KR102080344B1 - An apparatus and method for detecting fraud evaluation of online contents that distinguish heavy user from fraud user - Google Patents

Abstract

The present invention relates to a system, apparatus, method and program for detecting false evaluation information of online content related to a specific service provider, a specific service, a text, a photo, a book, a store including a restaurant such as a restaurant or a cafe, a movie, and the like. To this end, the online content is uploaded and configured in a platform service connected to the user's client so that the user can access and enter evaluation information on the online content, and for the detection of false evaluation information of the online content, A determination assistance information generation module for generating determination assistance information corresponding to a specific user; And a platform service configured to be connected with the determination assistance information generation module to receive the determination assistance information, and determine whether false evaluation information is input to the online content or whether the user inputs the false evaluation information based on the determination assistance information. False evaluation information determination module; false evaluation information detection apparatus can be provided, including. According to this, there is an effect that the false evaluation information for the online content can be automatically detected in the platform service.

Description

An apparatus and method for detecting fraud evaluation of online contents that distinguish heavy user from fraud user}

The present invention relates to a system, apparatus, method and program for detecting false evaluation information of online content that distinguishes a heavy user from a false evaluation user by using period information from a subscription date to an input date of evaluation information, and more specifically, a specific service provider. , A system, an apparatus, a method, and a program for detecting false evaluation information of online content related to a specific service, an article, a photo, a book, a restaurant including a restaurant or a cafe, a movie, and the like.

Recently, a lot of platform services of a double-sided network have been developed on the Internet where service providers who upload online content to acquire users and users who want to use such online content gather. These platform services include, for example, open markets such as ebay and Amazon.com, shared services such as Uber and Airbnb, and Yelp and Mango Plates. It may mean a restaurant information service such as (Mangoplate). In particular, smartphones have made it easier to become a service provider than the traditional web era, and users' access to online content through platform services on their smartphones has been significantly improved. This has led to an explosion of online content and platform services on the Internet. Users can use platform services such as the field where online contents are accumulated through mobile applications, mobile web, etc. to obtain enough information as they want, thereby helping them to select, make decisions or make judgments.

In particular, the user's decision is to provide rating information, which is information about ratings of other users using ratings or comments on specific online content. Providing such evaluation information has been greatly effected on eBay, and is mostly used to improve the user's experience in platform services such as Amazon.com, Uber, Airbnb, and Yelp.

This rating information may include rating information, comment information (including user review information and user comment information) about certain online content, crush information such as 'like' or 'bad', shareability information, recommendation information, and inquiry information. Can be. The user is provided with at least one online content uploaded to the platform service, and the evaluation information on the online content is very helpful for the user's decision, which leads to an improvement of the user experience.

However, if the evaluation information is provided to the artisan platform service where the online contents are gathered, the platform service itself is exposed to the risk of becoming a lemon market by the false evaluation of a specific user such as a malicious evaluation of a competitor or a favorable evaluation of a related person. . False evaluation information of a specific user threatens the ecosystem of the online content by losing trust in the evaluation information of the online content of the user who receives the online content.

Korea Patent Publication 10-2012-0101188 Korea Patent Publication 10-2013-0113038

However, since the rating information for the existing online content is simply performed as an arithmetic mean of all ratings recorded in one online content, it is difficult to detect false rating information of a specific user. In addition, the provision of comment information and crush information on the existing online contents has not been actually carried out the detection of false evaluation information, only the extent to which the 'report' function that users report themselves is used.

Therefore, in the past, in order to detect false rating information of a specific user, for each online content, a person manually determines whether the rating information, comment information, and affinity information of the online content is included in the calculation of the rating, The user decides whether to print or not.

As described above, the method of capturing false evaluation information manually performed by humans is a lemon marketization of platform service that causes huge costs (especially manpower costs), is not fast and sustainable, and the online content and traffic is exploding. It was not enough to prevent.

In addition, as the number of heavy users such as holics in a particular store surges, it is difficult to distinguish and classify whether the user who inputs specific evaluation information is a heavy user or a false evaluation user who inputs false evaluation information. Traditional methods were still inadequate. This is because usage behaviors in the service, such as the frequency of input of evaluation information of the heavy user and the false evaluation user, are very similar.

Accordingly, an object of the present invention is to evaluate from a subscription to quickly and automatically detect false evaluation information of online content related to a specific service provider, a specific service, a text, a photo, a book, a store including a restaurant or a restaurant such as a cafe, a movie, and the like. The present invention provides a system, apparatus, method, and program for detecting false evaluation information of online content that distinguishes a heavy user from a false evaluation user by using the period information up to the date of inputting the information.

Hereinafter, specific means for achieving the object of the present invention will be described.

An object of the present invention is configured in a platform service connected to the client of the user so that the online content is uploaded and the user can access and input the evaluation information for the online content, the detection of false evaluation information of the online content A decision assistance information generation module for generating decision assistance information corresponding to a specific online content or a specific user; And configured to be connected to the determination assistance information generation module to receive the determination assistance information, and to determine whether false evaluation information is input to the online content based on the determination assistance information, or that the user makes false evaluation information. False evaluation information determining module for determining whether or not input, wherein the determination assistance information, the frequency of at least one user's evaluation information input to each online content or each user input to at least one online content; Frequency information, which is information on the frequency of the evaluation information, and information on how long the user has passed until the user inputs the problem information after joining, or usage period information which is information on the number of evaluation information inputs from the membership to the present. False evaluation information, characterized in that it comprises a It can be achieved by providing to the device.

Another object of the present invention is a platform service that is connected to the client of the user so that the online content is uploaded and the user can access and input the evaluation information for the online content, the detection of false evaluation information of the online content A decision aid information generation module for generating decision aid information corresponding to the specific online content; And a false evaluation information determining module, configured in the platform service, connected to the determining assistance information generating module to receive the determining assistance information, and determining whether false evaluation information is input to the online content based on the determining assistance information. Wherein the determination assistance information is frequency information that is information about a frequency of at least one user's evaluation information input to each online content or a frequency of each user's evaluation information input to at least one online content. And usage period information which is information on how long the user has passed until the user inputs the evaluation information after signing up or information about the number of evaluation information inputs from the date of joining to the present, wherein the false evaluation information determination module includes: Using the period information as a weight of the frequency information It can be achieved by providing a false evaluation information detection device, characterized in that.

Another object of the present invention is to provide a platform service connected to the client of the user so that the online content is uploaded and the user can access and input the evaluation information for the online content, and to detect the false evaluation information of the online content. An evaluation information receiving step of performing, by the false evaluation information detection device, evaluation information of a specific user regarding a specific online content at the client; A determination assistance information generating step of generating, by the determination assistance information generating module which is one component of the false evaluation information detection device, the determination assistance information for the online content or the user; And a false evaluation information determining module, which is one component of the false evaluation information detecting device, determines whether false evaluation information is input to the online content or whether the user inputs false evaluation information based on the determination assistance information. An evaluation information determination step, wherein the determination assistance information is related to a frequency of at least one user's evaluation information input to each online content or a frequency of evaluation information each user inputs to at least one online content. Frequency information, which is information, and usage period information which is information on how long a period has passed until the user inputs problem information after joining, or information on the number of evaluation information inputs from the membership to the present. This can be achieved by providing a method for capturing false assessment information.

Another object of the present invention is to provide a platform service connected to the client of the user so that the online content is uploaded and the user can access and input the evaluation information for the online content, and to detect the false evaluation information of the online content. An evaluation information receiving step of performing, by the false evaluation information detection device, evaluation information of a specific user regarding a specific online content at the client; A determination assistance information generating step of generating, by the determination assistance information generating module, which is one component of the false evaluation information detection device, determination assistance information for the online content; And a false evaluation information determination step of determining, by the false evaluation information determination module, which is one component of the false evaluation information detection device, whether false evaluation information is input to the online content based on the determination assistance information. The supplementary information includes frequency information that is information about a frequency of at least one user's evaluation information input to each online content or a frequency of evaluation information each user inputs to at least one online content, and after the user subscribes. It includes information on how long the period has passed until the input of the evaluation information in question, or information on usage period, which is information on the number of evaluation information inputs from the membership to the present, wherein the false evaluation information determination module is configured to read the usage period information. False evaluation information, characterized in that used as a weight of the frequency information It can be achieved by providing a method to.

Another object of the present invention is to provide a platform service connected to the client of the user so that the online content is uploaded and the user can access and input the evaluation information for the online content, and to detect the false evaluation information of the online content. An evaluation information receiving step of performing, by the false evaluation information detection device, evaluation information of a specific user regarding a specific online content at the client; A determination assistance information generating step of generating, by the determination assistance information generating module which is one component of the false evaluation information detection device, the determination assistance information for the online content or the user; And a false evaluation information determining module, which is one component of the false evaluation information detecting device, determines whether false evaluation information is input to the online content or whether the user inputs false evaluation information based on the determination assistance information. An evaluation information determination step, wherein the determination assistance information is related to a frequency of at least one user's evaluation information input to each online content or a frequency of evaluation information each user inputs to at least one online content. Frequency information, which is information, and usage period information which is information on how long a period has passed until the user inputs problem information after joining, or information on the number of evaluation information inputs from the membership to the present. To perform computer-based methods for capturing false assessment information. This can be accomplished by providing a program stored on the lock medium.

As described above, the present invention has the following effects.

First, according to an embodiment of the present invention, in the situation where a heavy user and a false evaluation user having similar usage behaviors in terms of frequency of input of evaluation information exist simultaneously, false evaluation information of online contents of the platform service is automatically detected. The effect is that it becomes possible.

Secondly, according to one embodiment of the present invention, since false detection information on the online content is automatically detected in the platform service, it is possible to reduce manpower costs.

Third, when using the Herfindahl-Hirschman Index (HHI) index in accordance with an embodiment of the present invention, there is an effect that can automatically detect the false evaluation information quickly with little information. It is very important in terms of usability of data analysis to detect false evaluation information with high probability by using less information in the detection of false evaluation information.

The following drawings, which are attached to this specification, illustrate exemplary embodiments of the present invention, and together with the detailed description thereof, serve to further understand the technical spirit of the present invention. It should not be interpreted.
1 is a block diagram showing a system for detecting false evaluation information of online content according to an embodiment of the present invention;
2 is a block diagram illustrating a platform service according to an embodiment of the present invention;
3 is a block diagram showing a false evaluation information detection device 107 according to an embodiment of the present invention,
4 illustrates a method for capturing false evaluation information of online content according to an embodiment of the present invention.
5 shows the actual reviews of Table 1 entered into the online content of the restaurant 902,
FIG. 6 shows some of the actual reviews of Table 2 entered into the online content of restaurant 137445,
FIG. 7 illustrates a distribution of a ratio of the number of restaurants / total comment inputs for inputting comments on 316 random samples of actual user data.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, in describing in detail the operating principle of the preferred embodiment of the present invention, if it is determined that the detailed description of the related known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted.

In addition, the same reference numerals are used for parts having similar functions and functions throughout the drawings. Throughout the specification, when a particular part is connected to another part, this includes not only the case where it is directly connected, but also the case where it is indirectly connected with another element in between. In addition, the inclusion of a specific component does not exclude other components unless specifically stated otherwise, it means that may further include other components.

In the present invention, the 'online content' may refer to information related to a profile of a specific service provider, a profile of a specific service, a text, a photo, a book, a shop including a restaurant such as a restaurant or a cafe, a movie, and the like. Users can browse and browse these online content through applications on the PC or smartphone or the web. Users can provide rating information for these online content.

In the present invention, 'rating information' is a rating information, comment information (including user review information, user comment information) for a particular online content, crush information such as 'like' or 'bad', sharing information, recommendation information, inquiry It may mean information provided by the user about the specific online content including the information. Such rating information may refer to any information that is rated, rated, suggested, liked, disliked, recommended, or disliked for specific online content.

In the present invention, the 'platform service' may mean an online double-sided network of service providers trying to acquire users by uploading online contents and users who want to use such online contents.

False information dissemination system of online content

Regarding a False Evaluation Information Detection System of Online Content, FIG. 1 is a block diagram illustrating a false evaluation information detection system of online content according to an embodiment of the present invention. As shown in FIG. 1, the system for detecting false information of online content according to an embodiment of the present invention 100 may include a client 101, a network 102, and a platform service 103.

The client 101 may include at least one or more client devices, for example, one or more smartphones, mobile devices such as tablets, laptops or desktop PCs. The user browses / searches online content or inputs evaluation information of specific online content through the client 101.

Network 102 may include connections that provide a wireless or wired communication connection between data processing systems, computers, servers, mobile devices. The network 102 according to an embodiment of the present invention may be configured as a connection for providing a wired or wireless communication connection between the client 101 and the platform service 103. The network 102 according to an embodiment of the present invention may include a World Wide Collection of networks and gateways using TCP / IP to communicate with each other, an intranet, a LAN, and a WAN, and may mean the Internet.

The platform service 103 is connected to the client 101 through the network 102 to transmit online content to the client 101, and receives the evaluation information by the client 101. ) May be a web page that is available to a user through an application of a smartphone / Mac OS or a web browser of the smartphone / PC. Platform service 103 according to an embodiment of the present invention may include various computers, servers, databases, routers, switches, connections and other devices.

2 is a block diagram illustrating a platform service according to an embodiment of the present invention. As shown in FIG. 2, the platform service 103 according to an embodiment of the present invention includes an online content providing module 104, an online content database 105, an evaluation information receiving module 106, and a false evaluation information detecting device ( 107).

The online content providing module 104 is configured to receive online content from the online content database 105 and provide it to the client 101. The online content providing module 104 allows the client 101 to browse / browse online content in the platform service 103.

The online content database 105 may refer to a configuration for storing online content provided by a user through the client 101 or input by a platform service provider.

The evaluation information receiving module 106 is a module that receives the evaluation information from the client 101 and delivers the evaluation information to the false evaluation information detection device 107.

The false evaluation information detection device 107 is configured to receive evaluation information of a specific user in the evaluation information receiving module 106 and detect false evaluation information. 3 is a block diagram showing a false evaluation information detection device 107 according to an embodiment of the present invention. As shown in FIG. 3, the apparatus for detecting false evaluation information 107 according to an embodiment of the present invention includes a determination assistance information generating module 108, an evaluation information database 109, a false evaluation information determining module 110, It may include a penalty processing module 111.

The determination assistance information generation module 108 is a module that generates the determination assistance information that facilitates the evaluation by processing the evaluation information received in order to classify or predict the evaluation information as the false evaluation information in the false evaluation information determination module 110. . The determination assistance information generating module 108 receives the evaluation information and the user's information corresponding to the evaluation information in the evaluation information receiving module 106 and facilitates the evaluation information database 109 or the false evaluation information determination module 110. The determination assistance information obtained by processing the evaluation information is transmitted for evaluation.

The evaluation information database 109 is connected to the determination assistance information generating module 108, the evaluation information receiving module 106, and the penalty processing module 111 to store the determination assistance information, the evaluation information, and the penalty processing information. In particular, the penalty processing information may be delivered to the online content providing module 104 and the online content database 105.

The false evaluation information determination module 110 is connected to the determination assistance information generating module 108 and the evaluation information receiving module 106 to receive determination assistance information and evaluation information, and analyzes it to determine whether false evaluation information is present. . The false evaluation information may mean that the corresponding evaluation information is false evaluation information, or may indicate that false evaluation information is input to the online content. When the evaluation information of a specific user is determined to be false evaluation information, the corresponding evaluation information and corresponding user information may be transmitted to the penalty processing module 111.

The penalty processing module 111 is a module that receives the evaluation information determined as the false evaluation information and the corresponding user information in the false evaluation information determination module 110 and generates penalty processing information that has been penalized thereto.

How to Catch False Assessment Information of Online Content

Regarding a method for detecting false evaluation information of online content, FIG. 4 illustrates a method for detecting false evaluation information of online content according to an embodiment of the present invention. As shown in FIG. 4, in the method for detecting false evaluation information of online content according to an embodiment of the present invention, the evaluation information receiving step S10, the determination assistance information generating step S20, and the false evaluation information determining step S30. ), It may include a penalty processing step (S40).

The evaluation information receiving step S10 is a step in which the evaluation information receiving module 106 receives the evaluation information transmitted from the client 101.

The determination assistance information generation step S20 transmits the evaluation information received by the evaluation information receiving module 106 to the false evaluation information detection device 107, and generates determination assistance information that is one component of the false evaluation information detection device 107. The determination assistance information is generated based on the evaluation information, the user information, and the like received by the module 108.

False evaluation information determination step (S30) is based on the determination assistance information generated in the determination assistance information generation module 108, the false evaluation information determination module 110 determines that the evaluation information is false evaluation information, or to the specific online content It is a step of determining that the evaluation information about the likelihood is false evaluation information. The false evaluation information may be classified into positive false evaluation information to increase the evaluation of the online content, and negative false evaluation information to lower the evaluation of the online content.

Penalty processing step (S40), if the false evaluation information determination module 110 determines that at least some of the evaluation information or the evaluation information of the online content is false evaluation information, the user, the evaluation information or of the online content A penalty process is performed on at least some of the evaluation information. In the case of positive false evaluation information, the penalty processing may include suspending or canceling the account of the user who transmitted the evaluation information, and giving a weight for the evaluation information to 0 (excluding the sum of evaluation information) or negative. In the case of negative false evaluation information, the penalty processing may include suspending or canceling the account of the user who sent the evaluation information, and giving the weight of the evaluation information as 0 (excluding the sum of evaluation information) or a positive number.

According to an embodiment of the present invention, a method for capturing false evaluation information of online content may include a program stored in a recording medium such as an HDD, an SSD, a CD-ROM, a USB, and the like, which is executed on a computer.

Example

Hereinafter, a specific embodiment of a system, apparatus, method, and program for detecting false evaluation information of online content according to an embodiment of the present invention will be described.

Regarding the determination assistance information, the determination assistance information that may be generated by the determination assistance information generating module 108 which is one component of the false evaluation information detection system of the online content according to one embodiment of the present invention is comment information (comment information). , Textual mining information of the review information, etc.), frequency information of the evaluation information, network information between users, and usage behavior information of the user.

(1) Comment / Review Information

Regarding providing the text mining information as the judgment aid information, the textual mining of the comment / review information among the evaluation information enables the emotional analysis of the positive and negative degree of the comment / review information, and the comment / Through the emotion analysis result of the review information, the frequency information of the comment / review of a specific emotion may be provided to the false evaluation information determination module 110 as decision assistance information.

Alternatively, the comment / review information showing a similar pattern may be determined as the false evaluation information by assigning a plurality of attributes to the comment / review information previously determined as false evaluation information. One can be used. In this case, when the existing false evaluation information and the input evaluation information are classified into the same group, the classification information may be provided to the false evaluation information determination module 110 as decision assistance information. Attributes that may be included in the comment / review information may include a length of the comment / review, emotion analysis information of the comment / review, a time taken for input of the comment / review, and whether to use an emoticon in the comment / review.

Alternatively, various attributes may be assigned to existing evaluation information and newly received evaluation information and clustered, and the probability of false evaluation information of the cluster including the corresponding evaluation information may be used. To this end, one of several models of Unsupervised Machine Learning can be utilized. According to this, when specific evaluation information is received, the probability value of the false evaluation information of the cluster to which the corresponding evaluation information belongs may be provided to the false evaluation information determination module 110 as decision assistance information. In this case, the false evaluation information determination module 110 may determine the false evaluation information by setting a cut-off value of a probability value of the false evaluation information of the cluster to which the corresponding evaluation information belongs.

Alternatively, the frequency / presence of a specific word may be grasped through text mining of the comment / review information of the evaluation information, and provided to the false evaluation information determination module 110 as decision assistance information. For example, a restaurant owner's comment information about online content related to his or her own restaurant often contains words that are not used by ordinary users, such as "guarantee" and "guarantee," "I often use reverence without knowing it." Information about a specific expression method of such a comment / review may be provided to the false evaluation information determination module 11 as decision assistance information.

(2) Frequency information of evaluation information

In connection with providing the frequency information of the evaluation information as the determination assistance information, when the frequency information of the evaluation information is provided as the determination assistance information in the false evaluation information determination module 110, if the frequency of the evaluation information is rapidly increased, the corresponding online At least some evaluation information about the content may be determined as false evaluation information. For example, the normal distribution corresponding to the frequency of the evaluation information may be obtained, and the position of the normal distribution may be analyzed with respect to how much evaluation information is output for each online content. If the frequency of evaluation information changes drastically, the position in the normal distribution is changed. According to an embodiment of the present invention, such frequency information may be an example of the number of evaluation information for each date. According to another embodiment of the present invention, statistical distributions such as probability distribution, t distribution, and chi distribution may be used instead of the normal distribution.

(3) network information between users

With regard to providing network information between users as decision assistance information, network information may refer to social network information between users. For example, if location information is very similar among users who input evaluation information into specific online content, the information may be provided to the false evaluation information determination module 110 as decision assistance information. Alternatively, when all users who enter rating information in a particular online content are friends of a social network or when there are many mutual friends of social networks between users who input the information, the information is false as evaluation aid information. The information determination module 110 may be provided.

By the way, the technique of the false evaluation users inputting false evaluation information is evolving day by day. As detection and penalty regarding false evaluation information continue, false evaluation users do not input all false evaluation information in a short period of time, but enter false evaluation information into specific online content such as a restaurant over a long period of time. In such a situation, the above social network friend information (connection information) may be used as one of the most suitable problem solving methods. According to an embodiment of the present invention, the number of connection information between users who input evaluation information into specific online content may be used to detect false evaluation users. According to this, remarkable accuracy is generated in the detection of the false evaluation user.

As an example of online content for a restaurant, assume that there are A, B, C, D, and E with actual rating information for specific restaurant information, and a, b, c, d, and ㅁ with false rating information. do. At this time, there are two connections among the party evaluation users A, B, C, D, and E that are connected in a friend relationship, the maximum number of possible connections is 10, the number of friends of the whole network is 150, and the distinction of the whole network is distinct. Assume that the number of friends is 140 (the number of nonoverlapping friends). In addition, there are 8 connections with friends among the false rating users a, b, c, d, and ㅁ, the maximum number of possible connections is 10, the number of friends in the entire network is 30, the distinction in the whole network Assume that you have 29 friends. As the relationship between users who input the evaluation information becomes closer, it can be assumed that there is one group that inputs the false evaluation information.

In the above case, in order to detect a, b, c, d, and k as a false evaluation user, the ratio of the number of friendship relations between the users who input the evaluation information and the maximum possible number of connections is determined by the assistant assistance information. It can be used as. Alternatively, the ratio of the number of friend relation connections between users who input the evaluation information and the number of distinct friends of the entire network may be used as the determination assistance information. Alternatively, the ratio of the number of friend relation connections between users who input the evaluation information and the total number of friends of the entire network may be used as the determination assistance information. Alternatively, a ratio of the total number of friends of the entire network and the distinct number of friends of the entire network among users who input the evaluation information may be used as the determination assistance information.

To this end, according to an embodiment of the present invention, the decision assistance information generation module 108 has a maximum number of connection relations and the maximum number of possible connections between users who have input the evaluation information for the user group uploading the evaluation information to the specific online content. The ratio of the number of friends, the number of friendship connections between users who entered evaluation information, and the ratio of the number of distinct friends of the whole network, the total number of friendship connections between users who entered evaluation information, and the total number of networks At least one of the ratio of the number of friends or the ratio of the total number of friends of the entire network and the distinct number of friends of the entire network among users who input the evaluation information may be generated as the determination assistance information. As a method of determining whether or not the false evaluation information is based on the determination assistance information, a method using a distribution of ratio or a distribution of HHI may be used.

According to an embodiment of the present invention, the ratio of the number of friend-connection connections and the maximum possible number of connections between users who have input the evaluation information with respect to a group of users who have uploaded the evaluation information to specific online content is determined as the assistance information. In this case, the false evaluation information input group composed of several users who want to enter the false evaluation information can be detected even when the false evaluation information is input to specific online content over a certain period of time, rather than one day. .

(4) user behavior information

Regarding providing the user's behavior information as decision aid information, within a few seconds of entering the platform service, whether the user enters the evaluation information, browses other online content, enters the evaluation information, and the evaluation information in the session of the day Input information, which page is inputted into the evaluation information input, and the like can be analyzed and used as the determination assistance information. According to this, when specific evaluation information is received, behavior information of the user who inputs the evaluation information may be provided to the false evaluation information determination module 110 as decision assistance information. Or, for example, if a user enters rating information while like all of the reviews of the restaurant, the user enters positive rating information for a particular restaurant while entering negative rating information for restaurants of the same type nearby. The user behavior information, such as a case where a user promises compensation by leaving a contact number such as a phone number / email in a comment of other users who actively input evaluation information, may be provided to the false evaluation information determination module 110 as decision assistance information. have.

(5) time period from signup to upload specific rating information

In particular, as the number of heavy users such as holics in a specific store increases rapidly, it is difficult to distinguish and classify whether a user who inputs specific evaluation information is a heavy user or a false evaluation user who inputs false evaluation information. Traditional methods were still inadequate. This is because usage behaviors in the service, such as the frequency of input of evaluation information of the heavy user and the false evaluation user, are very similar. However, according to an embodiment of the present invention, when a specific user joins the platform service and uses information on a period until uploading specific evaluation information as the determination assistance information, the distinction between the heavy user and the false evaluation user is determined. The effect which becomes easily possible is produced. According to an embodiment of the present invention, when the frequency information of the evaluation information on the specific online content is used as the determination assistant information, the information on the joining and the period until the input date of the evaluation information may also be used as the determination assistant information. Alternatively, the information on the number of reviews from the subscription can also have the same effect as the information on the period from the date of entry of the subscription and the corresponding evaluation information.

 For example, if it is assumed that a statistical result is obtained that the shorter the period from the subscription to the upload of specific evaluation information, the higher the probability of being a false evaluation user, the technical feature according to an embodiment of the present invention may be performed as follows. have. If the frequency of the toxic evaluation information is high in a specific period in the distribution composed of the frequency information of the evaluation information on the specific online content, if only the frequency information is used, the period is unconditionally determined as the period in which the false evaluation information is inputted. . In this case, a problem arises in that evaluation information by the heavy users may be mistaken for false evaluation information. In addition, since the administrator must directly confirm that the evaluation information caught by the false evaluation information detection device is not false evaluation information, a problem occurs in that human resources are consumed.

However, according to an embodiment of the present invention, when the information on the period from the subscription to the upload of the evaluation information together with the frequency information of the evaluation information on the specific online content is used as the decision assistance information, The rate of mistaken evaluation information for false evaluation information may be lowered. That is, according to the determination assistance information, the ratio of the evaluation information by the heavy users among the false evaluation information caught by the false evaluation information detection apparatus according to an embodiment of the present invention can be reduced.

To this end, according to an embodiment of the present invention, in the determination assistance information generation module 108, as the determination assistance information, information about how long has passed until the user uploads the evaluation information after joining the evaluation information for each evaluation information; Can be generated.

According to another embodiment of the present invention, the number of reviews / comments from the date of joining to the present may be used as the determination assistance information, and the same effect as described above may be obtained.

EXAMPLE

Tables 1 and 2 below are examples for demonstrating the effect generated when the period from joining to uploading specific evaluation information is used as the determination assistant information. Based on the Applicant's actual data, it is immediately available on the Mango Plate, Applicant's actual application service. Some nicknames have been deleted for privacy reasons.

Restaurant ID Name of Restaurant User ID nickname Joined Reviewed by Total review 902 Column *** 556755 n ***** 2016-05-16 4:38:43 AM 2016-05-16 4:58:01 AM One 902 Column *** 556763 Kim***** 2016-05-16 4:54:29 AM 2016-05-16 5:03:00 AM One 902 Column *** 556786 foil***** 2016-05-16 4:41:07 AM 2016-05-16 5:11:59 AM One 902 Column *** 556785 tablet***** 2016-05-16 6:17:23 AM 2016-05-16 7:09:40 AM One 902 Column *** 556802 Five***** 2016-05-16 7:03:27 AM 2016-05-16 7:25:09 AM One 902 Column *** 557005 J ***** 2016-05-16 2:17:43 PM 2016-05-18 2:50:43 PM One

Table 1 above shows a case in which six reviews are written within 7 days for a specific restaurant with restaurant ID 902. If you review the frequency of reviews based on the date of the review, you'll see 5 reviews written on May 16, 2016 and 1 review written on May 18, 2016. The frequency of reviews of restaurant ID 902 is expressed as HHI described below. HHI ₉₀₂ = (1/6) ^ 2 + (5/6) ^ 2 = 26/36

Restaurant ID Name of Restaurant ID nickname Joined Reviewed by Total review 137445 circa******** 27907 L ***** 2014-08-01 4:18:26 AM 2016-01-25 4:31:46 AM 279 137445 circa******** 52544 A ***** 2015-01-05 9:50:19 AM 2016-01-26 7:50:14 AM 245 137445 circa******** 435605 One***** 2015-11-14 11:16:59 AM 2016-01-26 2:36:45 PM 107 137445 circa******** 114662 mother***** 2015-04-27 2:37:10 AM 2016-01-26 2:44:13 PM 498 137445 circa******** 65460 K ***** 2015-03-02 10:04:04 AM 2016-01-26 3:48:08 PM 222 137445 circa******** 68 Z ***** 2012-11-21 9:25:29 PM 2016-01-29 1:03:08 PM 571

Table 2 above shows a case where six reviews are written within 7 days for a specific restaurant with restaurant ID 137445. If you review the frequency of reviews based on the date of the review, you'll see 1 review written on January 25, 2016, 4 reviews written on January 26, 2016, and 1 review written on January 29, 2016. The frequency of reviews of restaurant ID 137445 is expressed as HHI described below. HHI ₁₃₇₄₄₅ = (1/6) ^ 2 + (4/6) ^ 2 + (1/6) ^ 2 = 18/36

In the above, the values of HHI ₉₀₂ and HHI ₁₃₇₄₄₅ are quite high, and they are similar to each other. Accordingly, when the false information is determined by generating the frequency information of the review as the determination assistant information according to an embodiment of the present invention, both the restaurant 902 and the restaurant 137445 may be determined as online content in which the false evaluation information is input. All restaurants can be classified for visual analysis. The frequency information alone does not require the manager to see all the restaurants' information, resulting in significant human resource savings.

However, when reviewing the actual reviews, it can be seen that restaurant 902 is closer to online content with false rating information than restaurant 137445.

5 shows the actual review of Table 1 entered into the online content of the restaurant 902. As shown in FIG. 5, the actual review of Table 1 includes artificial and artificial phrases such as "it is unforgettable once you eat it," "^^", and exposes the name of the restaurant "Lanna Thai". He also shows the name of the menu precisely and artificially, such as "Champapon Curry" and actively recommends it to other people, such as "We recommend highly recommended restaurants." In addition, it can be confirmed that the level of the picture taken is professional. Such reviews or comments can be visually determined with false evaluation information.

6 shows some of the actual reviews of Table 2 entered into the online content of restaurant 137445. As shown in Figure 6, the actual review of Table 2 is characterized by not completing the sentence to the end, all of the pros and cons of the restaurant is described, generally have a sinemic problem, the level of the picture taken You can see that the public level. Such reviews or comments can be visually determined to be not false evaluation information.

After all, in the case of using the review frequency information as the determination assistance information according to an embodiment of the present invention, a group including an online content group (restaurant 902, restaurant 137445, etc.) having a high probability that false evaluation information is input among numerous restaurant online contents. ) Can be determined. However, online content groups that are more likely to have false ratings entered may actually have false ratings entered (correct decision, e.g. restaurant 902), and otherwise (error verdict, e.g. restaurant 137445). There is).

According to an embodiment of the present invention, when the information on the period from the subscription to the upload of the evaluation information together with the frequency information of the evaluation information on the specific online content is used as the determination assistance information, the evaluation information by the heavy users The rate of mistaken for false evaluation information can be lowered.

For restaurant 902, if you review the frequency of reviews based on the date of review, you can see above that all 5 reviews were written within 24 hours of joining on May 16, 2016, and 1 review was written after 24 hours on May 18, 2016. . The HHI value of the review frequency of the restaurant ID 902 was 26/36 above, but if the review written within 24 hours from the membership is regarded as false evaluation information and the evaluation information is nulled rather than the false evaluation information, the following HHI value is obtained. Hereinafter, the HHI value in which the evaluation information is invalidated rather than the false evaluation information is referred to as HHI '.

HHI ' ₉₀₂ = 0 * (1/6) ^ 2 + 1 * (5/6) ^ 2 = 25/36

In the case of the restaurant 902, when the frequency information of the evaluation information and the period information from the joining date to the evaluation date (HHI ') are still used, a high HHI value may still be output, and thus the false evaluation information may be determined as online content.

For restaurant 137445, if you review the frequency of reviews based on the date of review, one review on January 25, 2016, four reviews on January 26, 2016, and one review on January 29, 2016, will be available after 24 hours (especially more than 10 days). You can confirm that it was created. The HHI value of the review frequency of the restaurant ID 137445 was 18/36 above, but if the review written within 24 hours from the membership is regarded as false evaluation information and the null information is invalid, the following HHI 'value is obtained. .

HHI ' ₁₃₇₄₄₅ = 0 * (1/6) ^ 2 + 0 * (4/6) ^ 2 + 0 * (1/6) ^ 2 = 0/36

In the case of the restaurant 137445, the HHI value using only the frequency information of the evaluation information was 18/36, which was higher than that when using the frequency information of the evaluation information and the period information from the join date to the evaluation date (HHI '). The false evaluation information may not be determined as input online content.

As a result, the HHI value using the frequency information of the evaluation information and the period information from the subscription date to the evaluation date is much more accurate than the HHI value using only the frequency information of the evaluation information. Is generated. In this embodiment, HHI is applied for the convenience of explanation of the concept of the present invention, and the content of the present invention is not limited.

(5) The ratio or difference between the number of evaluations and the number of restaurants

According to one embodiment of the present invention, the determination assistance information may be frequency distribution information for each online content of evaluation information left by a specific user. For example, when the online content is restaurant information, the heavy user shows a pattern of inputting a lot of evaluation information in various restaurants, while the false user shows a pattern of inputting a lot of evaluation information only at a specific restaurant. Therefore, if the distribution or the ratio of the input frequency of the evaluation information for the at least one or more online content for each user is confirmed, it is possible to clearly analyze whether the user is a heavy user or a false evaluation user. Therefore, according to another embodiment of the present invention, the determination assistance information may be a distribution or ratio of input frequencies of evaluation information for at least one or more online contents for a specific user.

EXAMPLE

Table 3 below shows a case where the ratio of the number of inputs of the total comments (comments inputted to a specific review is defined as comments) and the number of restaurants (inputs of online content in which comments) are inputted as a comment assistance information is generated. It is a practical example to demonstrate the effect. Based on the Applicant's actual data, it is immediately available on the Mango Plate, Applicant's actual application service. Some nicknames have been deleted for privacy reasons. According to an embodiment of the present invention, the ratio of the total number of comment inputs and the number of restaurants in which a comment is input may be defined between 0 and 1.

User ID nickname Total comments
Input count Number of restaurants ratio
(Number of restaurants / total comment input) 449089 Five****** 15 One 0.067 520630 God****** 14 One 0.071 523323 this****** 7 One 0.143 199760 J ****** 6 One 0.167 261309 Y ****** 6 One 0.167 494654 way****** 6 One 0.167 541606 Kim****** 12 5 0.417

As shown in Table 3, the user 449089 inputs a total of 15 comments to the platform service according to an embodiment of the present invention, all of which were entered into one restaurant. As a result, the ratio (number of restaurants / total comment input) was 0.067. In contrast, the user 541606 inputs a total of 12 comments to six restaurants in the platform service according to an embodiment of the present invention. In this case, a ratio of 0.417 was derived. According to one embodiment of the present invention, the lower the ratio of the number of restaurants inputted to each specific user / the number of inputs of the total comments, the lower the user group who inputs false evaluation information. It was confirmed that close to. Table 4 below describes a review example of the user 449089 that is determined to have entered the false evaluation information. The user ID is the identification code of the user, and the review ID is the identification code of the review with the corresponding comment.

User ID review ID Name of Restaurant The ratio of the number of restaurants which inputted comment / the number of input of total comment Comment content 449089 1645232 Five****** 0.067 Hello. Oh ****** ^^ Thank you for visiting our store and leaving good feedback. We will always try to prepare more delicious and healthy. Thank you for your patronage and interest in the future ^^ 449089 1604497 Five****** 0.067 Hello. Oh ****** ^^ Thank you for visiting the store and leaving a review. We will do our best to provide better service based on your feedback. Thank you for your patronage in the future. Thank you ^^ 449089 1592723 Five****** 0.067 Hello. Oh ******. First of all, thank you for your valuable feedback. We will do our best to be an okipoki listening to every opinion. Thank you for your patronage in the future. Thank you ^^ 199760 957686 That****** 0.15 We will try harder and better ^^ Thank you for your advice! 253973 1143288 copper****** 0.25 Come try it out :) 375720 2158748 cowardice****** 0.25 I'll prepare for service training. Sorry. 494654 1622487 nose****** 0.15 A lot of people come here and delicious aunts service Choi? 520630 1885488 God****** 0.05 I'll give you a present. Please give me a text. ^^ 010 ********

In Table 4, only three of the user 449089's total comments are listed, and all the other comments have the same format. As shown in Table 4, it is apparent that the user 449089 is the restaurant official (representative false evaluation information input user). As a result, it is reasonable for the user 449089 to evaluate the evaluation information (eg, rating or star rating) that the user has rated for the restaurant (online content) as false evaluation information, and apply a penalty to the sum of the rating or the star rating. According to one embodiment of the present invention, when using the ratio of the number of restaurants inputted to the input number of the total comments inputted, it is possible to easily classify the user inputting the false evaluation information as described above. In addition to the user 449089, other examples in Table 4, user 199760, user 253973, user 376720, user 494654, and user 520630, can also be visually determined to be clear that the restaurant is related to the user by reading the comments. The distribution of the ratio of the number of the restaurant which inputs the comment about 316 random samples / the total number of comment input is shown. As shown in FIG. 7, it can be seen that a group in which false evaluation information is input and a group in which general evaluation information other than false evaluation information are input are clearly distinguished in a distribution. Therefore, according to an embodiment of the present invention, the ratio of the number of restaurants inputted to the total number of comments inputted for each specific user / total comment inputted number is calculated, plotted on the distribution map, and then a specific cutoff value is defined or clustering is performed. The user group may be classified by inputting false evaluation information.

(6) Determination of false evaluation information using HHI

Regarding the determination of the false evaluation information, the determination of the false evaluation information in the false information evaluation information determination module 110 according to an embodiment of the present invention may use the HerfindahlHirschman Index (HHI). Determination of false evaluation information of online content is used, and using the HHI value causes a noticeable difference than simply shown in the normal distribution. Therefore, according to this, without having to review the evaluation information of all the online contents, it is possible to find the online contents which change the frequency of the rating rapidly.

HHI is a measure of market concentration in a particular industry. The total market is 100, which is the square of the market share of each company (organization), plus the sum. This is a weighted sum of market share, weighted by the market share of each company itself. The closer the market share of the parent company is to monopoly, the higher the HHI, and the closer to free competition, the lower the HHI value. Generally, 1,800 or more are concentrated, 1,000 or less are concentrated, and the middle is called a centralized industry.

When the HHI is used to determine the false evaluation information using the frequency information of the evaluation information, it can be expressed as Equation 1 below.

In Equation 1 above, HHI may represent an HHI value, N may represent the number of days in which evaluation information is input into the online content, and S _i may represent a share of the evaluation information number of the date i. By drawing a normal distribution with these HHI values, plotting each online content from the normal distribution, and looking at the HHI change in the normal distribution, it is possible to find out the online content whose rating information fluctuates rapidly. For example, when the frequency of evaluation information of online content in which the evaluation information is input once a day increases suddenly, a noticeable difference occurs in the method of using the HHI value than any conventional method. In Equation 1, the square of S _i is taken for convenience of explanation, but n square is possible.

For example, assuming that a total of 10 evaluation information is input to a specific online content, if the evaluation information is input once a day and a total of 10 times are input, the HHI is 0.1. However, if the evaluation information is input six times once a day for six days, and the evaluation information is input four times twice a day for two days, the HHI becomes 0.14. When one evaluation information is input in one day, six evaluation information is input in one day, and three evaluation information is input in one day, the HHI increases rapidly to 0.46. This has the effect of allowing a person to find a restaurant whose frequency of evaluation information fluctuates rapidly without having to review the evaluation information of all restaurants.

HHI may also be used for other decision assistance information. For example, assuming that User A left 30 reviews in 30 restaurants and User B left 30 reviews in 15 restaurants, input of rating information for at least one online content for a particular user When the distribution or occupancy of the frequency is expressed by HHI, it can be seen that the difference becomes clear.

(7) The percentage of overlapping content among the rated online content

In addition, according to another embodiment of the present invention, the overlapping number of online contents in which evaluation information is input may be used as determination assistance information for a user group in which evaluation information is uploaded to the online content in question. More specifically, when the online content is restaurant information, the ratio of restaurants overlapping among a plurality of restaurants evaluated by a specific user group uploading evaluation information to specific restaurant information may be determination assistance information. As a method of determining whether the false evaluation information is based on the determination assistance information, a cluster analysis may be performed based on the determination assistance information, or a method using a distribution of ratio or distribution of HHI may be used.

According to another embodiment of the present invention, when a ratio of overlapping restaurants among the plurality of restaurants evaluated by a specific user group uploading evaluation information to specific restaurant information is used as the determination assistance information, the plurality of users cooperate with each other. Thus, an effect of detecting a false evaluation group for inputting false evaluation information is generated. Such a false evaluation group is difficult to detect using the network information according to an embodiment of the present invention because there is no mutual network. In addition, since the false evaluation group inputs false evaluation information in the long term, it is difficult to detect using frequency information according to an embodiment of the present invention. Accordingly, the false evaluation group is detected as a false evaluation group by using a ratio of overlapping restaurants among a plurality of restaurants evaluated by a specific user group uploading evaluation information to specific restaurant information according to another embodiment of the present invention. To achieve the highest accuracy.

According to another embodiment of the present invention, when a ratio of restaurants overlapping among a plurality of restaurants evaluated by a specific user group uploading evaluation information to specific restaurant information is used as the determination assistant information, the following steps are provided. A detection method of a user group for inputting false evaluation information may be performed. Finding a group of users who overlap three or more pieces of online content with rating information among all users can be a huge load on the server or database. However, the following method can significantly reduce the load on the server or DB.

Firstly, calculate how many online contents overlapped with other users by inputting rating information for each user (grouped by 2 people)

Second, screening user groups that have a certain number of overlapping online contents or a number of overlapping online contents / number of total evaluation information inputs.

Third, n iterative search for at least a portion of online content where input of evaluation information of a user group consisting of two people overlaps and a third user whose evaluation online content overlaps (grouping of three or more)

According to another embodiment of the present invention, a group of users who input false evaluation information by two persons is grouped into two persons, a sample size is reduced, and a third group in which the two groups and the online content in which the evaluation information is input partially overlaps. When searching for a user group in order of searching for users and grouping them into three or more users, the load generated on the server or DB is considerably reduced, thereby facilitating the detection of false evaluation information input user groups.

EXAMPLE

Hereinafter, Table 5 lists detailed information of specific user groups that overlap with restaurants uploading reviews. Table 5 is a practical example for demonstrating the effect that occurs when using the overlapping number of restaurant information for review in a specific user group as the decision assistance information. Based on the Applicant's actual data, it is immediately available on the Mango Plate, Applicant's actual application service. Some nicknames have been deleted for privacy reasons.

User ID nickname Review date User total reviews Representative review contents 557404 foil****** 3 times including 2016-05-17 12:28:02 PM 3 Bossa's splendid savory vegetables, freshly boiled fish, a cup of Dong Dongju 557432 W ****** 2016-05-17 1:25:12 PM 3 times 3 On a hot day, it's packed with ice dongdongju and Omusul Makguksu. 557735 foil****** 2016-05-18 9:00:50 AM 3 times 3 It's delicious all the time I've eaten it! Makguksu-chan 6557735 this****** 2016-05-18 9:05:29 AM 3 times 3 Live red crab 25,000 won unlimited refills! I ate it for 25,000 won. When you're in a hurry, you can also enjoy crab ramen for lunch! The bathroom is so clean 558123 *cold***** 2016-05-19 7:04:46 AM 3 times 3 Delicious Rural Scenery 588272 *pick***** 3 times including 2016-05-19 12:13:43 PM 3 My family's regular house sesame water Makguksu diet is the strongest but I am happy even if you become a bossam and Makguksu noodles ㅋㅋ 588899 * Benefit ***** 2016-05-21 1:56:34 AM 3 times 3 Of course, the traditional Makguksu noodle restaurant in Chuncheon is delicious. 588912 four****** 2016-05-21 2:02:18 AM
3 times 3 The truth of Bosamu's fantastic collaboration of boiled meat and vegetables

As shown in Table 5, it can be seen that a user group consisting of eight users rotates about three restaurants, leaving a good rating for each restaurant on different dates. The three restaurants are sewage Makguksu with Restaurant ID 141464, Sambat Village with Restaurant ID 141642, and Sewage Makguksu with Restaurant ID 258706. The user group shown in Table 5 can confirm that the purpose is so clear that each user has the same number of restaurants that all overlap with the total user reviews. The purpose is to enter a good rating for that restaurant group.

When the frequency information of the evaluation information reviewed above is used, it is difficult to catch false evaluation information because the reviews are not written in one day but distributed in different dates. However, when analyzing the information of a specific user group overlapping restaurants uploaded reviews, according to another embodiment of the present invention, even if the false rating information is input on various dates, it is possible to determine the false rating information. do.

When reading each of the reviews described in Table 5, it is very difficult to determine whether or not it is false evaluation information. According to another embodiment of the present invention, when reviewing detailed information of a specific user group that overlaps restaurants uploaded with each other, even if the actual review is difficult to determine whether the false rating information is effective, the false rating information can be caught. Is generated.

As described above, those skilled in the art will appreciate that the present invention can be implemented in other specific forms without changing the technical spirit or essential features. Therefore, the above-described embodiments are to be understood in all respects as illustrative and not restrictive. The scope of the present invention is shown by the following claims rather than the detailed description, and it should be construed that all changes or modifications derived from the meaning and scope of the claims and equivalent concepts are included in the scope of the present invention.

100: system for capturing false evaluation information of online content
101: client
102: network
103: platform services
104: online content delivery module
105: Online content database
106: evaluation information receiving module
107: false evaluation information detection device
108: judgment assistance information generation module
109: Evaluation Information Database
110: false evaluation information determination module
111: penalty processing module

Claims (2)

And a platform service connected to the client of the user so that the online content can be uploaded and the user can access and input evaluation information about the online content, and correspond to the online content for the detection of false evaluation information of the online content. A determination assistance information generation module for generating determination assistance information to be generated; And
A false evaluation information determination module, configured to be connected to the determination assistance information generating module, to receive the determination assistance information, and to determine whether false evaluation information is input to the online content based on the determination assistance information;
Including,
The determination assistance information,
A first evaluation information monopoly rate, which is information on how much a specific period unit monopolizes the number of evaluation information among period units divided by a specific period of time; And
The specific period unit is evaluated among the period units in which the specific online content is divided into the specific period intervals by invalidating a weight of the evaluation information whose input date of specific evaluation information of a specific user is after a predetermined period from the subscription of the specific user. A second evaluation information monopoly rate, which is information on how monopolistic the number of information is;
Including;
The first evaluation information monopoly rate and the second evaluation information monopoly rate for the specific online content are calculated according to the following equation,
The false rating information determining module determines whether the rating information of the specific online content is by a heavy user or a false rating user based on a difference between the first rating information monopoly rate and the second rating information monopoly rate. Classify and perform false evaluation information determination on the specific online content,
The false evaluation information determining module determines that the evaluation information of the specific online content is caused by the false evaluation user when the difference between the first evaluation information monopoly rate and the second evaluation information monopoly rate is equal to or less than a specific value. Characterized by
Device for detecting false rating information of online content that distinguishes heavy users from false rating users.
[Equation]

Here, HHI is the first evaluation information monopoly rate and the second evaluation information monopoly rate, N is the number of days that the evaluation information is input to the specific online content, S _i is the evaluation of the specific period unit i Information share share.
False evaluation information detection device configured on a platform service connected to the client of the user so that the online content is uploaded and the user can access and input the evaluation information on the online content, and perform false detection of the online content. A rating information receiving step of receiving evaluation information of online content at the client;
A determination assistance information generation step of generating, by the determination assistance information generation module, one component of the false evaluation information detection device, determination assistance information for the online content; And
A false evaluation information determining step of determining, by the false evaluation information determining module, which is one component of the false evaluation information detecting device, whether false evaluation information is input to the online content on the basis of the determination assistance information;
Including,
The determination assistance information,
A first evaluation information monopoly rate, which is information on how much a specific period unit monopolizes the number of evaluation information among period units divided by a specific period of time; And
The specific period unit is evaluated among the period units in which the specific online content is divided into the specific period intervals by invalidating a weight of the evaluation information whose input date of specific evaluation information of a specific user is after a predetermined period from the subscription of the specific user. A second evaluation information monopoly rate, which is information on how monopolistic the number of information is;
Including;
The first evaluation information monopoly rate and the second evaluation information monopoly rate for the specific online content are calculated according to the following equation,
In the false evaluation information determining step, the false evaluation information determining module determines whether the evaluation information of the specific online content is caused by a heavy user based on a difference between the first evaluation information monopoly rate and the second evaluation information monopoly rate. Determining whether or not by a false rating user to perform false judgment information determination on the specific online content,
The false evaluation information determining module determines that the evaluation information of the specific online content is caused by the false evaluation user when the difference between the first evaluation information monopoly rate and the second evaluation information monopoly rate is equal to or less than a specific value. Characterized by
A method of capturing false rating information of online content that distinguishes heavy users from false rating users.
[Equation]

Here, HHI is the first evaluation information monopoly rate and the second evaluation information monopoly rate, N is the number of days that the evaluation information is input to the specific online content, S _i is the evaluation of the specific period unit i Information share share.

Images