JP7365152B2 - Content matching system and content matching method - Google Patents

Description

The present invention relates to a content matching system and a content matching method, and more particularly, to a content matching system and a content matching method suitable for displaying content of interest to consumers and promoting sales in a booth displaying products and services. Regarding matching methods.

It has become common practice to set up large displays at event booths and street advertisements to inquire about products and services to event participants and passersby. At this time, event exhibitors and advertisement exhibitors are concerned with how to attract customers' attention and efficiently provide information and advertisements.

For example, Patent Document 1 discloses an advertising system that analyzes images of potential customers, determines their interest level, and controls advertisement content based on the interest level. Additionally, Patent Document 2 discloses a display device that analyzes clothing worn by a captured image and displays product information and advertising information to a product purchaser of a vending machine in accordance with the analysis. .

Japanese Patent Application Publication No. 2013-50945 Japanese Patent Application Publication No. 2018-84775

The above-mentioned prior art Patent Document 1 discloses a technique for determining a customer's degree of interest based on the customer's line of sight and body posture direction. Further, Patent Document 2 discloses a technique for providing product information and advertising information according to customer attributes of captured images.

However, it cannot be said to be a sufficient indicator of the interest of people present at an event booth.

Event booths and street advertising can use pausing as an important indicator of interest. Furthermore, the level of interest of people who enter the booth is even higher, and it is necessary to select the content to be displayed with such customers in mind. Furthermore, as customer attributes, categorization in terms of gender and age group is also an important factor in content display. In addition, for event booths, factors such as exhibition location, advertising location, and timing are also factors to consider when providing information and advertising. The above-mentioned conventional technology does not show how to take this viewpoint into consideration and display appropriate content to potential customers.

An object of the present invention is to provide a content matching system and content matching suitable for promoting sales by displaying content of interest to consumers according to customer attributes, location, and time in a booth exhibiting products and services. The purpose is to provide a method.

The configuration of the content matching system of the present invention is preferably a content matching system that matches content content, content provision location, and provision date and time,
A means for storing location information of multiple booths connected to a network and their attributes in association with each other as a booth installation location, a sensing means for capturing an image of surrounding information of the booth, and a degree of customer attraction of the booth based on the sensing results. , a means for analyzing the degree of attention from passersby for each booth, and selecting recommended booth candidates for the desired content provision date and time received from the user based on the analysis results and the attributes. This is what I did.

According to the present invention, a content matching system suitable for promoting sales by displaying content of interest to consumers according to customer attributes, location, and time in booths displaying products and services and in street advertisements. and a content matching method.

1 is a diagram showing a system configuration of a content matching system. It is a functional block diagram of a management server. It is a figure showing an example of an equipment data table. FIG. 3 is a diagram showing an example of a location data table. It is a figure showing an example of an application/use situation data table (part 1). It is a figure which shows an example of an application/usage status data table (part 2). It is a figure showing an example of a passerby attribute table. It is a figure showing an example of a matching data table. 3 is a flowchart showing the overall processing of the content matching system. It is a flowchart which shows sensing with respect to a passerby and its attribute analysis process. It is a flowchart which shows a customer attraction effect analysis process. It is a flowchart which shows matching analysis processing (pull type). It is a flowchart which shows matching analysis processing (push type). It is a functional block diagram of a management server. FIG. 3 is a functional block diagram of a control terminal.

Hereinafter, each embodiment of the present invention will be described using FIGS. 1 to 13.

[Embodiment 1]
Embodiment 1 of the present invention will be described below with reference to FIGS. 1 to 11B.
First, the configuration of the content matching system according to the first embodiment will be described using FIGS. 1 and 2.

First, the system configuration of the content matching system according to the first embodiment will be described using FIG. 1.
The content matching system has a configuration in which a management server 100 and a control terminal 200 installed in each booth 50 are connected via a network 5, as shown in FIG. The network 5 may be a LAN (Local Area Network) or a global network such as the Internet. The booth 50 is an area where information about products and services set up at an exhibition event or the like is provided to appeal to customers. It is assumed that the booth 50 of this embodiment has a system configuration in which an exhibitor performs user authentication and can be used by authenticated persons.

The management server 100 is a server that analyzes information collected by the control terminals 200 and distributes matching content to the control terminals 200 of each booth. The control terminal 200 is an information processing device that collects information on people inside and outside the booth, uploads it to the management server 100, and displays content distributed from the management server 100 on a display device such as a display.

The management server 100 can be realized by a general server device having the hardware configuration shown in FIG.

The management server 100 has a processor (Central Processing Unit) 101, a main memory 102, a communication I/F 103, an input/output I/F 104, and an HDD 110 connected by a bus.

The processor 101 controls each part of the management server 100, loads necessary programs into the main memory 102, and executes them.

The main memory 102 is usually configured with a volatile memory such as a RAM, and stores programs executed by the processor 101 and data to be referenced.

Communication I/F 103 is an interface for connecting to network 5.

The input/output I/F 104 is an interface for connecting the input/output device 120 and the display device 130. As the input/output device 120, for example, a keyboard, a pointing device mouse, a printer, etc. are connected.

The HDD 110 has a large storage capacity, and stores programs and necessary data for executing the present embodiment.

The control terminal 200 can be realized by a general information processing device having the hardware configuration shown in FIG.

The control terminal 200 has a configuration in which a processor 201, a main memory 202, a communication I/F 203, an input/output I/F 204, and an HDD 210 are connected via a bus. The functions of each part are similar to the components of the corresponding server device.

A user authentication device 220 , an internal sensing device 230 , an external sensing device 240 , and a content display device 250 are connected to the input/output I/F 204 of the control terminal 200 .

The user authentication device 220 is a device that authenticates exhibitors to use this booth, and may be a biometric authentication device such as fingerprint or face authentication, or a device for password authentication. Good too. The internal sensing device 230 is a device that collects information on people inside the booth 50, and is, for example, a video camera installed inside the booth 50. Conversely, the external sensing device 240 is a device that collects information on people outside the booth 50, and is, for example, a video camera installed outside the booth 50. The content display device 250 is, for example, a large-screen display device that displays content related to products and services.

Next, the functional configuration of the management server and the data it holds will be explained using FIG.
As shown in FIG. 2, the management server 100 has a functional configuration including a reservation management section 301, an entrance/exit management section 302, a customer attraction effect analysis section 303, a matching analysis section 304, an image recognition/attribute analysis section 305, and a user authentication section. 306, a sensing device I/F section 307, and a content display device I/F section 308.

The reservation management unit 301 is a functional unit that manages reservations for users of each booth and appropriately allocates reservations to users. The entry/exit management unit 302 is a functional unit that appropriately manages entry/exit of the booth 50 so that authenticated users are allowed to enter/exit and unauthenticated users are not allowed to enter/exit. The customer attraction effect analysis unit 303 is a functional unit that executes a customer attraction effect analysis process that analyzes the customer attraction effect of content. The matching analysis unit 304 is a functional unit that performs matching analysis processing to select appropriate content by matching determined from the passerby attributes of the content. The image recognition/attribute analysis unit 305 has a function of performing image recognition on images taken by passersby (people inside and outside the booth 50) and analyzing their attributes (such as gender, age, and occupation estimated based on clothing). Department. The user authentication unit 306 is a functional unit that performs user authentication for users who use content. The sensing device I/F unit 307 is a functional unit that exchanges information with the internal sensing device 230 and the external sensing device 240. The content display device I/F unit 308 is a functional unit that exchanges information with the content display device 250.

The functions of each part described above can be realized by installing programs to execute them and having the processor 101 execute those programs loaded into the main memory 102.

The details of the process of sensing passersby and performing image recognition/attribute analysis, customer attraction effect analysis process, and matching analysis process will be explained later.

Furthermore, the management server 100 holds an equipment data table 410, a location data table 420, an application/usage status data table 430, a passerby attribute table 440, and a matching data table 450 as data. Note that details of each table will be explained later.

Next, the data structure used in the content matching system will be explained using FIGS. 3 to 7.

The equipment data table 410 is a table that holds information regarding booth equipment, and as shown in FIG. 3, it consists of fields such as a booth ID 410a and an installation location 410b. The booth ID 410a stores an identifier that uniquely identifies the booth. The installation location 410b describes the installation location of the booth, such as a station or place name.

The location data table 420 is a table that describes the characteristics of the location where the booth is installed, and as shown in FIG. 4, it consists of fields such as an installation location 420a and a location attribute 420b. The installation location 420a stores the name of the location to be described. The location attribute 420b stores information about attributes of the installation location 420a, such as business, sightseeing, transit, event, and academic city.

The application/usage status data table 430 is a table that stores information regarding applications made by users to use booths, information on recommended booths, and past usage results, and as shown in FIGS. 5A and 5B, Business operator 430a, purpose 430b, content keyword 430c, content content 430d, target customer group 430e, desired start date and time 430f, desired end date and time 430g, desired booth 430h, recommended start date and time 430i, recommended end date and time 430j, recommended booth 430k, reservation ID 430l , reservation start date and time 430m, reservation end date and time 430n, reservation booth 430o, performance start date and time 430p, performance end date and time 430q, and performance booth 430r.

The business operator 430a, purpose 430b, content keyword 430c, content content 430d, and target customer group 430e are user attributes input when the user applies for a booth. The desired start date and time 430f, the desired end date and time 430g, and the desired booth 430h are the desired conditions for the user to use the booth. The recommended start date and time 430i, the recommended end date and time 430j, and the recommended booth 430k are conditions for the booth recommended by the matching process. The reservation ID 430l, the reservation start date and time 430m, the reservation end date and time 430n, and the reservation booth 430o are the conditions under which the user made a reservation to use the booth. The performance start date and time 430p, the performance end date and time 430q, and the performance booth 430r are performance information on the user's use of the booth.

A name or an identifier for identifying a user is stored in the business operator 430a. The usage 430b stores the usage of the user's booth, such as "retail", "exhibition", and "advertisement". The content keyword 430c stores keywords that the user should pay attention to in the content displayed at the booth. The content content 430d stores a brief description of the content displayed by the user at the booth. The target customer group 430e stores information on the classification targeted by the user exhibiting the content as the main customer group. The desired start date and time 430f and the desired end date and time 430g respectively store the start date and time and end date and time of the booth that the user desires to use. The desired booth 430h stores the contents of the booth desired by the user. For example, "within XX range" specifies that it is an "exhibition booth within 5 minutes of a Yamanote Line station." The recommended start date and time 430i and the recommended end date and time 430j respectively store the start date and time and end date and time of the booth recommended by the matching process. The recommended booth 430k stores a unique identifier of a booth recommended through matching processing. A unique identifier indicating that a reservation has been made is stored in the reservation ID 430l. The reservation start date and time 430m and the reservation end date and time 430n respectively store the start date and time and end date and time of the reserved booth. A unique identifier representing the reserved booth is stored in the reserved booth 430o. The performance start date and time 430p and the performance end date and time 430q respectively store the start date and time and end date and time as the performance of the booth used. The performance booth 430r stores a unique identifier representing a booth that has been used successfully.

The passerby attribute data 440 is a table that stores information on the performance of the booth operation, and as shown in FIG. 6, the booth ID 440a, start date and time 440b, end date and time 440c, content presence/absence 440d, and number of passersby outside the booth 440e. , the attributes of passersby outside the booth 440f, the number of people watching the booth 440g, the attributes of people watching the booth 440h, the number of stopped passersby 440i, the stopped passersby attribute 440j, the number of people inside the booth 440k, and the attributes of people inside the booth 440l.

A unique identifier for identifying a booth is stored in the booth ID 440a. The start date and time 440b stores the start date and time of the booth operation. The end date and time 440c stores the end date and time of the booth operation. The content presence/absence field 440d stores a flag indicating whether or not content is displayed in the booth. For example, when the content display device is displaying content that the exhibitor wants to present, it will be "present," and when no content is being displayed, it will be "absent." The number of passersby outside the booth measured by the external sensing device 240 is stored in the number of passersby outside the booth 440e. The attribute of the passerby outside the booth is stored in the attribute 440f of the passerby outside the booth as a result of the image being captured and image recognized by the external sensing device 240. The number of people gazing at the booth 440g stores the number of passersby who are imaged by the external sensing device 240 and analyzed through image recognition, line of sight analysis or posture analysis as gazing at the target booth. The booth gazing person attribute 440h stores attributes as a result of image recognition for the person gazing at the booth. The number of stopped passersby 440i stores the number of passersby who are imaged by the external sensing device 240 and determined to have stopped by image recognition. The stopped passerby attribute 440j stores an attribute as a result of image recognition for a passerby who is determined to have stopped. The number of people in the booth 440k stores the number of people entering the booth, which is imaged by the internal sensing device 230. The in-booth person attribute 440l stores attributes as a result of image recognition of the person who entered the booth.

In general, it can be estimated that the interest in the content displayed at the booth increases in the order of mere passers-by outside the booth ≦ people watching the booth ≦ people who stopped ≦ people who entered the booth.

The matching data table 450 is a table for storing information for evaluating the actual customer attraction effect of a booth, and is a table for extracting booths that are recommended candidates by searching this table using various keys and aggregating them. be. As shown in FIG. 7, the matching data table 450 includes a booth ID 450a, purpose 450b, content keyword 450c, target customer group 450d, installation location 450e, aggregation date and time 450f, attribute ratio of passersby without content 450g, and number of customers. It consists of fields such as 450h, attention level 450i, attention person attribute ratio 450j, high interest level 450k, high interest attribute ratio 450l, nearby booth held content keyword 450m, matching degree 450n, and nearby booth related index 450o.

A unique identifier for identifying a booth is stored in the booth ID 450a. The purpose 450b stores the purpose of the user's booth, such as "retail", "exhibition", and "advertisement". The content keyword 450c stores keywords of interest for content exhibited by the user at the booth. The target customer group 450d stores information on the classification that the user who exhibits the content targets as the main customer group. The purpose 450b, content keyword 450c, and target customer group 450d are information corresponding to the purpose 430b, content keyword 430c, content content 430d, and target customer layer 430e, respectively, of the application/usage status data table 430 shown in FIG. 5A. The installation location 450e stores the installation location of the booth. The aggregation date and time 450f stores the date and time when information regarding the booth usage record was collected. The attribute ratio of passersby when no content is present 450g stores the result of attribute analysis for passersby when no content is displayed (when the content presence/absence 440d is "absent"). The number of customers attracted 450h stores the difference in the number of passersby per unit time when the content is being displayed and when the content is not being displayed. This makes it possible to estimate the effect of displaying the content. The attention level 450i stores the percentage of people who stopped (attention people) while displaying the content. The degree of attention 450i may be defined as the rate of the person who gazed at the content while displaying the content. The attention person attribute ratio 450j stores attribute ratio information of people who stopped or gazed while displaying the content. The high interest level 450k stores the percentage of people who actually entered the booth (high interest people) among the people who stopped or gazed. The high interest attribute ratio 450l stores attribute ratio information of people who entered the booth while the content was being displayed. The held content keywords of nearby booths 450m store content keywords defined for booths held in the vicinity of the applicable booth (for example, on both sides and two neighbors of the applicable booth). This is because passersby may be interested in the contents of not only the booth, but also booths held near the booth, and may visit several booths. The matching degree 450n stores the percentage of people who have the attributes specified by the user as the target customer group among the people of interest. For example, if there are 100 people of interest, 50 are women in their 30s, 39 are women in their 40s, and the user specifies the target customer group 430e as women in their 30s to 40s, matching The degree of coincidence is 89%. Furthermore, the degree of matching may be defined as the percentage of people with attributes specified by the user as a target customer group among people with high interest. The neighboring booth related index 450o is an index representing the degree of similarity between the content keywords of the relevant booth and its neighboring booths. Although not shown, for example, you can keep a score table for each keyword (such as "cosmetics - cosmetics: 0.8", "cosmetics - lotion: 0.6", etc.), and when there is a person who matches the score, It may be determined by adding up the scores and calculating the sum.

Next, the processing of the content matching system will be explained using FIGS. 8 to 12.

First, the overall processing of the content matching system will be explained using FIG. 8.
The processing up to the first half of the overall processing will be explained separately into pull type and push type. Here, a pull type is one in which the user who uses the booth acts on the system to cause the user to use the booth, and a push type, on the other hand, is one in which the user who uses the booth acts on the system to increase the possibility of using the booth. The system selects a certain user and presents recommended booths.

In the case of the pull type, the user registers registration information such as the content keyword of the content to be exhibited, desired date and time, content content, and target customer group in the application/usage status data table 430 (S100a).

Next, the content matching system searches the matching data table 450 for the combination of content keywords that the user wants to exhibit, and outputs the combination of location, date and time that is effective in attracting customers (matching analysis) (S101a). The meaning of "has the effect of attracting customers" and the details of the pull-type matching analysis process will be explained later.

Next, the content matching system refers to the application/usage status data table 430 for the output date/time combinations, and outputs the combination of place/date/time that is reservable and has the most effect on attracting customers (S102a).

On the other hand, in the case of the push type, the content matching system refers to the application/usage status data table 430, checks the usage status of the booth, and extracts candidates for available locations and dates and times (S100b).

Next, the matching data table 450 is searched using vacant locations and dates and times as keys, and the results are totaled to output content keyword combinations that are effective in attracting customers at the corresponding locations and time slots (S101b). The meaning of "has the effect of attracting customers" and the details of the push-type matching analysis process will be explained later.

Next, the system searches for a user who has the combination of content keywords that is most effective in attracting customers, and recommends and displays a booth for that user at the location and time (S102b).

Next, the user refers to the location, date and time presented by the system and makes a reservation for a booth at that location, date and time (S103). The contents of the reservation are reflected in the application/usage status data table 430.

Next, the user goes to the reserved booth on the reserved date and time and performs user authentication using the user authentication device 220 (S104).

Next, the user enters and exits the reserved booth and starts using it (S105).

Next, the allocated content such as advertisement is displayed on the content display device 250 of the booth (S106).

Next, when the use is completed, the user leaves the room, and the user's usage record at the booth is registered in the application/use status data table 430 (S107).

Next, the matching data table 450 is updated using the usage history of the application/usage status data table 430 and the information of the passerby attribute data table 440 (customer attraction effect analysis) (S108). The details of the customer attraction effect analysis process will be described later.

Next, sensing of passersby and their attribute analysis processing will be explained using FIG.
First, the external sensing device 240 and the internal sensing device 230 installed inside and outside the booth 50 sense the outside and inside of the booth 50 and accumulate information about passersby (S200).

It is determined whether the fixed interval timing has come (S201), and when the fixed interval timing has arrived (S201: YES), the process goes to S202, and when the fixed interval timing has not come (S201: NO), S205 go to.

If the fixed interval timing has not arrived, it is determined whether or not it is outside the business hours of the booth exhibition (S205), and if it is outside the business hours of the booth exhibition (S205: YES), the process is ended and the booth is closed. If it is within the business hours of the exhibition (S205: NO), the process returns to S200.

When the timing of the fixed interval comes, based on the sensing information of the fixed interval, the included passersby are detected by image recognition and attribute analysis processing, and attributes such as gender, clothing, age group, etc. are determined (S202 ).

Next, especially from the sensing information outside the booth, when content is being displayed in the booth, passersby who gazed at the booth and passersby who stopped are extracted (S203).

Next, the number of people is counted for each attribute of the sensing information inside and outside the booth and recorded in the passerby attribute data table 440 (S204).

Next, the customer attraction effect analysis process will be explained using FIG.
The customer attraction effect analysis process is performed after the exhibition using the user's booth is completed and the usage record information (actual start date and time 430p, performance end date and time 430q, and performance booth 430r) shown in FIG. 5A is registered.

First, the percentage of passerby attributes with no content is calculated from the records of the passerby attribute data table 440 for the corresponding location (S300).

Records from the passerby attribute data table 440 that are included in the corresponding location and time period are extracted (S301).

Next, the number of customers attracted (the difference in the number of passersby outside the booth with content and without content per unit time) is calculated (S302).

Next, the degree of attention (percentage of people who stopped or looked at the content while displaying the content (attention people)) and the ratio of attention person attributes (attention person attribute distribution) are calculated (S303 ).

Next, calculate the degree of high interest (the percentage of people who actually entered the booth (highly interested people) among the people who paid attention) and the percentage of high interest attributes (the attribute distribution of people with high interest). S304).

Next, we search for records of booths exhibiting content with content in a certain nearby range (for example, on both sides of the relevant booth) during the same time period, check the content keywords, and find the neighborhood that shows the similarity of the content keywords. A booth related index is calculated (S305).

Next, a matching degree indicating the degree of correspondence between the target customer group desired by the user and the attributes of the highly interested person is calculated (S306).

Finally, each value calculated in S302 to S306 is stored in each field of the matching data table 450 (S307).

Next, matching analysis processing (pull type) will be explained using FIG. 11A.
This process corresponds to S101a in FIG. 8.
First, the content matching system searches for records in the matching data table 450 using the OR of combinations of content keywords 430c that the user wants to exhibit (S400).

Next, from the search results, records whose matching degree 450n is smaller than a predetermined threshold are excluded from analysis (S401).

Next, for the narrowed down records of the matching data table 450, the number of customers attracted 450h x degree of attention 450i = number of people paying attention per unit time is calculated, and the records are sorted in descending order of the number of people paying attention per unit time (S402 ).

Next, among the records sorted in S402, the records are sorted in descending order of the value of the portion matching the target user group in the attention person attribute ratio 450j, and records smaller than a predetermined threshold are excluded from analysis. (S403).

Next, among the records sorted in S403, the data is sorted in descending order of high interest level 450k, and records smaller than a predetermined threshold are excluded from analysis (S404).

Next, among the records sorted in S404, the data is sorted in descending order of the value of the part that matches the target user group with the high interest person attribute ratio 450l, and records smaller than a predetermined threshold are excluded from the analysis. (S405).

Next, several booth candidates from the top (for example, from the top to the third, this value may be defined by the user) are extracted from the sorted results (S406). At this time, booth candidates may be extracted by taking into consideration the keyword related index 450o of nearby booths. For example, among the records sorted in the process up to S405, for records with a small difference in high interest level 450k and high interest person attribute ratio 450l, the keyword related index 450o is greater than a predetermined value in the lower record. Processing such as sorting the records to a higher level may also be performed. Alternatively, the frequency distribution of past performance booths 450r may be calculated for the determined top candidate booths, and the candidates may be extracted in descending order of appearance frequency.

Next, from among the candidates obtained in the processing up to S406, booths that match the user's desired time slot (desired start date and time 430f, desired start date and time 430g) are presented to the user (S407). Here, the same time zone can include viewpoints such as the user's desired time zone and the dates and times of past exhibition booths that match or are similar in terms of season, month, day, or time. . That is, even if the user's desired time zone does not match, the user's desired time zone and the extracted candidate date and time are compared, and if the time zones are close to some extent (for example, if the user's desired time zone is 13 :00 to 15:00 and the date and time of the extracted candidate is 13:30 to 15:30), that fact and the date and time may be presented to the user.

In the above matching analysis processing (pull type) processing, there are a certain number of people who are paying attention (people who gazed, people who stopped) and people who are highly interested (people who entered the booth), and among them, the target customer that the user desires A booth that matches the attributes to some extent is selected. Therefore, the meaning of "has the effect of attracting customers" mentioned in S101a is defined as the existence of a certain number of such people who are attracting attention and people with high interest, and among them, the attributes of the target customers desired by the user match to some extent. be done.

Next, matching analysis processing (push type) will be explained using FIG. 11B.
This process corresponds to S101b in FIG. 8.
The content matching system refers to the application/usage status data table 430 to find the vacant booth and date and time, and searches the matching data table 450 using the vacant booth and date and time as a key (S500). Regarding date and time, the search target is not an exact match, but also the vicinity (for example, 30 minutes before and after, 1 hour before and after).Next, from the search results of S500, records whose matching degree is smaller than a predetermined threshold are analyzed. (S501).

Next, for the records of the matching data table 450 narrowed down in S501, the number of customers attracted 450h x degree of attention 450i = number of people paying attention per unit time is calculated, and the records are sorted in descending order of the number of people paying attention per unit time. , records smaller than a predetermined threshold are excluded from analysis (S502).

Next, among the records sorted in S502, the records are sorted in descending order of high interest level 450k, and records smaller than a predetermined threshold are excluded from analysis (S503).

Next, from the results of sorting in S503, the number of occurrences of the content keywords determined as top candidates is counted, and a keyword set is created and outputted from a combination of keywords with a large number of occurrences (S504).

In the matching analysis processing (push type) described above, content keywords of booths with a certain number of people who are paying attention (people who gazed, people who stopped) and people with high interest (people who entered the booth) are selected. Therefore, the meaning of "has the effect of attracting customers" mentioned in S101b is defined as the presence of a certain number of such people who are paying attention or who are highly interested.

According to the content matching system of this embodiment, based on the performance of booths exhibited in the past, the target customer group of the user matches, and the content displayed at the booths to be exhibited is similar and attracts attention. We recommend that users use booths with high interest (many people were staring or stopping) and booths with high interest (many people entering the booth), so users who display products and services at the booth It is possible to efficiently display content of interest to consumers and promote sales.

[Embodiment 2]
The content matching system of the second embodiment will be described below with reference to FIGS. 12 and 13.

In this embodiment, compared to Embodiment 1, only the differences will be mainly explained.
The hardware system configuration of the content matching system of this embodiment is the same as that of the first embodiment. The table and the contents of each process are also the same as in the first embodiment. In this embodiment, only the functional configurations of the management server 100 and the control terminal 200 are different.

As shown in FIG. 12, the functional configuration of the management server 100 of this embodiment includes a reservation management section 301, an entrance/exit management section 302, a customer attraction effect analysis section 303, and a matching analysis section 304.

That is, compared to the first embodiment, the image recognition/attribute analysis section 305, user authentication section 306, sensing device I/F section 307, and content display device I/F section 308 are omitted.

As in the first embodiment, the management server 100 holds an equipment data table 410, a location data table 420, an application/usage status data table 430, a passerby attribute table 440, and a matching data table 450 as data.

On the other hand, the control terminal 200 of this embodiment has a functional configuration as shown in FIG. 208.

That is, the image recognition/attribute analysis section 205, user authentication section 206, sensing device I/F section 207, and content display device I/F section 208 perform image recognition/attribute analysis of the management server 100 shown in FIG. 2 of the first embodiment. The functions of the analysis section 305, user authentication section 306, sensing device I/F section 307, and content display device I/F section 308 are executed by the control terminal 200.

As a result, data is processed on the control terminal 200 side of this embodiment and the results are stored in each table, so the amount of data communication between the management server 100 and the control terminal 200 is reduced, reducing the load on the management server 100. can do.

100... Management server, 120... Input/output device, 130... Display device,
200... Control terminal, 220... User authentication device, 230... Internal sensing device, 240... External sensing device, 250... Content display device,
301... Reservation management department, 302... Entrance/exit management department, 303... Customer attraction effect analysis department, 304... Matching analysis department, 305... Image recognition/attribute analysis section, 306... User authentication section, 307... Sensing device I/F section, 308...Content display device I/F section,
410...Equipment data table, 420...Location data table, 430...Application/usage status data table, 440...Passerby attribute table, 450...Matching data table

Claims (10)

A content matching system that matches content content, content provision location, and provision date and time,
means for associating location information and attributes of multiple booths connected to a network and storing them as booth installation locations;
sensing means for capturing peripheral information of the booth;
means for analyzing the degree of customer attraction of the booth and the degree of attention from passersby for each booth based on the sensing results;
Based on the analysis results and the attributes, select booth candidates to recommend for the desired content provision date and time received from the user ;
It has a management server and a control terminal connected to the management server,
The management server is
a first table that stores exhibition conditions desired by the user;
A second table that stores passerby information as the booth's exhibition results;
maintains a third table that associates and stores the results of analyzing each booth exhibition information and passerby information, including the content keywords of the content exhibited at the booth and the exhibition date and time of the booth;
The control terminal has a sensing device that images passersby inside and outside the booth,
The management server receives, from the user, a desired date and time for displaying the booth, content keywords of content to be displayed at the booth, and target customer attributes targeted by the user, and stores them in the first table;
The control terminal transmits an image of a passerby captured by the sensing device,
The management server performs an image analysis of the image of the passerby, performs attribute analysis of the passerby, and stores attribute analysis information of the passerby at each booth in the second table;
The management server refers to the second table,
The difference in the number of passersby outside the booth per unit time when content is displayed on the booth and when no content is displayed on the booth is calculated as the number of customers attracted,
When content is displayed on a booth, the person who gazed at the displayed booth among passersby or the person who stopped outside the booth is considered the person who is paying attention, and the person who is paying attention to the person who is paying attention to the person passing by as a whole. Calculate the percentage as the degree of attention,
Calculating the attribute ratio of the passerby of the noted person as the attention person attribute ratio,
storing the number of customers of each booth, the degree of attention, and the attention attribute ratio in the third table;
The management server inputs the desired date and time of exhibiting the booth input by the user, the content keyword of the content to be exhibited at the booth, the target customer attributes targeted by the user, and the information of each booth stored in the third table. A content matching system that selects recommended booth candidates based on content keywords, exhibition dates and times of booths, the calculated number of customers for each booth, the degree of attention, and the attention attribute ratio. . Furthermore, when the content is displayed in the booth, the management server determines that the person who entered the booth is a high interest person, and calculates the ratio of the high interest person to the attention person as a high interest level,
Calculating the ratio of the attributes of the high interest person as a high interest person attribute ratio,
storing the high interest level and the high interest person attribute ratio of each booth in the third table;
2. The content matching system according to claim 1 , wherein the management server further selects booth candidates to recommend based on the high interest level and the high interest person attribute ratio. The management server calculates, as a matching degree, the proportion of people with attributes specified by the user as a target customer group among people of interest or people with high interest for each booth, and stores it in the third table;
3. The content matching system according to claim 2 , wherein the management server selects recommended booth candidates by excluding booths for which the matching degree is smaller than a predetermined threshold. The third table includes fields for content keywords of booths held in the vicinity of each booth and a nearby booth related index that compares the degree of similarity between the content keywords of each booth and the content keywords of booths held in the vicinity. ,
2. The content matching system according to claim 1 , wherein recommended booth candidates are selected by also referring to the neighboring booth related index. A content matching system that recommends the most suitable booth to display content,
It has a management server and a control terminal connected to the management server,
The management server is
a first table that stores exhibition conditions desired by the user;
A second table that stores passerby information as the booth's exhibition results;
maintains a third table that associates and stores the results of analyzing each booth exhibition information and passerby information, including the content keywords of the content exhibited at the booth and the exhibition date and time of the booth;
The control terminal has a sensing device that images passersby inside and outside the booth,
The management server receives content keywords of content to be exhibited at the booth and stores them in the first table,
The control terminal transmits an image of a passerby captured by the sensing device,
The management server performs an image analysis of the image of the passerby, performs attribute analysis of the passerby, and stores attribute analysis information of the passerby at each booth in the second table;
The management server refers to the second table,
The difference in the number of passersby outside the booth per unit time when content is displayed on the booth and when no content is displayed on the booth is calculated as the number of customers attracted,
When content is displayed on a booth, the person who gazed at the displayed booth among passersby or the person who stopped outside the booth is considered the person who is paying attention, and the person who is paying attention to the person who is paying attention to the person passing by as a whole. Calculate the percentage as the degree of attention,
storing the number of customers and the degree of attention of each booth in the third table;
The management server refers to the first table and searches for vacant booths,
Selecting candidates from the vacant booths based on the number of customers and the degree of attention regarding the vacant booths;
Outputting a keyword set of combinations in which the content keywords of the candidate booth have a large number of appearances;
The management server selects recommended booth candidates by referring to the content keyword of the content to be exhibited at the booth, the output keyword set, and the exhibition date and time of the candidate booth,
The management server is characterized in that the management server displays the recommended booth candidates to the target user based on the output keyword set and the content keyword of the content to be exhibited at the booth input by the user. Content matching system. Furthermore, when the content is displayed in the booth, the management server determines that the person who entered the booth is a high interest person, and calculates the ratio of the high interest person to the attention person as a high interest level,
storing the high interest level of each booth in the third table;
Furthermore, based on the high interest level, candidates are selected from the vacant booths,
6. The content matching system according to claim 5, wherein the content matching system outputs a keyword set of combinations in which content keywords of the candidate booth have a large number of appearances. A content matching method that matches content content, content provision location, and provision date and time using a content matching system,
The content matching system is
It has a management server and a control terminal connected to the management server,
The management server is
a first table that stores exhibition conditions desired by the user;
A second table that stores passerby information as the booth's exhibition results;
maintains a third table that associates and stores the results of analyzing each booth exhibition information and passerby information, including the content keywords of the content exhibited at the booth and the exhibition date and time of the booth;
The control terminal has a sensing device that images passersby inside and outside the booth,
the management server receiving from the user a desired date and time for displaying the booth, content keywords of content to be displayed at the booth, and target customer attributes targeted by the user, and storing the received information in the first table;
a step in which the control terminal transmits an image of a passerby captured by the sensing device to the management server;
a step in which the management server analyzes the image of the passerby transmitted from the control terminal , performs attribute analysis of the passerby, and stores attribute analysis information of the passerby at each booth in the second table; and,
The management server ,
Referring to the second table,
The difference in the number of passersby outside the booth per unit time when content is displayed on the booth and when no content is displayed on the booth is calculated as the number of customers attracted,
When content is displayed on a booth, the person who gazed at the displayed booth among passersby or the person who stopped outside the booth is considered the person who is paying attention, and the person who is paying attention to the person who is paying attention to the person passing by as a whole. Calculate the percentage as the degree of attention,
a step of calculating a ratio of attributes of passersby of the person of interest as a ratio of attributes of the person of interest;
the management server storing the number of customers, the degree of attention, and the attention person attribute ratio of each booth in the third table;
The management server inputs the desired date and time of exhibiting the booth entered by the user, the content keyword of the content to be exhibited at the booth, the target customer attributes targeted by the user, and the information of each booth stored in the third table. The method further comprises the step of selecting recommended booth candidates based on the content keyword, the exhibition date and time of the booth, the calculated number of customers for each booth, the degree of attention, and the attention attribute ratio. content matching method. Furthermore, when the content is displayed in the booth, the management server determines a person who enters the booth as a high interest person, and calculates the ratio of the high interest person to the attention person as a high interest level,
the management server storing the high interest level and the high interest person attribute ratio in a third table;
8. The content matching method according to claim 7, wherein the management server further comprises the step of selecting booth candidates to recommend based on the high interest level and the high interest person attribute ratio. A step in which the management server calculates, as a matching degree, the proportion of people who have the attributes specified by the user as the target customer group among the people who are attracting attention or people who are highly interested in each booth, and storing it in the third table. and,
9. The content matching method according to claim 8 , wherein the management server selects recommended booth candidates by excluding booths for which the degree of matching is smaller than a predetermined threshold. The third table includes fields for content keywords of booths held in the vicinity of each booth and a nearby booth related index that compares the degree of similarity between the content keywords of each booth and the content keywords of booths held in the vicinity. ,
8. The content matching method according to claim 7 , wherein the management server selects booth candidates to recommend by also referring to the nearby booth related index.

Images