KR101873319B1 - System for providing real time advertising and event service for small business owners - Google Patents

Abstract

An embodiment of the present invention relates to a real-time publicity and event providing system, and a technical problem to be solved is to provide a system and a method for realizing a real-time publicity and event providing system by allowing a customer to browse and receive franchise information or event information desired by a customer in a desired area, , And the merchant is provided with the information of the customer's review rating to enable more efficient promotion and event provisioning.
To this end, according to an embodiment of the present invention, a publicity and event management application is installed, and various merchant information is searched based on current location information or local information set by a customer on the web connected by the execution of the publicity and event management application A customer terminal for receiving promotional and event information provided by each merchant, registering a visit date of the merchant, and receiving review rating information on the visit date of the merchant; A publicity and event management application is installed and registers the merchant's business card, publicity and event information on the connected web by the execution of the publicity and event management application, and stores customer analysis information on customers visiting or interested in the affiliated store, The merchant terminal receiving the review analysis information on the merchant visit history information and the review rating information and the search analysis information on the merchant information and the search information on the publicity and event information; And providing the promotion and event management application, the merchant information, the promotion and event information, the review rating information, the customer analysis information, the review analysis information, and the search analysis information in cooperation with the customer terminal and the merchant terminal, And a management server for calculating the review analysis information through a predetermined review analysis algorithm.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system and method for real-

An embodiment of the present invention relates to a real-time publicity and event providing system.

In general, many of the mainstream customers of in-store shops, especially those operated by small merchants, are residing (or working) around shops or passing around shops. Or regular customers who are satisfied with the goods or services of those shops and come to visit continuously. Small-scale offline stores often open discounts and gift-giving events, and are attempting event marketing to promote to customers while reducing inventories. Stores that require inventory management, such as grocery stores, restaurants, and fashion-conscious fashion stores, are more actively using event marketing methods.

Actively marketing through events is effective, but the way to notify customers of the event is very limited. For example, an advertisement leaflet carrying an advertisement of a shop is distributed to passersby passing around the shop, or an advertisement or an event is displayed on a local handbag. However, since the event takes place within a certain period of time, the advertising methods such as the advertisement flyers and the local handbags, which are less responsive, are limited.

Recently, as the development and distribution of communication networks have expanded, local advertisements using the Internet or wireless communication networks are being developed. For example, an offline store owner promotes his own exclusive website and promotes his or her own store in a local information service such as a portal site. The former has the advantage that the store owner can provide his or her own advertisement or event as desired, but the problem is that the construction cost is high and the number of visitors is low. While the latter has the advantage of being able to promote to the visitors of the portal site, it has a disadvantage in that it is very limited in positively providing its own advertisement or event.

Japanese Patent Application Laid-Open No. 10-2014-0122477 (published on October 10, 2014)

One embodiment of the present invention is to improve the usability of the information and the customer convenience by allowing the customer to view and receive the merchant information or event information desired by the customer in an area desired by the customer and to provide the merchant with the review rating information And provides a real-time publicity and event providing system that enables events to be provided.

Other objects and advantages of the present invention will become apparent from the following description taken in conjunction with the accompanying drawings, in which: FIG.

A PR and event providing system according to an embodiment of the present invention is provided with a PR and event management application and is configured to receive current location information or local information set by a customer on the connected web by executing the PR and event management application A customer terminal that receives information on promotional and event information provided by each merchant, registers a visit date of the merchant, and provides review rating information on the visit date of the merchant; A publicity and event management application is installed and registers the merchant's business card, publicity and event information on the connected web by the execution of the publicity and event management application, and stores customer analysis information on customers visiting or interested in the affiliated store, The merchant terminal receiving the review analysis information on the merchant visit history information and the review rating information and the search analysis information on the merchant information and the search information on the publicity and event information; And providing the promotion and event management application, the merchant information, the promotion and event information, the review rating information, the customer analysis information, the review analysis information, and the search analysis information in cooperation with the customer terminal and the merchant terminal, And a management server for calculating the review analysis information through a predetermined review analysis algorithm.

The customer analysis information may include at least one of peripheral customer information located within a predetermined radius of the merchant store, merchant card information, merchant information, interest customer information obtained by searching for promotional and event information, The average information of the surrounding customers, the interested customers, and the purchasing customers, the number of the customers who have viewed or downloaded the publicity and event information within the predetermined period of time, Public relations / event status and performance information.

The review analysis information includes at least one of the number of registered items of the last visit to the merchant shop, the review rating information, the daily review rating trend information, the gender or age review registration rate information, the review registration ratio information per visit type, And may include star rating information.

The customer terminal can arrange the searched various merchant information in descending order based on the location information or the area information, or sort and display the searched merchant information in the review rating order based on the review rating information of the merchant.

The management server can provide event information of the franchisees located within a predetermined radius in real time based on the location information of the customer.

Wherein the review analysis information includes a merchant conditional expectation value, an observation signal value, and review rating information calculated by a review rating calculation engine, and the merchant conditional expectation value includes an nth customer As an evaluation value at the time of evaluation,

Lt; / RTI >

)은At this time, the signal value observed by the nth customer

)silver

은 개인의 엄격함 변수,

은 개별평점,

은 가중치,

,

은 분산,

은 고객의 평가시의 기대치)에 의하여 산출될 수 있다.(here,

Is an individual's strictness variable,

Individual ratings,

Weight,

,

Is distributed,

Can be calculated by the expectation value at the time of evaluation of the customer.

)를 결정짓는 요소는 평가자가 기존에 평가한 평가의 개수(

), 그 평가자가 일정 기간 동안에 제출한 평가의 개수(

), 그 평가자와 평가대상 가맹점 사이의 적합도(matching)의 차이(

), 그 평가자와 평가대상 가맹점 사이의 방문 다양성의 지표(

) 및 그 평가자가 평가대상 가맹점을 방문한 시간 간격(

) 지수의 산술적 합The individual's severity variable (

) Is the number of evaluations that the evaluator has previously evaluated (

), The number of evaluations submitted by the evaluator over a period of time (

), The difference in the matching between the evaluator and the participating merchant

), An indicator of visit diversity between the evaluator and the merchant

) And the time interval during which the evaluator visited the merchant

) The arithmetic sum of the exponent

는 상수, α는 계수)(here

Is a constant, and? Is a coefficient)

. ≪ / RTI >

은 The individual rating

silver

은 개인의 엄격함 변수,

,

는 가중치,

은 신호입력값,

은 기존에 방문한 사람들이 평가한 평점에 대한 개별평점 정보(

) 및 신호입력값(

)을 고려한 가맹점의 참품질에 대한 예측값)(here,

Is an individual's strictness variable,

,

Weight,

A signal input value,

Is the individual rating information about ratings that people have visited

) And the signal input value (

) Is a predicted value for the true quality of the franchisee,

. ≪ / RTI >

The real-time public relations and event providing system according to an embodiment of the present invention can improve the usability of information and the customer convenience by allowing the customer to view and receive the merchant information or event information desired by the customer in an area desired by the customer, Rating information can be provided to enable more efficient promotion and event provisioning.

1 is a block diagram schematically showing a real-time publicity and event providing system according to an embodiment of the present invention.
2 is a block diagram schematically showing the management server of FIG.
3 is a diagram illustrating an example of a Web page item provided to an affiliate terminal in a real-time publicity and event providing system according to an embodiment of the present invention.
4 is a diagram illustrating an example of a Web page item provided to a customer terminal in a real-time publicity and event providing system according to an embodiment of the present invention.
5A to 5I are diagrams showing a state in which the publicity and event management application is executed in the merchant terminal of FIG.
6A to 6L are diagrams showing a state in which a publicity and event management application is executed in the customer terminal of FIG.

The terms used in this specification will be briefly described and the present invention will be described in detail.

While the present invention has been described in connection with what is presently considered to be the most practical and preferred embodiment, it is to be understood that the invention is not limited to the disclosed embodiments. Also, in certain cases, there may be a term selected arbitrarily by the applicant, in which case the meaning thereof will be described in detail in the description of the corresponding invention. Therefore, the term used in the present invention should be defined based on the meaning of the term, not on the name of a simple term, but on the entire contents of the present invention.

When an element is referred to as " including " an element throughout the specification, it is to be understood that the element may include other elements as well, without departing from the spirit or scope of the present invention. Furthermore, the term " part " or the like described in the specification means a unit for processing at least one function or operation, which may be implemented by hardware or software, or a combination of hardware and software.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In order to clearly illustrate the present invention, parts not related to the description are omitted, and similar parts are denoted by like reference characters throughout the specification.

FIG. 1 is a block diagram schematically showing a real-time publicity and event providing system according to an embodiment of the present invention, FIG. 2 is a block diagram schematically showing a management server of FIG. 1, FIG. 4 is a diagram illustrating an example of a Web page provided to a merchant terminal in a real-time publicity and event providing system according to an embodiment of the present invention. FIG. 5A to 5I are diagrams showing a state in which a publicity and event management application is executed in the merchant terminal of FIG. 1, and FIGS. 6A to 6L are views showing an example of an item of a publicity and event And the management application is executed.

1, a real-time publicity and event providing system according to an embodiment of the present invention includes a customer terminal 10, a merchant terminal 20, and a management server 30.

The customer terminal 10 is a terminal device used by a customer who desires to receive the real-time publicity and event providing service, and is a portable terminal device having a network function of a wireless network (or mobile communication network) such as a smart phone, a tablet PC, . Of course, the customer terminal 10 should be a portable terminal capable of executing applications and capable of generating and collecting location information.

4, the customer terminal 10 is installed with a publicity and event management application, and based on the current location information or the local information set by the customer on the web connected by the execution of the publicity and event management application Searches for various merchant information, receives promotional and event information provided by each merchant, registers a merchant visit date, and provides review rating information on a merchant visit date.

The customer terminal 10 can arrange the searched various merchant information items in descending order based on the location information or the regional information, or sort and display the searched merchant information items in the review rating order based on the review rating information of the merchants.

6A to 6L, the state in which the PR and event management application is executed in the customer terminal 10 will be described in turn.

6A, a menu slide, a menu tab menu, and store list information are displayed on the home page of the connected web by the execution of the publicity and event management application. At this time, the menu slide can be displayed by sliding the menu list area. The menu tab menu is composed of a main screen tab menu and can be moved to the next tab menu (i.e., page) by swiping the screen. A representative image, a name of a merchant, a distance from a current position, a business type, a representative menu (item), and whether or not a favorite is added, so that the user can move to the store detail page when the store is selected.

Also, as shown in FIG. 6B, the customer can search for desired merchant information or merchant information through a search, a search word input, and a recent search term menu on the screen.

As shown in FIG. 6C, the customer can set a region / sorting criterion on the screen. For this purpose, menus such as region setting, sorting criterion, region setting popup, sorting criterion selection popup are displayed on the screen. In this case, when the current location icon is selected, the current location information is provided in the region setting, the region setting pop-up is displayed when the region selection region is selected, the sorting reference setting pop- When selecting the city / state button, you can expose the picker at the bottom and select the city / county / district when you select the city / county / district button. So that it is immediately applied when the sorting criterion is selected and the pop-up is closed.

6D, the customer can use the merchant detail menu (i.e., store details) on the screen. For this purpose, the customer can store detailed information, favorite shop registration, store basic information, business card SNS share, Connection, 1: 1 inquiry, homepage connection, event, review item are displayed. At this time, when the store is selected from the store detailed menu, it is moved to the store detail page. In the store basic information, the business card information registered by the merchant manager is provided. Profile information is provided when the business card is not registered. FIG. 6E). In an event, an event currently in progress of the store may be exposed, a phrase may be displayed if there is no event in progress, and a review may provide review information on the store.

Further, as shown in FIG. 6F, the customer can browse information about the coupon by operating buttons (before downloading, download completion, coupon use completion, expiration of time) according to the coupon status in the store details.

As shown in FIG. 6G, the customer can be provided with a review registration and a store review rating menu in the store details. At this time, the customer goes from the review registration to the review registration page. In the store review rating, Can be provided.

Also, as shown in FIG. 6H, the customer can receive review information, review review, store review rating, review sorting criteria, and review list information from the review list. At this time, when the review button of the store detailed page is selected, the review information is moved to the review page of the store, the review is made in the review, and the rating information of the overall review rating and detail item is provided in the store review rating. In the review sorting list, the list is sorted according to the latest registration order, the highest ranking order and the lowest ranking order. If the default number of exposures is 10 in the review list, 10 is additionally loaded for the bottom scroll.

As shown in FIG. 6I, the customer registers a review and selects the location icon of the event page, as shown in FIG. 6J, to check the location of the store. As shown in FIG. 6K, , It is possible to confirm the detailed event information and confirm desired coupon information and usage information as shown in FIG.

The merchant terminal 20 is a terminal device possessed by a store owner or a manager of a small and medium-sized business enterprise having a computer terminal including a general PC having a network function and a portable terminal (smart phone, tablet PC). At this time, the small and medium-sized business may include a store having a physical space regardless of the business type, such as a restaurant, an apparel shop, a cosmetics shop, a beauty shop, and a bookstore.

As shown in FIG. 3, the merchant terminal 20 has a publicity and event management application installed therein. The merchant terminal 20 registers merchant card, promotional and event information on the web connected by the publicity and event management application, Review analysis information on customer's visit history information and review rating information, and search analysis information on merchant information and search information on publicity and event information.

5A to 5I, a description will be given of a state in which the promotion and event management application is executed in the merchant terminal 20 of FIG.

First, the merchant terminal 20 accesses the management server 30 to set an inquiry date. Then, the merchant terminal 20 sets up an inquiry date and stores it in the shop (i.e., a customer who used coupons at shops within 500 meters of the shop within the period of time) In other words, customers who view, download, and use the event, publicity flyer, shop card, and store added to their favorite stores), purchasing customers (ie customers using coupons), average number of customers per day, (Ie, the number of event visits / downloads in the period covered), publicity performance, number of review registrations, review ratings, daily review rating trends, visits to the homepage , Number of new visitors, number of visits to the homepage, page views, number of new visits and revisits.

In addition, registration of affiliation event, event search (that is, progress status, inquiry period), event list recommendation event, event period, event performance, publicity registration / correction, publicity result, member registration / Or may be provided.

The management server 30 is a server device of a company providing the real-time publicity and event providing service while providing an application for establishing an interface environment through network connection with the merchant terminal 20 and the customer terminal 10. [ That is, the management server 30 is interlocked with the customer terminal 10 and the merchant terminal 20, and is connected to the customer terminal 10 and the merchant terminal 20, for example, for the promotion and event management application, the merchant information, the promotion and event information, the review rating information, The analysis information is provided through search analysis information or by receiving the information, and the review analysis information is calculated through a predetermined review analysis algorithm. Also, the management server 30 provides event information of the franchisees located within a predetermined radius in real time based on the location information of the customer.

At this time, the network may be a wired / wireless environment communication network, various interface environments, a network environment, and a platform environment. For example, the merchant terminal 20 can be provided with a publicity and event management application by accessing a homepage provided on the web by the management server 30 through a web browser. Further, the merchant terminal 10 can be provided with a wireless network The real-time publicity and event providing service viewed from the management server 30 can be provided by the publicity and event management application installed through the mobile communication network.

2, the management server 30 includes a communication unit 310, a member authentication unit 320, a customer information analysis unit 330, a review information analysis unit 340, a search information analysis unit 350, A review analysis information calculation unit 360, a payment processing unit 370, and a storage unit 380. [

The communication unit 310 receives information from the customer terminal 10 and the merchant terminal 20 connected to the wired and wireless communication network through a publicity and event management application, merchant information, publicity and event information, review rating information, And provides search analysis information or receives the corresponding information.

The member authenticating unit 320 receives the customer unique information (ID and password) received from the registered customer terminal 10 and the affiliated shop terminal 20 connected to the web page provided by the management server 30 via the wireless network, ) With the member information stored in advance in the storage unit, and allows the connected customer and the franchisee to use the present real-time publicity and event providing service according to whether the comparison result is matched.

The customer information analyzer 330 may participate in a corresponding event based on surrounding customer information, merchant card information, merchant information, interested customer information, and event information that are located within a predetermined radius based on the merchant Purchase using coupon provided by the merchant Promotion of the number of customers who browsed or downloaded public information and event information within a predetermined period of time, average information of nearby customers, interested customers and purchasers, customer information, a predetermined date and time zone / Customer information including event status and performance information is analyzed by a predetermined analysis program.

The customer information analyzer 330 may provide the analyzed customer information to the storage unit 380 to store and manage the customer information.

The review information analyzing unit 340 may include information on the number of registrations of the merchant's visit, review rating information, daily review rating trend information, gender or age-based review registration ratio information, review type registration ratio information per visit type, Review information including rating information for each item is analyzed by a predetermined analysis program. The analyzed review information is provided to the storage unit 380 and stored and managed.

The search information analyzing unit 350 is a device for analyzing a homepage visitor, and is a device for analyzing a homepage visitor. The search information analyzing unit 350 is a device for analyzing a homepage visitor, and includes a homepage visitor count, a number of new visitors, a homepage visit count, a page view, Pages, and the like are analyzed by a predetermined analysis program. The analyzed search information is provided to the storage unit 380 and stored and managed.

The review analysis information calculation unit 360 includes the franchise conditional expectation value, the observation signal value, and the review rating information calculated by the review rating calculation engine.

Hereinafter, the process of calculating the franchise conditional expectation value, the observation signal value, and the review rating information by the review rating calculation engine will be described in more detail.

,

) Consideration of Evaluator Group (

,

)

The evaluator of the merchant satisfaction is divided into the registered evaluator and the unregistered evaluator, and the registered evaluator is divided into the excellent evaluator and the general evaluator.

A registered reviewer (identified by subscript "re") refers to an evaluator who has been given an ID by entering customer information on the website, and a not-registered reviewer (subdivided by subscript nr) The evaluator who made the evaluation on the spot without any action.

The evaluation evaluator is further divided into an elite reviewer (divided by subscript e) and a non-elite reviewer (divided by subscript ne). In the case of the evaluation of the franchise stores, those who had a lot of evaluations were classified as excellent evaluators considering that they were positive in participating in the evaluation. As a result, . The number of evaluations to be used as a criterion is based on the customer information stored in the database, and when the parameter values of the various parameters included in the model to be introduced below are estimated, .

On the other hand, when there are not many evaluators at the beginning as in the evaluation of new merchants, when the evaluator data are not accumulated, the average rating is calculated using the simple average method as in the conventional method, and then the data about the evaluator is accumulated, When the number of evaluations per group (excellent group, general group, and non-registered group) becomes 10 to 30 or more, the optimum aggregation method can be used by estimating the parameters for each evaluator.

)으로, 기존 평가자들의 평가에 대한 관심도를 가중치(

)로 측정한다. 이러한 가맹점 품질에 대한 평가신호의 정확성 지표로서의 분산(

)과 기존 평가자들의 평가에 대한 관심도 지표로서의 가중치(

)를 구하는 까닭은 예를 들어, 우수평가자는 일반평가자에 비해 평판(기존 평가자들의 평가)을 꼼꼼히 읽어보는 것과 같이 기존 평가자들의 평가에 대해 관심이 많고(

) 신호의 정확성이 더 높다(

)고 판단되기 때문에 이러한 차이를 반영하고자 함이다. The accuracy of the evaluation signal for the quality of the franchise acquired after each visit to the franchisee by the excellent evaluator, the general evaluator, and the unregistered evaluator, which can be defined in this way,

), And weights the interest of the existing evaluators in the evaluation (

). The evaluation of the quality of such merchant

) And the weight as an index of interest in the evaluation of existing evaluators (

For example, a good evaluator is more interested in evaluating existing evaluators than the general evaluators (for example, reading reputation (evaluation of existing evaluators) carefully)

) The accuracy of the signal is higher (

), So we want to reflect these differences.

) 그 까닭은 미등록평가자는 웹사이트에서 평점을 매기지 않고 가맹점 방문시에 현장에서 평점을 매기므로 기존 평가자의 평가보다 자신이 현장에서 직접 관찰한 신호에 더 높은 가중치를 줄 것으로 기대되기 때문에 그러한 차이를 반영하기 위함이다. 반면에 신호의 정확성은 사전적으로 어느 그룹이 더 높을지 정해지지 않는다. On the other hand, if the general evaluator and the non-registered evaluator are compared, the unregistered evaluator is expected to be less interested in reputation than the general evaluator

) Because the unregistered evaluators are expected to give a higher weight to the signals they themselves see on the field rather than the ratings of the existing evaluators because they are graded on site when visiting merchants without rating on the website, . On the other hand, the accuracy of the signal is not determined in advance for which group is higher.

)의 고려 The rigor of the evaluator (

) Consideration of

)를 만들어 낸다. 이렇게 개인별 엄격함 변수(

)를 계산하는 이유는 가맹점 관찰 후 발견하는 품질에 대한 객관적인 신호 이외에도 개인적으로 성향에 따라, 과거 경험에 따라 실제 가맹점의 실제의 참품질보다 더 높게 혹은 더 낮게 평가할 수 있기 때문이다. 예를 들어 가맹점을 여러 다양한 목적(예를 들어 가족/친구/사업 등)으로 방문한 경험이 있는 평가자는 새로운 종류의 가맹점에 대해서 반감이 적어 비교적 후하게 평가하는 반면, 가맹점을 하나의 목적, 예컨대 가족 목적으로만 가맹점을 방문했던 평가자는 사업 목적으로 가맹점을 방문할 경우 필요한 서비스를 제공받지 못하는 것으로 인해 가맹점에 대해서 매우 박하게 평가를 할 수 있는 것과 같은 개인적 경험 내지 성향을 고려한 변수이다. Using the individual profile data from the evaluator, the stringency variable (

). Thus,

) Is calculated because, in addition to the objective signal about the quality that is found after the franchisee observation, it can be judged according to the personality and past experience to be higher or lower than the actual true quality of the actual franchisee. For example, an evaluator who has visited a merchant for a variety of purposes (eg, family / friend / business, etc.) evaluates a new type of merchant relatively less with less antipathy, The evaluator who visited the franchise only for the purpose is a variable considering the personal experience or tendency such as being able to evaluate the franchisee very positively due to the lack of the necessary service when visiting the franchise for business purpose.

In the present invention, the following five factors are factors that can affect such an evaluator's strictness.

)이다. 이는 기존에 많은 가맹점을 방문한 경험이 있는 사람과 그렇지 않은 사람의 평가가 다를 수 있음을 고려한 것이다. First, the number of evaluations that the evaluator has already evaluated (

)to be. It is considered that the evaluation of people who have already visited many franchise stores and those who have not.

)이다. 이를 통해 얼마나 자주 평가를 하는 사람인지를 알 수 있다. 예를 들어, 10개월간 2개의 평가를 하는 사람과 일주일간 2개의 평가를 하는 사람을 다르게 취급할 수 있다. 계산식은 총 평가의 개수를 첫 평가 이후 지나간 날짜수로 나눈 수, 즉 일간 평균 평가수이다. 예를 들어, 10일간 2개의 평가를 남긴 사람은 2/10=0.2의 값을 가지며, 100일간 2개의 평가를 남긴 사람은 2/100=0.02의 값을 가지고, 100일간 3개의 평가를 남긴 사람은 3/100=0.03의 값을 가진다. 오직 하나의 평가만을 남긴 사람은 하나 이상의 평가를 남긴 사람들의 평균값으로 대체한다. 일간 평균 평가수가 작은 사람은 가맹점을 매우 드물게 방문하는 사람이고, 반대로 높은 사람은 자주 가맹점에 방문하는 사람으로 볼 수 있다. 이론적으로 생각해 보면 가맹점을 드물게 방문하는 사람은 가맹점에 대한 평가가 후하고 가맹점을 자주 방문해서 비교의 기회가 많은 사람은 가맹점의 평가가 매우 박할 수가 있는 것을 고려하기 위함이다.The second is the number of evaluations submitted during a given unit period (

)to be. This shows how often people evaluate. For example, a person who performs two assessments over a 10-month period and a person who performs two evaluations per week may treat it differently. The formula is the number of total evaluations divided by the number of days since the first evaluation, that is, the average number of evaluations per day. For example, a person who left two evaluations for 10 days has a value of 2/10 = 0.2, a person who left two evaluations for 100 days has a value of 2/100 = 0.02, Has a value of 3/100 = 0.03. A person who leaves only one assessment is replaced by the average value of those who have left at least one assessment. A person with a low average daily rating is a person who visits a merchant very rarely, while a high person is often seen as a person visiting a merchant. Theoretically, those who rarely visit a franchisee are after an evaluation of the franchisee and frequently visit the franchisee to consider that the evaluation of the franchisee can be very poor for those who have a lot of opportunities for comparison.

)이다. 이 차이를 계산함으로써 평가자와 제품 사이에 얼마나 매칭(matching)이 잘 되었는지를 판단할 수 있다. 먼저 개별 가맹점의 특징값을 산정한다. 예를 들어, 가맹점에 방문하는 목적이 (가족/커플/친구/솔로/사업) 등으로 분류가 된다고 가정하자. 세 명의 방문객이 어떤 가맹점에 방문했을 때 그 목적으로 두 명이 가족을, 한 명이 친구로 표시했다면 그‘가맹점의 특징값’은 (0.66/0/0.33/0/0)이 된다. 이때 어떤 방문객(V1)이 그 가맹점(A)을 방문하게 되면 그 가맹점(A)의 산정된 특징값이 방문객(V1)의 아이디에 기록이 되고, 그 방문객(V1)이 다른 가맹점(B)을 방문하게 되면 그 가맹점(B)의 특징값을 기존의 가맹점(A)의 특징값과 합산해서 평균 특징값을 계산한다. 이것이‘그 방문객(V1) 개인의 가맹점 적합도 특징값’이 된다. 여기서 그 방문객(V1)이 새롭게 가맹점(C)을 방문할 때 그 방문객(V1) 개인의 특징값과 가맹점의 특징값의 차이의 제곱들을 모두 더한 값이

값이다. 이 값이 크다는 것은 평가자(V1)가 방문한 가맹점이 그 평가자(V1)가 그 전까지 방문했던 가맹점들의 성향에 비추어볼 때 많이 다르다는 것을 의미하고, 작다는 것은 평가자의 기존 가맹점 방문 성향과 매우 유사하다는 것을 의미한다.The third is the matching difference between the evaluator and the product (

)to be. By calculating this difference, it is possible to determine how well the evaluator has matched the product. First, the feature values of the individual merchants are calculated. For example, suppose that the purpose of visiting a merchant is (family / couple / friend / solo / business). When three visitors visited a certain merchant, the two characteristics of the merchant and the characteristics of the merchant were (0.66 / 0 / 0.33 / 0/0). At this time, when a visitor V1 visits the merchant A, the estimated characteristic value of the merchant A is recorded in the ID of the visitor V1, and the visitor V1 registers the other merchant B And when it is visited, the feature value of the franchisee B is added to the feature value of the existing franchisee A to calculate an average feature value. This is the 'merchant fitness value characteristic value of the visitor (V1)'. Here, when the visitor V1 newly visits the merchant C, a value obtained by adding the squares of the difference between the feature value of the visitor V1 and the feature value of the merchant

Value. This large value means that the franchisee visited by the evaluator (V1) differs greatly in view of the tendency of the evaluator (V1) to visit the previous merchant, and the smallness is very similar to the tendency of the evaluator to visit the existing merchant it means.

)를 만들 수 있다. 이는 평가자가 기존에 방문했던 가맹점들의 각 특징값, 즉 (가족/커플/친구/솔로/사업)에 해당하는 각 점수들의 분산값을 구한 후 이 값들을 모두 합한 후 제곱근을 취하여 계산한다. 이 다양성 지표가 크면 평가자는 다양한 목적으로 가맹점을 방문하는 것이고, 이 지표가 작으면 평가자는 비교적 동일한 목적으로 가맹점을 방문하는 것이다. Fourth, the diversity indicator of how the evaluator has experienced the various merchants

). This is calculated by calculating the variance value of each score corresponding to each characteristic value (family / couple / friend / solo / business) of the franchisees visited by the evaluator, summing these values together and then taking the square root. If the diversity index is large, the evaluator visits the merchant for various purposes. If the indicator is small, the evaluator visits the merchant for the same purpose.

)도 중요한 변수이다. 경험적으로 시간이 지남에 따라 평균등급은 하락하는 경향이 있다. 이 현상의 원인에 대한 해석은 다양한데, 그 한 예로서 최근의 평가자가 초기의 평가자보다 더 엄격한 경향이 있기 때문이다. Fifth, time (

) Is also an important variable. Empirically, the average rating tends to decline over time. There are many different interpretations of the cause of this phenomenon, for example, recent assessors tend to be more stringent than initial assessors.

) 산출모형 Evaluator's individual rating (

) Output Model

After evaluating the merchant, the evaluator models the rating by using the following formula.

은 개인의 엄격함 변수,

,

는 가중치,

은 신호입력값,

은 기존에 방문한 사람들이 평가한 평점에 대한 개별평점 정보(

) 및 신호입력값(

)을 고려한 가맹점의 참품질에 대한 예측값) (here,

Is an individual's strictness variable,

,

Weight,

A signal input value,

Is the individual rating information about ratings that people have visited

) And the signal input value (

) Is a predicted value for the true quality of the franchisee,

) 및 이번에 새롭게 방문해서 관찰한 신호값(

)를 사용하여 평가한 가맹점의 참 품질에 대한 예측값

에

의 가중치를 주고, 본인이 새롭게 관찰해서 얻은 신호값에

의 가중치를 주어서 계산한 후 개인별 엄격함 변수(

)를 포함하여 최종적으로 자신의 평점

을 매긴다고 가정한다. In other words, information about ratings that people previously visited on your website

) And a signal value newly observed at this time (

), The estimated value of the true quality of the merchant

on

, And the signal value obtained by the new observation

, And then the individual strict variable (

) And finally his rating

.

This model can be explained more easily as follows.

을 매길 때, 본인이 알고 있는 정보는 ① 본인의 타입(우수평가자인가 혹은 일반평가자인가), ② 그에 따른

값, ③ 엄격함

, ④ 평가자가 관찰한 가맹점의 품질에 대한 신호값

, ⑤

, 즉, 기존의 평가와 내가 관찰한 정보를 종합했을 때 가맹점의 참품질에 대한 예측값이다. That is, the evaluator himself or herself uses the individual rating

, The information I know is ① my type (excellent evaluator or general evaluator), ②

Value, ③ strictness

, ④ the signal value of the quality of the franchisee observed by the evaluator

, ⑤

, That is, a predicted value for the true quality of the merchant when we combine the information from the existing evaluation and the information I have observed.

뿐이며 이 정보로부터 통계학적인 과정을 거쳐서 ② 평가자의

값, ③ 엄격함

, ④ 평가자가 관찰한 가맹점의 품질에 대한 신호값

⑤

등을 모두 사후적으로 계산해내는 것이다. However, what the system according to the present invention knows is that (1) the type of evaluator (whether it is an excellent evaluator or general evaluator) and the evaluator's personal information,

From this information, through the statistical process, ② evaluator

Value, ③ strictness

, ④ the signal value of the quality of the franchisee observed by the evaluator

⑤

And so on.

에 주어진 공식대로 평점을 매기는 것으로 가정한다. 평가자 본인이 스스로 매긴 값이

이지만 마구잡이로 매기는 것이 아니라 공식에 의해 시스템적으로 매기는 것으로 가정한다. ‘평가자 본인이 스스로 매기는 것’과 ‘공식에 의해 시스템적으로 매기는 것’ 의 두 문장이 모순이 아닌데 이에 대한 이해는 경제학적 사고방식으로 설명될 수 있다. 즉, 경제학에서는 개인이 자유롭게 의사결정을 내리지만 아무렇게나 결정하는 것이 아니라 일정한 합리성 하에서 논리적으로, 공식을 염두에 두고 결정을 한다고 가정한다. 간단한 예를 들면 우유나 과자를 살 때 집 앞에 있는 편의점에서 사지 않고 굳이 멀리 있는 이마트까지 차를 타고 가서 사는 이유를 설명하기 위해서 이마트의 물건 가격, 교통비, 시간, 쇼핑의 즐거움, 폭넓은 선택의 자유, 편의점의 물건 가격, 편리함 등을 x, y, z 등의 변수로 표현하여 부등호가 이마트쪽이 더 클 때 이마트를 방문하는 결정을 내린다고 모형화하는 것이다. 따라서 소비자의 선택은 자유이고 스스로 결정을 내리는 것이지만, 그 결정은 아무렇게나 마구잡이가 아니라 소비자의 생각과 행동을 묘사하는 일정한 ‘공식’에 의해 결정된다고 보는 것이 경제학에서의 기본 분석방법이다. Next, as the evaluator (1)

Is given by the formula given in. The value that the evaluator himself / herself

But it is assumed that it is systematically wagered by the formula, not the random one. An understanding of this can be explained in an economic way of thinking, since the two sentences "Evaluator himself or herself" and "Systematically wagering by formula" are not contradictions. In other words, in economics, it is assumed that an individual freely makes decisions but does not make any decisions, but rather makes decisions based on a certain rationality, logically, with formulas in mind. A simple example is the price of goods, transportation costs, time, shopping pleasure, wide choice of goods, etc. to explain the reasons for buying a milk or a snack in a car rather than buying it at a convenience store in front of the house. , Convenience store's goods price, convenience, etc. by variables such as x, y, and z, and make a decision to visit E-Mart when the inequality is larger. Therefore, it is the basic analysis method in economics that consumers 'choice is free and makes their own decisions, but the decision is determined by a certain' formula 'that depicts consumers' thoughts and actions rather than randomness.

을 매기는 수식으로 돌아가

을 구성하는 수식의 내용을 살펴보면, 기본적으로 평가자는 가맹점의 품질에 대한 평가를 하기를 원한다. 이때 평가자에게는 두 가지 목표가 있을 수 있다. 하나는 방문한 가맹점에 대해서 남들이 뭐라 하든 신경쓰지 않고 자신이 느끼고 경험한 바를 솔직하게 평가하는 것입니다. 이것이

이다. 쉬운 예로

은 가맹점의 서비스에 큰 감동을 받았다던가, 반대로 서비스가 엉망이라서 무척 화가 났다던가 하는 경우에 감각적으로 느끼는 바에 따른 평가이다. 다른 하나는 일종의 공공서비스처럼 다른 사람들에게 가맹점의 참품질을 알리는 것을 사명처럼 여기는 것으로, 기존 평가자들의 평점 정보(

)도 참고하고 본인이 직접 관찰한 가맹점의 품질정보(신호,

)를 종합하여 가맹점의 참품질에 대한 최선의 예측치

를 계산해 내는 것이다. 평가자의 최종적인 평가

는 이 두 목표 사이의 가중평균으로 결정한다. 따라서 공공서비스적인 정신에 두는 가중치를

라고 하고, 다른 사람이 뭐라 하건 자기 감정을 솔직하게 표현하는 것에 두는 가중치를

로 해서 가중평균을 구한다. 여기에 개인적인 엄격함에 대한 조정을 해서 최종적인 평점

을 결정하는 것이다. 이 모형에 따르면 우수평가자는 일반평가자에 비해

가 크다. 즉, 가맹점의 참품질에 대한 최선의 예측치에 더 큰 가중치를 두는 것으로 그 이유로 가맹점의 참품질을 알고자 할 때 우수평가자를 더 중요시 하는 것이다. From this point of view,

Go back to the equation

, The evaluator basically wants to evaluate the quality of the merchant. At this time, the evaluator may have two goals. One is to honestly evaluate what you feel and experience, without worrying about what others say about the merchant you are visiting. this

to be. As an easy example

It is an evaluation according to the sensation felt when the service of the franchisee is greatly impressed or the service is bad and it is very angry. The other is to regard the franchise's true quality as a mission to other people, like a kind of public service,

), And the quality information (signal,

) To give the best estimate of the true quality of the franchisee

. Evaluator's final evaluation

Is determined by the weighted average between these two goals. Therefore, the weight of public service spirit

And weigh the weight of what others express in their honest expressions of their feelings.

To obtain the weighted average. Here we make adjustments to personal strictness,

. According to this model, the superior evaluator

. In other words, we place a greater weight on the best predictions of the true quality of the franchisee, which is why we value the best franchisee when we want to know the true quality of the franchisee.

) True quality of merchants (

)

)는 다음과 같은 식에 의해 변한다고 모형화한다. Next, the true quality of the merchant at time t (

) Is modeled by the following equation.

(여기서

는 품질에 영향을 미치는 변수)

(here

Is a variable that affects quality)

는 정규분포를 따른다고 가정한다. 즉, 현재의 가맹점의 참품질(

)은 이전의 기(期)의 가맹점의 참품질에 새롭게 발생한 변수(

)가 영향을 미쳐서 결정된다고 가정한다. 이 새로운 변수(

)에는 예를 들어 가맹점 식당 웨이터의 사소한 서빙 실수도 포함되고, 새로운 가맹점 지배인의 경영철학에 따른 종업원들의 서비스 마인드의 변화 등도 포함되며 가맹점 설비 등에 대한 제도(법령)의 강화와 같은 비예측적 변수의 발생도 이에 해당한다. 만약 새로운 변화가 계속해서 전혀 발생하지 않는다면(

, t=1,2,...) 참품질의 값도 변화가 없을 것이지만 이는 비현실적인 가정이다. here

Is assumed to follow a normal distribution. That is, the true quality of the current merchant

) Is the newly generated variable in the true quality of the merchant in the previous period

) Are determined to be influenced. This new variable (

) Includes, for example, minor service mistakes at the franchisee's restaurant, including changes in the service mindset of employees according to the management philosophy of the new franchise manager, This is also the case. If a new change does not occur at all (

, t = 1, 2, ...) Although the true quality value will not change, this is an unrealistic assumption.

In the modeling of the change in the quality of the franchisee, the variation (variance) of the true quality with the change of time is expressed as follows.

.

.

) 변화분(

)이 누적되어 참품질의 값이 처음부터 크게 달라지게 된다. 따라서 평가자들의 평가로부터 가맹점의 참품질을 알고자 할 때 n-1번째 평가자 이후에 시간이 많이 지난 후 n번재 평가자가 평가를 남기게 되면 n-1번째 평가 이후로 가맹점의 품질이 크게 변했을 것이므로 상대적으로 n-1번째 평가자보다 n번째 평가자에게 더 높은 가중치를 주어야 할 것이다. In other words, the true quality of merchants

) Change

) Are accumulated so that the value of true quality greatly changes from the beginning. Therefore, if we want to know the true quality of the franchisee from the evaluation of the evaluators, if the n-th evaluator leaves the evaluation after a lot of time since the n-1 evaluation, the quality of the franchisee will change greatly after the n-1 evaluation. it is necessary to give a higher weight to the n-th evaluator than the n-1-th evaluator.

) Evaluator's observation signal (

)

은 예를 들어 가맹점이 썩 나쁘지 않을 경우 평가자가 관찰한 가맹점의 점수는 “이 정도면 별 4개(

)는 되겠네”와 같은 주관적 판단이다. 이 판단은 평가자 본인은 알고 있지만 시스템 분석자의 입장에서는 모르는 값이며 분석자가 알 수 있는 값은 평가자가 최종적으로 매기는 평점인

과 평가자의 개인정보뿐이다. 따라서 최적집계분석을 하기 위해서는 시스템은

과 평가자의 개인정보로부터 평가자의

을 ‘사후적으로’ 계산해 내는 것이다.In the present invention, the evaluator evaluates the quality of a franchisee

For example, if a merchant is not bad enough, the rating of the merchant that the evaluator observed is "

"It is a subjective judgment like" It will be. " This judgment is known to the evaluator but is unknown to the system analyst, and the value that the analyst knows is the evaluator's final rating

And the personal information of the evaluator. Therefore, in order to perform optimal aggregation analysis,

And from the personal information of the evaluator.

'Post-mortem'.

(여기서

은 신호값,

은 평가대상 가맹점의 참품질값,

은 신호값에 영향을 미치는 변수) On the other hand, the signal observed by the n-th evaluator can be expressed by the following equation.

(here

Lt; / RTI >

Is the true quality value of the merchant to be evaluated,

Is a variable that affects the signal value)

)는 정규분포를 따른다고 가정하고, 그것의 조건부 분산인

이 바로 앞서 언급한 n번째 평가자의 신호의 정확성이다. 이 분산이 작다는 것은 n번째 평가자가 가맹점의 참품질에 대한 신호를 정확하게 포착한다는 의미이고, 분산이 크다는 것은 신호를 정확하게 포착하지 못한다는 의미이다. Here, the variable that affects the signal value (

) Is assumed to follow a normal distribution, and its conditional variance

This is the accuracy of the nth evaluator signal just mentioned. This small variance means that the nth evaluator correctly captures the signal about the true quality of the merchant, and the large variance means that it can not capture the signal correctly.

은 n-1번째까지 평가자들의 평가 정보(

)를 사용하여 계산한 n번째 평가 당시의 가맹점의 품질에 대한 기대치로서 통계학적으로 조건부 기대치 혹은 예측치이다.

(N-1) evaluation information of the evaluators

) Is statistically a conditional expectation or forecast as an estimate of the quality of the franchise at the time of the n-th evaluation.

)의 분산 혹은 신호의 정확성이 언급되는 이유는 다음과 같다. 즉, 평가자가 평가한 가맹점의 품질에 대한 평가값(신호값)인

은

라고 말할 수 있는데(

=신호값,

=가맹점의 참품질,

은 가맹점의 참품질에 영향을 미치는 에러) 평가자가 예를 들어 가맹점 구석구석을 잘 살피지 못해서(에러

이 발생) 참품질과 약간 다른 평가를 내릴 수 있기 때문이다. 우수평가자는 구석구석을 잘 살펴서 참품질을 더 잘 알 수 있기 때문에 더 작은

값을 갖게 되고 따라서

의 분산, 다시 말해서 신호의 분산이 작게, 즉 신호의 정확성이 높아진다. 반면 일반평가자는 구석구석을 잘 못보기 때문에 에러

가 크게 되고

의 분산도 커지고 다시 말해서 신호의 정확성이 낮아지게 된다. signal(

) Or the accuracy of the signal is mentioned as follows. That is, the evaluation value (signal value) of the quality of the merchant evaluated by the evaluator

silver

You can say (

= Signal value,

= True quality of merchants,

Is an error that affects the true quality of the merchant.) The evaluator, for example,

Because it can make a slightly different evaluation than the true quality. Top performers are better able to see the true quality by taking a closer look so they are smaller

Value and thus

That is, the variance of the signal is small, i.e., the accuracy of the signal is high. On the other hand, the general evaluator

Becomes larger

And the accuracy of the signal is lowered.

Optimal Aggregation Algorithm Model

Given the definition of the variables model to derive the optimal aggregation algorithm according to the present invention, in order to obtain the true quality of the merchant using the optimal aggregation method, several parameter values are calculated using the data input to the data shall.

,

, 등과 신호의 정확성인

,

,

, 그리고 개인별 엄격함인

,

을 구해야 한다. (미등록평가자는 로그인하여 개인 정보를 입력하지 않으므로

을 계산하지 않는다) 개인별 엄격함은 앞서 언급한 다섯 가지 항목이 있으므로 이들 항목들의 합산으로 계산한다. 즉, 각 항목에 해당하는 계수값을 계산한다. First, the reputation of reputation, reputation,

,

, Back and signal accuracy

,

,

, And individual strictness

,

. (Unregistered reviewers do not enter personal information by logging in.

) Individual strictness is calculated by summing up these items because there are five items mentioned above. That is, the coefficient value corresponding to each item is calculated.

는 상수이고 α는 계수). (here

Is a constant and alpha is a coefficient).

이고 분산이

인 정규분포를 따르는 확률변수의 확률밀도함수는 다음과 같이 표현된다. These coefficients can be obtained by using the Maximum Likelihood Estimation Method among the statistical estimation models. In general,

And dispersion

The probability density function of a random variable following a normal distribution is expressed as follows.

와

이 주어졌을 때

의 확률밀도함수값을 계산하는 식이다. 그런데 이 식을 거꾸로 해석해서 데이터인

가 주어졌을 때 파라미터가

와

일 확률(우도, likelihood)을 나타내는 식으로 볼 수도 있다. 이 해석에 착안해서 주어진 데이터를 가장 잘 설명하는, 혹은 주어진 데이터를 만들어냈음직한 그럴 듯한 파라미터

와

를 찾는 방법이 최우추정법이다. 이 식은 실제 참값인

와

를 대입했을 경우 이 함수식이 최대값을 갖고, 다른 파라미터값을 대입하게 되면 함수값이 작아지게 된다. 따라서 데이터의 실제 참값인

와

를 찾기 위해서는 위 함수값을 최대화하는

와

를 찾으면 된다. 이 방법은 보통 수치최적화(numerical optimization) 방법에서 예를 들면 newton-raphson algorithm등과 같은 알고리즘을 사용하여 최대값을 찾아내게 된다. 위 식은 데이터가 하나 있을 경우의 식이고 데이터가 N개 있을 경우에는 벡타와 행렬을 사용하여 다음과 같은 식으로 표현된다. This expression is a parameter

Wow

When given

The probability density function of However, by reversing this equation,

Parameter is set to

Wow

It can also be seen as an expression that expresses probability (likelihood). By focusing on this interpretation, it is possible to define a plausible parameter that best describes the given data,

Wow

Is the best estimate. This expression is actually true

Wow

, The function expression has the maximum value, and if the other parameter value is substituted, the function value becomes smaller. Thus, the actual true value of the data

Wow

To maximize the above function value

Wow

. This method usually finds the maximum value in a numerical optimization method using an algorithm such as the newton-raphson algorithm. The above equation is an expression when there is one data, and when there are N data, it is represented by the following expression using a vector and a matrix.

,

는

열벡터이고,

는

행렬이며,

은 벡터

를 전치(transpose)한 것이다. here

,

The

Column vector,

The

Matrix,

Vector

. ≪ / RTI >

이 정규분포를 따른다는 가정을 사용하여 우도함수(likelihood function)을 만들고 이 우도함수를 최대화하도록 앞서 열거한 파라미터 값들을 구한다. 이때 주의할 점은 우도함수는 본래 평점 자료

를 사용한다. 하지만 실제 계산과정에서 참 파라미터 값을 계산하는데 시간이 오래 걸리고 계산 결과도 오차가 큰 현상이 발견되어서

대신에 그 차이인

데이터를 사용하여 우도함수를 추정하도록 한다. 그리고 또한

를 직접 사용하면 우도함수식이 매우 복잡해지므로 그 대신에 신호의 차이인

를 사용하여 우도함수를 계산한다. 둘 사이의 관계는 위의 식 (1)로부터

와

에 대한 전환식(자코비안 행렬)을 계산하여 얻는다. 정리하면,

이라 하고,

을 분산행렬이라고 할 때 파라미터값들은 다음 함수를 최대화(최우추정법)함으로써 구한다. The model parameters are estimated using the best estimate method as follows. In particular,

Using the assumption that it follows this normal distribution, we construct a likelihood function and obtain the parameter values listed above to maximize the likelihood function. Note that the likelihood function is originally the rating data

Lt; / RTI > However, in the actual calculation process, it takes a long time to calculate the true parameter value and the calculation result has a large error.

Instead,

Let us estimate the likelihood function using the data. likewise

Is used, the likelihood function becomes very complicated,

To calculate the likelihood function. The relationship between the two is based on the above equation (1)

Wow

(Jacobian matrix). In short,

And,

Is a dispersion matrix, the parameter values are obtained by maximizing (maximum likelihood estimation) the following function.

은 변수

를

에 대해 미분한 편미분 행렬인 자코비안(Jacobian) 행렬이다. 이 행렬 역시 위의 식(1)을 사용하여 편미분을 적용함으로써 계산한다. Where N is the total number of evaluators and N-1 is used because it is known statistically that it is better to use N-1 instead of N.

Is a variable

To

Is a Jacobian matrix that is a differential partial differential matrix for This matrix is also calculated by applying the partial derivatives using the above equation (1).

Once the parameter values have been determined by the Maximum Likelihood Method, the true quality of the merchant using the optimal aggregation method should now be calculated. The true quality due to the optimal aggregation is the expectation of the quality of the merchant when the evaluators'

의 값을 구함으로써 얻을 수 있다. 여기서

은 평가자 n까지의 정보가 모두 주어졌을 때 n번째 평가자 때의 품질의 조건부 기대치(conditional expectation)를 의미한다. 이때

와

의 에러항들이 모두 정규분포를 따르므로

의 조건부 기대치와 분산은 이변량 정규분포(bivariate normal distribution)에 대해 잘 알려진 통계학의 결과를 사용하여 다음과 같은 평균

과 분산

의 일반식을 얻을 수 있다. In other words

. here

Means the conditional expectation of the quality at the nth evaluator when all the information up to the evaluator n is given. At this time

Wow

Since all the error terms follow a normal distribution

The conditional expectation and variance of the bivariate normal distribution are calculated using the well-known statistical results for the following mean

And dispersion

Can be obtained.

이고

이다.

은 앞서 정의한 각 그룹별 신호의 정확성을 나타내는 분산이다. 우수평가자는

를 갖고, 일반평가자는

, 그리고 미등록평가자는

의 값을 가지게 된다. here

ego

to be.

Is a variance that indicates the accuracy of the signal for each group defined above. A good evaluator

, And the general evaluator

, And the undeclared evaluator

. ≪ / RTI >

은

과

의 가중평균으로 계산된다. 여기서

에는

까지의 모든 정보가 포함되어 있고,

에는 n번째 평가자의 정보만 있으므로 ‘일반적으로’

의 가중치가

의 가중치보다 더 크다. 하지만

의 가중치에는 이 포함되어 있는데, 이는 앞에서 언급한 바와 같이 n-1번째 평가자 이후부터 n번째 평가자까지의 시간 동안 가맹점의 품질이 얼마나 변했는가를 측정하는 부분이다. 따라서 만약에 n-1번째 이후 n번째 평가자까지 많은 시간이 흘러서 가맹점의 품질이 변했다면 비록

에 기존 평가자들의 모든 정보가 포함되어 있다 하더라도 이번에 새로운 n번째 평가자의 신호인

에 최근 가맹점에 대한 정보가 많이 포함되어 있을 것으로

에 대한 가중치가 상대적으로 더 커질 수도 있다. In the above equation,

silver

and

Is calculated as a weighted average of. here

There

All the information is included,

, The 'n'

Weight of

Is greater than the weight of. But

The weight of , Which is a measure of how much the quality of the merchant has changed since the n-1th evaluator to the nth evaluator as mentioned above. Therefore, if the quality of the merchant has changed since much time has passed since the n-1th to the nth evaluator,

Even if all of the information of the existing evaluators is included in the new evaluator's signal,

There should be a lot of information about recent merchants

May be relatively larger.

을 구하기 위해서는

을 알아야 한다. 이 값은 평가자의 평점

이 주어지고, 주어진 평가자의 프로파일에 따라

이 계산되어졌을 때 다음 두 식(1)(2)의 연립방정식에 의해 다음과 같이 계산한다. In the above equation

To get

. This value is the evaluator's rating

Given the profile of the given evaluator

Is calculated by the simultaneous equations of the following two equations (1) and (2) as follows.

(2)

(2)

(3)

(3)

In other words,

을 좌변으로 넘기면 다음과 같은 식을 얻게 된다. , And this equation is rearranged

, The following equation is obtained.

을 사용하여 최적집계 방식에 의한 평균평점을 계산할 수가 있다. So-called

Can be used to calculate the average rating by the optimal aggregation method.

The payment processing unit 370 charges the merchant through the result of the use of the real time publicity and event providing service provided to the customer or the merchant in the management server 30 and processes the payment.

The storage unit 380 stores the membership information analyzed by the customer information analysis unit 330, the review information analysis unit 340 and the search information analysis unit 350 provided from the customer terminal 10 and the merchant terminal 20 A ROM, an EEPROM (Electrically Erasable and Programmable Read Only) device, and the like, for storing various analysis information, review analysis information calculated by the review analysis information calculation unit 360, and payment information processed by the payment processing unit 370. [ Memory, an SDRAM (Synchronous Dynamic Random Access Memory), or an HDD, but the type of the storage unit is not limited thereto.

The storage unit 380 includes an affiliate store information database 381 for storing information related to an affiliate shop, a customer information DB 382 for storing information about the customer, an analysis information DB 383 for storing various analysis information, And an application information DB 384 for storing publicity and event management application information provided to the merchant terminal 10 or the merchant terminal 20.

As described above, the present invention is not limited to the above-described embodiments, but can be applied to the real-time publicity and event providing system according to the present invention. It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

10: customer terminal 20: merchant terminal
30: management server 310:
320: member authentication unit 330: customer information analysis unit
340: review information analysis unit 350: search information analysis unit
360: Review analysis information calculation unit 370: Payment processing unit
380: storage unit 381: merchant information DB
382: customer information DB 383: analysis information DB
384: Application information DB

Claims (8)

A publicity and event management application is installed and searches for various merchant store information based on current location information or local information set by the customer on the connected web by the execution of the publicity and event management application, And a customer terminal receiving the event information, registering the latest visit to the merchant, and receiving review rating information about the latest visit to the merchant;
A publicity and event management application is installed and registers the merchant's business card, publicity and event information on the connected web by the execution of the publicity and event management application, and stores customer analysis information on customers visiting or interested in the affiliated store, The merchant terminal receiving the review analysis information on the merchant visit history information and the review rating information and the search analysis information on the merchant information and the search information on the publicity and event information; And
And provides the promotion and event management application, the merchant information, the promotion and event information, the review rating information, the customer analysis information, the review analysis information, and the search analysis information in cooperation with the customer terminal and the merchant terminal, And a management server for calculating the review analysis information through a predetermined review analysis algorithm,
Wherein the review analysis information calculated by the management server includes at least one of a franchise conditional expectation value calculated by a review rating calculation engine in consideration of a weight based on the group, a strictness of the individual, an individual rating of the evaluator, Include review rating information,
The merchant conditional expectation value is an expectation value at the time of evaluation of the n-th customer when the information up to the customer n performing the review evaluation is given,

Lt; / RTI >
At this time, the signal value observed by the nth customer

)silver

(here,

Is an individual's strictness variable,

Individual ratings,

Weight,

,

Is distributed,

Is the expectation of the customer's evaluation)
Lt; / RTI >

The individual's severity variable (

) Is the number of evaluations that the evaluator has previously evaluated (

), The number of evaluations submitted by the evaluator over a period of time (

), The difference in the matching between the evaluator and the participating merchant

), An indicator of visit diversity between the evaluator and the merchant

) And the time interval during which the evaluator visited the merchant

) The arithmetic sum of the exponent

(here

Is a constant, and? Is a coefficient)
Lt; / RTI >

The individual rating

silver

(here,

Is an individual's strictness variable,

,

Weight,

A signal input value,

Is the individual rating information about ratings that people have visited

) And the signal input value (

) Is a predicted value for the true quality of the franchisee,
Lt; / RTI >

The true quality of the merchant at time t

)silver,

(here

, Which is a variable affecting the quality, including the service mistakes of the waiters at the restaurant, changes in employees' service minds according to the management philosophy of the new merchant manager,
Lt; / RTI >

The evaluator's observation signal observed by the nth evaluator (

),

(here

Lt; / RTI >

Is the true quality value of the merchant to be evaluated,

Is a variable that affects the signal value)
Lt; / RTI >

Wherein the merchant terminal is connected to the management server to set an inquiry date and then receives information on a customer around the shop using the coupon at a store within 500 meters of the shop through the GPS within the relevant period of time, Delivery system. The method according to claim 1,
The customer analysis information may include at least one of peripheral customer information located within a predetermined radius of the merchant store, merchant card information, merchant information, interest customer information obtained by searching for promotional and event information, The average information of the surrounding customers, the interested customers, and the purchasing customers, the number of the customers who have viewed or downloaded the publicity and event information within the predetermined period of time, A publicity / event status, and performance information. The method according to claim 1,
The review analysis information includes at least one of the number of registered items of the last visit to the merchant shop, the review rating information, the daily review rating trend information, the gender or age review registration rate information, the review registration ratio information per visit type, And a star rating information. The method according to claim 1,
Wherein the customer terminal arranges the searched various merchant information in descending order based on the location information or the local information or sorts and displays the searched merchant information in the order of review rating based on the review rating information of the merchants. . The method according to claim 1,
Wherein the management server provides event information of the franchisees located within a predetermined radius in real time based on the location information of the customer. delete delete delete

Images