be selected and input, and the above-described survey / analysis system The market research / analysis results (customer data table, waveform graph, synchronization rate list, etc.) obtained in 1 may be browsed.

  As described above, according to the market research / analysis system 1, it is possible to conduct market research and analysis with high accuracy, and to provide alliance companies as users of the market research / analysis system in an easy-to-understand format. It is possible to provide a system that can

1 Market Research / Analysis System 2 Server 3 Store Terminal 4 Communication Network 5,6 Customer Terminal 7 Headquarters Terminal

Claims (15)

  Waveform generating means for waveformizing customer data accumulated in the database, means for calculating a difference between at least two waveform points generated by the waveform generating means, and at least generated by the waveform generating means Means for calculating a difference between line segment angles of two or more waveforms, means for determining whether or not the difference between the points falls within a threshold value, and whether or not the difference between the line segment angles falls within a threshold value. A means for determining, a means for calculating a ratio at which the difference between the points falls within a threshold, a means for calculating a ratio at which the difference between the line segment angles falls within the threshold, and a difference between the points within the threshold A market research / analysis system comprising means for calculating a degree of approximation of at least two or more waveforms from a ratio at which a difference between a ratio and a line segment angle falls within a threshold.   The market research / analysis system according to claim 1, wherein the waveform generation means generates a line graph.   The market research / analysis system according to claim 1, wherein the waveform generation means includes means for calculating averaged customer data.   A means for selecting one or more destination data and a means for selecting one or more destination data are provided. The waveform generation means generates a waveform of the source data and a waveform of the destination data, and calculates the degree of approximation of the waveform. 4. The market research / analysis system according to claim 1, wherein the means for calculating calculates a degree of approximation of the waveform of the destination data with respect to the destination data.   5. The market research and analysis system according to claim 1, further comprising a table creating means for displaying a list of approximations of the destination data and the destination data.   6. The market research / destination data according to claim 1, wherein the destination data and the destination data can be selected from at least one of the categories of customers, companies, stores, products, services, and areas. Analysis system.   7. The market research / analysis system according to claim 1, further comprising prediction processing means for calculating predicted data based on actual data accumulated in a database.   The market research / analysis system according to claim 1, wherein the actual data is discretized, and predicted data is calculated by a prediction processing unit based on the discretized actual data.   9. The market research / analysis system according to claim 1, wherein the forecast data calculated by the forecast processing means is customer data.   9. The market research / analysis system according to claim 1, wherein the customer data is created by complementing the missing items of the actual data with the forecast data calculated by the forecast processing means.   The market research / analysis system according to claim 7, wherein the forecast data includes a probability value or a forecast value obtained by threshold determination of the probability value.   The market research / analysis system according to any one of claims 1 to 11, further comprising table creation means for displaying a list of customer data for customer IDs.   13. A recommendation means for associating a recommendation medium of a company, a store, a product / service or an area for which an arbitrary degree of approximation is calculated with respect to an arbitrary customer ID. Market research and analysis system.   A recommendation means for associating a customer ID having an arbitrary degree of approximation with a recommendation medium corresponding to an arbitrary company ID, store ID, product / service ID, or area ID. The market research and analysis system according to any one of 13 above.   The market research / analysis system according to claim 1, wherein the headquarters terminal includes destination data and destination data selection means and market research / analysis result browsing means.