KR20130040004A - Method for determining commendatory booth based on similarity of visiting booth and favorite order - Google Patents

Abstract

PURPOSE: A method for determining a visit booth and a recommended booth by considering the similarity of visiting preference is provided to accurately determine a recommended booth of a user by including a case visiting the same booth at different unit visiting times in the similarity calculation of the visit booth. CONSTITUTION: A recommended booth determining server searches for and extracts sequential visit booth information of surrounding users(S100,S200). The server calculates the separation unit visit time similarity of a separation unit visit time and the same unit visit time similarity of a same unit visit time and calculates entire similarity by booth after arranging the sequential visit booth information(S300-S600). The server determines a recommended booth based on the calculation result(S700). [Reference numerals] (AA) Start; (BB) End; (S100) Searching the sequential visit booth information of surrounding users; (S200) Extracting the sequential visit booth information of surrounding users; (S300) Arranging sequential visit booth information; (S400) Calculating same unit visit time similarity; (S500) Calculating separation unit visit time similarity; (S600) Calculating entire similarity by booth; (S700) Determining a recommended booth;

Description

Method for determining commendatory booth based on similarity of visiting booth and favorite order}

The present invention relates to a method for determining a recommendation booth, and more particularly, considering the similarity of the same visit booth between the user and the neighboring user, as well as the similarity of the visit order of the same booth and the visit preference of the same booth, The present invention relates to a method for determining a visit booth of a neighboring user having a similar tendency of visiting preference as a recommendation booth of a user.

An exhibition or fair is an important marketing tool that is held by exhibitors to inform visitors of new products or services. Recently, ubiquitous technology has been applied to meet various demands of visitors at exhibitions and fairs, and it is possible to determine the movement route of the terminal provided by the visitor to identify the user's visit route or information about the goods or services of the booth visited by the visitor. Provide to the audience through the terminal.

A method of determining a recommendation booth to be provided to a target user in the related art (hereinafter, referred to as the related art 1) is to search for and search for neighboring users having the same user information as the target user, for example, user gender, age, occupation, address, and income. The booth visited by a nearby user was recommended to the target user. That is, the prior art 1 determines neighboring users having similar tendencies as the target user based on user information, and determines booths visited by neighboring users having similar tendencies as the recommendation booths of the target user.

However, since the prior art 1 which determines the recommendation booth based on the similar tendency of the user information determines the similar tendency of the target user and the neighboring users based on the user information, considering the booth information that the target user actually visited at the exhibition or exhibition, There is a problem in that it does not effectively provide a recommendation booth of interest to the target user.

In order to solve the problems of the prior art 1, a method of determining a recommendation booth by a similar tendency of a booth actually visited by a target user and a neighboring user has been proposed (hereinafter, referred to as prior art 2). The prior art 2 determines whether the booth visited by the target user and the booth visited by the surrounding user is the same, calculates the similarity of the same booth visit between the target user and the neighboring user, and visits the same booth as the booth visited by the actual target user. The visitor's booth having a similar tendency is determined as the recommendation booth of the target user. The prior art 2 determines the recommendation booth by considering the booth visit tendency of the target user changing in real time by determining the visit booth of the neighboring users who have similar booth inclination in consideration of the visitor booth of the target user. Has

However, the related art 2 does not consider the movement pattern of the target user and merely considers that the target user and the neighboring users visited the same booth, so that there is a problem of recommending a booth that is irrelevant or difficult to visit. There is a problem in that it is not possible to determine a more accurate and effective recommendation booth considering the movement pattern of the target user.

Patent application No. 10-2011-0000 (hereinafter referred to as "prior application") filed on September 10, 2011 by the applicant of the present invention is to solve the problems of the prior arts 1 and 2 mentioned above, The prior application is to provide a method for determining a recommendation booth considering not only the similarity of visits of the same booth but also the similarity of the visit order of the same booth between the user and the surrounding users. An earlier application may determine the recommended booth that is convenient to visit from the user's current location, considering not only the similarity of visits to the same booth between the user and the neighboring users, but also the similarity of the visit order of the same booth. Recommend booths can be determined in real time.

However, since the previous application calculates the similarity of the sequential visit booth information of the user and the neighbors only when the same booth is arranged at the same unit visit time, the visitor with different unit visit times is arranged in the sequential visit booth information of the user and the neighbor users. The problem with the booth is that it is not considered in the similarity calculation. In addition, in case of visiting the same booth at different unit visit times in the sorted sequential visit booth information of the user and the surrounding user, the elector does not consider the separation time of the unit visit time, that is, the visitor preference of the user and the surrounding user. Has a problem.

The present invention is to solve the problems of the above-described prior art 1, 2 and the prior application, the object of the present invention to achieve the same booth at different unit visit time in the ordered visit booth information of the user and the surrounding users The present invention provides a method of determining a recommendation booth in consideration of the similarity calculation between the visited booths of a user and a neighboring user.

Another object of the present invention is to provide a user's and neighbor's visit preference in consideration of the separation time of different unit visit times when the user visits the same booth at different unit visit times in ordered visit booth information. It is to provide a way to determine the recommendation booth based on.

Another object of the present invention is to provide a method for determining a recommendation booth that is convenient to visit at a current location of a user in consideration of a movement pattern of the user.

In order to achieve the object of the present invention, the present invention for determining the recommendation booth in the recommendation booth determination server is a sequential booth visit of the peripheral user including the last visit booth of the user in the sequential booth visit information of the peripheral user stored in the database unit Retrieving the information, and extracting sequential booth visit information from the searched sequential booth visit information of the neighboring users to the next visit booth of the last visit booth, which is the previous visit booth, to a threshold number or less; Arranging the sequential booth visit information and the sequential visit booth information of the neighboring users extracted below the threshold number so that the last visit booth is the same unit visit time, and the sequential booth visit information of the sorted neighbor users and the sequential visit booth of the user Of the same unit visit time in the information Calculating similarity of the same unit visit time; calculating similarity of the spaced unit visit time of the spaced unit visit time from the sequential booth visit information of the sorted neighbor users and the sequential visit booth information of the user; Calculating the total similarity between the sequential visit booth information of the user and the sequential booth visit information of the surrounding users by summing the similarity of the visits of the distance units, and following the last visit booth in the sequential visit booth information of the surrounding users based on the result of the total similarity calculation Calculating a recommendation level of the visited booth and determining a booth having the highest recommendation rate as the recommendation booth.

The next visit booth with the highest recommendation calculated here is determined as the recommendation booth.

Preferably, the method further includes determining whether the determined recommendation booth is included in the sequential visit booth information of the user, and when the determined recommendation booth is included in the sequential visit booth information of the user, the next visit having the next highest recommendation degree. Determine the booth as the recommended booth.

Preferably, the same unit visit time similarity may be calculated as an index of the difference between the unit visit time of each same unit visit time and the last visit booth according to whether the same booth is visited based on the same unit visit time.

Preferably, the distance unit visit time similarity is calculated as an index of the difference between the unit visit time visiting the same booth and the unit visit time of the user visiting the same booth and the unit visit time of the last visit booth.

The recommendation booth determination method according to the present invention has various effects as follows.

First, the recommendation booth determination method according to the present invention includes the case where the user visits the same booth at different unit visit times in the arranged sequential visit booth information of the user and the neighboring user, by including the similarity calculation of the sequential visit booth of the user and the neighboring users. In addition, the booth not visited during the same unit visit time can be accurately determined to recommend the user's booth.

Secondly, the recommendation booth determination method according to the present invention considers the separation time of different unit visit times when the user and the neighbor users visit the same booth at different unit visit times in the ordered sequential visit booth information. By calculating the similarity of the sequential visit booths of, the recommendation booth may be determined based on the visit preferences of the user and the neighboring users.

Third, the recommendation booth determination method according to the present invention may determine the recommendation booth, which is convenient for the user to visit in the booth which the user visited last, by determining the recommendation booth in consideration of the movement pattern and the visit preference of the user.

1 is a functional block diagram illustrating a recommendation booth determination system according to the present invention.
2 is a flowchart illustrating a recommendation booth determination server according to the present invention in more detail.
3 is a flowchart illustrating a recommendation booth determination method according to the present invention.
4 illustrates an example of sequential visit booth information of a neighbor user stored in a database unit.
FIG. 5 illustrates an example of sequential visit booth information of an extracted sorted user and a neighboring user for calculating the same unit visit time similarity.
FIG. 6 illustrates an example of sequential visit booth information of an extracted sorted user and a neighboring user for calculating a distance unit visit time similarity.
7 is a flowchart illustrating the recommendation booth determination step according to the present invention in more detail.

1 is a functional block diagram illustrating a recommendation booth determination system according to the present invention.

In more detail with reference to Figure 1, a user who visits an exhibition or exhibition visits the exhibition booth of the museum or exhibition with a user terminal 100, respectively. The user terminal 100 is connected to the recommendation booth determination server 300 or the booth terminal 400 through the wired / wireless network 200. The user terminal 100 transmits an identifier of the user terminal 100 to the recommendation booth determination server 300 or the booth terminal 400 through the wired / wireless network 200. When the user carries the user terminal 100 and visits the exhibition booth, the user terminal 100 recognizes the visited exhibition booth and displays the recognized booth identifier, the user terminal 100 identifier, and the visit time of the recommended booth determination server ( 300). In another embodiment of the present invention, the booth terminal 400 recognizes a user terminal identifier visiting the exhibition booth, and recommends information on the recognized user terminal identifier, booth identifier, and visit time through the wired / wireless network 200. Send to booth decision server.

On the other hand, the user through the wired / wireless network 200 through the wired / wireless network 200 through the user interface (not shown) provided in the user terminal to request a recommendation booth at the exhibition or exhibition visited by the user terminal 100 It transmits to 300, and receives the information on the recommendation booth from the recommendation booth determination server 300.

The recommended booth determination server 300 receives and stores information on an exhibition booth visited by a user from the user terminal 100 or the booth terminal 400. When the recommendation booth determination server 300 receives a recommendation booth request command from the user terminal 100, the sequential visit booth information of neighboring users is extracted to a threshold number or less, and the last visit booth of the user who requested the recommendation booth is the same unit visit time. By arranging such that the user's sequential visit booth information and the neighbor's sequential visit booth information of the user and the neighboring users visit the same booth based on each unit visit time, the similar unit visit time similarity and the distance unit visit time similarity are calculated. Decide which booth to recommend to the user. The recommendation booth determined by the recommendation booth determination server 300 is transmitted to the user terminal 100 requesting the recommendation booth through the wired / wireless network 200.

2 is a flowchart illustrating a recommendation booth determination server according to the present invention in more detail.

Referring to FIG. 2 in more detail, the transceiver 310 receives information and time information about the booth visited by the user from the user terminal 100 or the booth terminal 400 through the wired / wireless network 200, and the database. The unit 320 stores information about booths visited by each user according to time, that is, sequentially visited booths. Preferably, the database unit 320 is divided by a user identifier, and stores booth information visited by a user sequentially as an identifier of a visited booth.

When the search unit 330 receives a recommendation booth request command from the user terminal 100 through the transceiver 310, the visitor booth information and surroundings of the user who requested the recommendation booth, which are stored in the database unit 320, are received. Based on the sequential visit booth information of the user, the sequential visit booth information of the neighbor user including the last visit booth of the user is searched among the sequential visit booth information of the neighbor users.

The arranging unit 340 extracts the sequential visit booth information from the searched sequential visit booth information of the neighboring user to the next visit booth of the last visit booth, which is the previous visit booth, to a threshold number or less. In addition, the alignment unit 340 aligns the sequential visit booth information and the sequential visit booth information of the neighboring users extracted below the threshold number so that the last visit booth is the same unit visit time. Here, the unit visit time refers to the order of booths in which a user or a neighboring user visits sequentially. For example, if a user visits the booths B001, B002, B003, and B004 sequentially, the unit visit time of the booths B001, B002, B003, and B004 that the user visited is irrespective of whether the user stayed at each booth or the actual time of visit. T1, T2, T3, and T4.

The similarity calculation unit 350 separates the same unit visit time similarity from the sorted sequential visit booth information of the neighboring users and the sequential visit booth information of the user based on whether the same booth was visited at each unit visit time based on each unit visit time. The unit visit time similarity is calculated, and the total similarity between the neighboring users and the sequential visit booth information of the user, which are aligned by adding the same unit visit time similarity and the spaced unit visit time similarity, are calculated. First, the similarity calculator 350 determines whether the same booth visited the same unit at the same unit visit time based on the visit time of each unit in the sorted sequential booth visit information and the sequential visit booth information of the user. Calculate Also, the similarity calculator 350 determines whether the same booth is visited at different unit visit times based on the sorted sequential booth visit information of the neighboring users and the sequential visit booth information of the user, and the distance of the unit visit time of the same booth visits the same booth. Calculate the similarity of visit time in the spaced unit accordingly.

The recommendation booth determination unit 360 calculates the recommendation of the next visit booth from the sum of the similarities of each of the next visit booths of the last visit booth in the sequential visit booth information of the neighboring users based on the overall similarity calculation result, and the highly recommended booth Decide on the recommendation booth. The transceiver 310 transmits the recommendation booth determined by the recommendation booth determination unit 360 to the user terminal requesting the recommendation booth through the wired / wireless network 200.

3 is a flowchart illustrating a recommendation booth determination method according to the present invention.

Referring to FIG. 3, the sequential visit booth information of the neighboring user including the last visit booth of the user is searched for from the sequential visit booth information of the neighboring user stored in the database unit (S100). 4 is an example of sequential visit booth information of neighboring users stored in the database unit. When a user who requests a recommendation booth sequentially visits B001, B002, B003, B005, and B006, the user sequentially visits booth B006. Search for neighbor user 1, neighbor user 2, neighbor user 3, and neighbor user 7 included in the visited booth information. Here, Peripheral User 1 and Peripheral User 2 visited B006 at Unit Visit Time (T8), Peripheral User 3 visited B006 at Unit Visit Time (T5), and Peripheral User 7 entered B006 at Unit Visit Time (T4). Visited That is, regardless of whether the last visit booth B006 is visited at the same unit visit time as the user, the neighboring user who includes the last visit booth in the sequential visit booth information is searched.

From the sequential visit booth information of the searched neighbor users, the sequential visit booth information from the last visit booth to the next visit booth of the last visit booth is extracted below a threshold number (S200). In the sequential visit booth information of neighboring users extracted below the threshold number, the last visit booth coincides with the unit visit time of the last visit booth of the sequential visit booth information of the user, that is, the sequential visit booth information of the user and the extracted sequential visit booth information of the neighboring users. In step S300, the sequential visit booth information of the user and the sequential visit booth information of the extracted neighboring users are arranged such that the last visit booth is located at the same unit time (S300).

From the sequential booth visit information of the sorted neighbor users and the sequential visit booth information of the user, the same unit visit time similarity is calculated based on whether the same booth is visited at the same unit visit time (S400), and the sequential booth visit information of the sorted neighbor users In step S500, the similarity is calculated based on whether the same booth is visited at different unit visit times in the user's sequential visit booth information (S500). The total similarity of the sorted sequential visit booth information of the neighboring user and the user is calculated from the sum of the calculated unit visit time similarity of the neighboring user and the sorted sequential visit booth information of the user and the distance visit unit time similarity (S600).

Preferably, the same unit visit time similarity is calculated as an index of the difference between the unit visit time of each same unit visit time and the last visit booth according to whether the same booth is visited on the basis of the same unit visit time, more preferably the same unit visit. The temporal similarity SDD is calculated by the following equation (1).

[Equation 1]

Where p is the threshold number, i is the unit visit time, f (α) is a weight function, and if the same booth is arranged at the same unit visit time, it has α value and if the same booth is not arranged at the same unit visit time, It has a value of 0, wherein α is characterized by a value of 1 or less.

Meanwhile, the distance similarity between the visits of the spaced units is calculated as an index of the difference between the unit visit times of users and neighboring users who visited the same booth and the difference of unit visit times of users who visited the same booth and the unit visit time of the last visit booth. The distance unit visit time similarity SDD is calculated by Equation 2 below.

&Quot; (2) "

Where p is the threshold number, i is the unit visit time, d is the difference between the unit visit times of visits to the same booth of user t and the surrounding user j, f (α) is a weight function, and the same booth is an α value. Different booths have a value of 0, wherein α has a value of 1 or less.

FIG. 5 illustrates an example of sequential visit booth information of an extracted sorted user and a neighboring user for calculating the same unit visit time similarity.

The threshold number was set to 6 to extract sequential visit booth information of neighboring users, and may be set to various numbers of threshold numbers according to the field to which the present invention is applied. Referring to FIG. 5, the sequential visit booth information from the sequential visit booth information of the neighboring users including the last visit booth of the user to the next visit booth of the last visit booth is extracted as a maximum of six unit visit times, that is, the threshold number 6. . In the case of the neighboring user 1, the sequential visit booth (B002, B007, B001, B005, B006, B013) is extracted to include up to six unit visit times from the last visit booth (B006) to the next visit booth (B013). The visit booth B006 is arranged to match the unit visit time T5 of the last visit booth of the user's sequential visit booth information. In the case of the neighboring user 2, the sequential visit booth (B001, B015, B010, B009, B006, B013) is extracted to include up to six unit visit times from the last visit booth (B006) to the next visit booth (B013). The visit booth B006 is arranged to match the unit visit time T5 of the last visit booth of the user's sequential visit booth information. In the case of the neighboring user 3, the sequential visit booth (B014, B001, B007, B008, B006, B011) is extracted to include up to six unit visit times from the last visit booth (B006) to the next visit booth (B011). The visit booth B006 is arranged to match the unit visit time T5 of the last visit booth of the user's sequential visit booth information. In the case of neighboring user 3, the sequential visit booth (B001, B003, B005, B006, B010) was extracted to include up to six unit visit times from the last visit booth (B006) to the next visit booth (B010). (B006) arranges to match the unit visit time (T5) of the last visit booth of the user's sequential visit booth information.

The same unit visit time similarity (SSD) between the user T and the neighboring user 1 is calculated as in Equation 3 below.

&Quot; (3) "

Similarly, the visit time similarity between the sequential visit booth information of all the surrounding users and the sequential visit booth information of the user arranged in the same manner is calculated.

FIG. 6 illustrates an example of sequential visit booth information of an extracted sorted user and a neighboring user for calculating a distance unit visit time similarity.

Referring to FIG. 6, in the ordered sequential visit booth information of the user and the neighboring user 1, the user and the neighboring user 1 visited the same booth B001 at the unit visit time (T1) and the unit visit time (T3), respectively. B002 was visited at unit visit time (T2) and unit visit time (T1), respectively. That is, the user and the neighboring user 1 visited the same booth B001 at a spaced unit visit time of 2 and the same booth B002 at a spaced unit visit time of 1.

An example of the distance unit visit time similarity (SDD) of the user and the neighboring user 1 may be calculated as in Equation 4 below.

Similarly, the visit time similarity between the sequential visit booth information of all the surrounding users and the sequential visit booth information of the user who are arranged in the same manner is calculated.

Based on the overall similarity calculation result above, the recommendation degree of the next visit booth of the last visit booth is calculated from the sequential visit booth information of the surrounding users, and the highly recommended booth is determined as the recommendation booth to be recommended to the user (S700). The recommendation degree is calculated by the next visit booth, which is calculated by the sum of the total similarity of the sequential visit booth information of the neighboring users and the sequential visit booth information of the sorted neighbors and the total similarity of each next visit booth.

For example, the total similarities between users, neighbors 1, neighbors 2, neighbors 3, and neighbors 7 are 0.714, 0.643, 0.5, and 0.789, respectively, and neighbors 1, neighbors 2, neighbors 3, and neighbors 7 If the next visit booth information of B013, B013, B011, and B010 in the sequential visit booth information is respectively, the recommendations for booths B013, B011 and B010 are respectively calculated as follows.

6 is a flowchart illustrating the recommendation booth determination step according to the present invention in more detail.

Looking in more detail with reference to Figure 6, to calculate the recommendation calculated for each of the next visit booth, calculates and compares the size of the recommendation for each visit booth (S710). Based on the recommendation size comparison, the next visit booth having the highest recommendation is determined as the candidate recommendation booth (S720). It is determined whether the determined candidate recommendation booth exceeds the threshold recommendation degree TH (S730), and when the determined candidate recommendation booth exceeds the critical recommendation degree, it is determined whether the determined candidate recommendation booth is included in the user's sequential visit booth information ( S740). If the candidate recommendation booth is not included in the user's sequential visit booth information, the candidate recommendation booth is determined as the final recommendation booth (S750).

Based on the recommendation calculated according to the field to which the present invention is applied, all of the next visit booths of a high recommendation high group may be determined as candidate recommendation booths, and in this case, the recommendation degree of the candidate recommendation booth determined exceeds the critical recommendation. It is determined whether only the candidate recommendation booths not included in the user's sequential visit booth information among the candidate recommendation booths exceeding the critical recommendation degree are determined as final recommendation booths.

Meanwhile, the above-described embodiments of the present invention can be written as a program that can be executed in a computer, and can be implemented in a general-purpose digital computer that operates the program using a computer-readable recording medium.

The computer-readable recording medium may be a magnetic storage medium (for example, a ROM, a floppy disk, a hard disk, etc.), an optical reading medium (for example, a CD-ROM, DVD, etc.) and a carrier wave (for example, the Internet). Storage medium).

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is evident that many alternatives, modifications and variations will be apparent to those skilled in the art. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

100: user terminal 200: network
300: recommended booth determination server 400: booth terminal

Claims (7)

In the method for determining the recommendation booth in the recommendation booth determination server using the neighboring user and the visited booth information of the user,
Retrieving sequential booth visit information of the neighbor user including the last visit booth of the user from the sequential booth visit information of the neighbor user stored in a database unit;
Extracting sequential booth visit information from the searched sequential booth visit information of the neighboring user to the next visit booth of the last visit booth, wherein the last visit booth is the previous visit booth;
Arranging sequential booth visit information of neighboring users extracted below the threshold number so that the last visited booth is the same unit visit time and sequential visit booth information of the user;
Calculating similarity of the same unit visit time of the same unit visit time from the sorted sequential booth visit information of the neighboring users and the sequential visit booth information of the user;
Calculating a similarity of spaced unit visit times of spaced unit visit times from the arranged sequential booth visit information of the neighboring users and the sequential visit booth information of the user;
Calculating the total similarity between the sequential visit booth information of the user and the sequential booth visit information of the neighboring user by adding the same unit visit time similarity and the distance unit visit time similarity; And
Calculating a recommendation degree of the next visit booth of the last visit booth based on the sequential visit booth information of a neighboring user based on the overall similarity calculation result, and determining the booth having the highest recommendation rate as the recommendation booth; Recommended booth determination method characterized by the above-mentioned. The method of claim 1, wherein the recommendation booth determination method is
And determining whether the determined recommendation booth is included in the sequential booth visit information of the user. 3. The method of claim 2,
The recommendation calculated above determines the next visit booth as the recommendation booth.
If the determined recommendation booth is included in the sequential visit booth information of the user, the recommendation booth determination method characterized in that for determining the next visit booth having the next highest recommendation as the recommendation booth. The similarity visit time of claim 3, wherein:
The booth determination method of claim 1, wherein the same booth determination time is calculated based on whether the same booth is visited or not, based on whether the same booth is visited. The method of claim 4, wherein the same unit visit time similarity is calculated by Equation 1 below.
[Equation 1]

Where p is the threshold number, i is the unit visit time, f (α) is a weight function, and if the same booth is arranged at the same unit visit time, it has α value and if the same booth is not arranged at the same unit visit time, Has a value of 0, wherein α is a value of 1 or less. 4. The similarity of claim 3, wherein the distance unit visit time similarity is
And a distance difference between unit visit times for visiting the same booth and a unit visit time of the user who visited the same booth and an index of the difference between unit visit times for the last visit booth. The method of claim 6, wherein the separation unit visit time similarity is calculated by Equation 2 below.
&Quot; (2) "

Where p is the threshold number, i is the unit visit time, d is the difference between the unit visit times of visits to the same booth of user t and the surrounding user j, f (α) is a weight function, and the same booth is an α value. And different booths have a value of 0, wherein α has a value of 1 or less.

Images