KR102644685B1 - Review reliability validation device and method of thereof - Google Patents

Abstract

An apparatus and method for verifying review reliability are disclosed. The review reliability verification device according to an embodiment of the present invention receives the customer's order information and contact information from the order input device through a communication unit that communicates with an external device, a memory that stores customer information, business information, and reviews, and a communication unit, and orders. A review request for each menu included in the information is sent to the contact information, and the reviews received in response to the sent review request are saved. If the number of saved reviews exceeds the preset number, some of the saved reviews are selected as reviews subject to verification. selection, and may include a processor that determines the reliability of each selected review subject to verification.

Description

Review reliability verification device and method {REVIEW RELIABILITY VALIDATION DEVICE AND METHOD OF THEREOF}

The present invention relates to a review reliability verification device and method. Specifically, the present invention relates to a review reliability verification device and method. Specifically, it collects reviews only for customers who have completed purchase verification, filters reviews suspected to be fake reviews for the collected reviews, and verifies the reliability of the filtered reviews. It relates to a review reliability verification device and method that can be performed.

When visiting a new area, such as when traveling, people want to try delicious restaurants. In these cases, people mostly look for reviews of restaurants. In everyday life, it has become common to look up reviews before visiting a new restaurant and then decide whether to visit or what to eat. Accordingly, the importance of reviews is increasing.

Paradoxically, as the importance of reviews increases, their reliability is decreasing. Originally, reviews are content written with objective evaluations based on experience, but there is a flood of reviews written after receiving compensation from companies, malicious reviews of competitors, and false reviews written without even visiting the company. In this regard, there is an increasing demand for the development of technology that can ensure the reliability of reviews.

The present invention is intended to solve the above-mentioned problems, and provides an efficient reliability verification function by granting the right to write reviews only to customers who have actually placed an order, and making the written reviews go through two stages of filtering and reliability verification. The purpose is to provide a review reliability verification device and method.

To achieve the above object, a review reliability verification device according to an embodiment of the present invention includes a communication unit that communicates with an external device, a memory that stores customer information, business information, and reviews, and a customer order from an order input device through the communication unit. Controls the communication unit to receive information and contact information and transmit a review request for each menu included in the order information to the contact information, and when a review is received in response to the transmitted review request, the received review is stored in the memory. , and if the number of the stored reviews exceeds a preset number, select some of the stored reviews as reviews to be verified, and may include a processor that determines the reliability of each of the selected reviews to be verified.

The processor counts the number of daily review requests and the number of input reviews, respectively, calculates the average value of the ratio of input reviews to review requests based on the counted number of review requests and the number of input reviews, and calculates the average value of the ratio of input reviews to review requests on a specific day. If the ratio of input reviews to review requests exceeds the calculated average value, the review on the specific day may be selected as the review subject to verification.

In addition, the processor counts the number of daily review requests for each route through which the order information and contact information were obtained, calculates an average value of the number of daily review requests for each route, and reviews requests triggered in a specific route on a specific day. If the number of cases exceeds the calculated average value, reviews written through a specific path on the specific day may be selected as reviews to be verified.

And the processor collects the writing time of all reviews of the author who wrote the selected review subject to verification, and when the number of reviews written by the author during a preset time exceeds the preset number, the reliability of the selected review subject to verification is increased. You can decide on low.

In addition, the processor collects the writing time of all reviews of the author who wrote the selected review target review and the location of the review target business, and divides the review writing time interval of the author by the location interval of the business to compare the time interval. If the value is less than the preset value, the reliability of the selected review target review may be determined to be low.

And the processor collects the ratings and number of reviews of all reviews for the review target business of the author who wrote the selected review target review, and if the number of collected reviews is greater than a preset value, the average review rating value of the author and the average value of all review ratings of the review target business, and if the difference between the two average values as a result of the comparison is more than a preset value, the reliability of the selected review target review can be determined to be low.

In addition, the processor collects and stores reviews determined to have low reliability in a memory, determines similarity by analyzing a plurality of characteristic information of the stored reviews, and verifies characteristic information for which the determined similarity is greater than or equal to a preset value. It can be set as a standard for selecting target reviews.

Meanwhile, the review reliability verification method according to an embodiment of the present invention to achieve the above object includes obtaining the customer's order information and contact information, and sending a review request for each menu included in the order information to the contact information. Transmitting, if a review is input in response to the transmitted review request, storing the input review; If the number of stored reviews exceeds a preset number, selecting some of the stored reviews as reviews to be verified; It may include determining the reliability of each of the selected verification target reviews.

The step of selecting the review subject to verification includes counting the number of daily review requests and the number of input reviews, and averaging the ratio of input reviews to review requests based on the counted number of review requests and input reviews. It may include calculating a value and, if the ratio of input reviews to review requests for a specific day exceeds the calculated average value, selecting the review of the specific day as a review subject to verification.

In addition, the step of selecting the review subject to verification includes counting the number of daily review requests for each route through which the order information and contact information were obtained, calculating the average number of daily review requests for each of the counted routes, and calculating the average value of the number of daily review requests for each route on a specific day. If the number of review requests triggered by a specific path exceeds the calculated average value, a step of selecting a review written through a specific path on the specific day as a review subject to verification may be included.

And the step of determining the reliability of each of the selected verification target reviews includes collecting the writing time of all reviews of the author who wrote the selected verification target review, and determining the number of reviews written by the author during a preset time. If the number is exceeded, the step of determining the reliability of the selected review subject to be verified as low may be included.

In addition, the step of determining the reliability of each of the selected reviews to be verified includes collecting the writing time of all reviews and the location of the business to be reviewed by the author who wrote the selected review to be verified, and the review writing time of the author. If the interval divided by the location interval of businesses to compare the time interval is less than a preset value, it may include determining the reliability of the selected review target review as low.

And the step of determining the reliability of each of the selected reviews to be verified includes collecting the ratings and number of reviews of all reviews for the business to be reviewed by the author who wrote the selected review to be verified, and the number of reviews of the collected reviews. If the number is more than a preset value, comparing the average review rating of the author and the average value of all review ratings of the review target business, if the difference between the two average values as a result of the comparison is more than a preset value, the selected review target review It may include determining the reliability to be low.

In addition, collecting and storing reviews determined to have low reliability in a database, determining similarity by analyzing a plurality of characteristic information of reviews stored in the database, and verifying characteristic information where the determined similarity is greater than or equal to a preset value. A step of setting a standard value for selecting a target review may be further included.

According to the various embodiments of the present invention as described above, the authority to write a review can be granted only to customers who have actually placed an order, and a two-step reliability verification process can be performed, so content that does not play the role of a review It has the effect of removing . Additionally, according to various embodiments of the present invention, small business owners can be protected from unexpected damage by removing malicious reviews.

1 is a diagram illustrating the configuration of a review reliability verification system according to an embodiment of the present invention;
Figure 2 is a block diagram schematically showing the configuration of a review reliability verification device according to an embodiment of the present invention, and
Figure 3 is a flowchart illustrating a review reliability verification method according to an embodiment of the present invention.

Hereinafter, various embodiments of this document are described with reference to the attached drawings. However, this is not intended to limit the technology described in this document to specific embodiments, and should be understood to include various modifications, equivalents, and/or alternatives of the embodiments of this document. Similar reference signs may be used for similar components in connection with the description of the drawings.

In this document, expressions such as “have,” “may have,” “includes,” or “may include” refer to the presence of the corresponding feature (e.g., a component such as a value, function, operation, or part). , does not rule out the existence of additional features.

In this document, expressions such as “A or B,” “at least one of A and/or B,” or “one or more of A and/or B” may include all possible combinations of the items listed together. For example, “A or B”, “at least one of A and B”, or “at least one of A or B” (1) includes at least one A, (2) includes at least one B, or ( 3) It can refer to all cases including both at least one A and at least one B. As used herein, expressions such as “first,” “second,” “first,” or “second” may modify various elements in any order and/or importance, and may refer to one element as another element. It is only used to distinguish from an element and does not limit the corresponding components.

The expression "configured to" used in this document may mean "suitable for" or "having the capacity to" depending on the context. It can be used interchangeably with ", "designed to," "adapted to," "made to," or "capable of." The term “configured (or set to)” may not necessarily mean “specifically designed to” in hardware. Instead, in some contexts, the expression “a device configured to” may mean that the device is “capable of” working with other devices or components. For example, a “processor configured (or set) to perform A, B, and C” means performing the operations by executing one or more software programs stored in a dedicated processor (e.g., embedded processor) or memory device to perform the operations. It may refer to a general-purpose processor (e.g., CPU or Application Processor) that can perform certain functions.

Figure 1 is a block diagram illustrating elements constituting a review reliability verification system 1000 according to an embodiment of the present invention. Referring to FIG. 1 , the review reliability verification system 1000 may include a review reliability verification device 100, an order input device 200, a review input device 300, and a review provision device 400.

The review reliability verification device 100 can collect reviews only for customers whose purchase status has been verified, filter reviews suspected to be fake reviews from the collected reviews, and verify the reliability of the filtered reviews. The review reliability verification device 100 may receive order information and customer contact information from the order input device 200. Through this, the review reliability verification device 100 can initially transmit a review request only to customers who actually placed an order. In Figure 1, the review reliability verification device 100 is expressed as a cloud, but it can of course be implemented as a general server device.

The order input device 200 may be an unmanned ordering device such as a kiosk at a business, a device such as a POS where the owner or employee directly enters order information at the business, or a device such as a smartphone that orders food through a delivery application or web page. You can. The order input device 200 can receive an order input from a customer and allow the customer to input contact information.

The review reliability verification device 100 requests a review from a customer who has actually placed an order. At this time, the review input device 300 is a device that can receive a review request and input a review. For example, if a customer places an order through a smartphone, the customer's smartphone can perform the functions of both the order input device 200 and the review input device 300.

The review providing device 400 is a device that can provide reviews selected as highly reliable by the review reliability verification device 100 to the general public. Any electronic device that can run a web page or application to provide a review can function as the review providing device 400.

Hereinafter, the review reliability verification method will be described focusing on the review reliability verification device 100. Figure 2 is a block diagram schematically showing the configuration of the review reliability verification device 100. Referring to FIG. 2 , the review reliability verification device 100 may include a communication unit 110, a memory 120, and a processor 130. Of course, in addition to the components shown, various components such as an input unit (not shown), an output unit (not shown), and a display unit (not shown) may be additionally included.

The communication unit 110 may communicate with external devices such as the order input device 200, the review input device 300, and the review providing device 400 using a wired or wireless network. For example, the communication unit 110 may receive information about order details (menu, quantity, etc.) and contact information (phone number, member ID, etc.) entered by the user from the order input device 300. The communication unit 110 can use various wireless communication methods such as Bluetooth, Zigbee communication, WiFi, InfraRed (IR) communication, and NFC (Near Field Communication). Additionally, the communication unit 110 may use a wired communication method such as High Definition Multimedia Interface (HDMI), Low Voltage Differential Signaling (LVDS), or Local Area Network (LAN).

The memory 120 may store various modules, software, and data for driving the review reliability verification device 100. For example, the memory 120 stores data such as received order information, contact information, reviews, business information, menu information, software for verifying review reliability, control signal data such as communication protocols with external devices, etc. It can be.

The memory 120 may be implemented in the form of flash memory, hard disk drive (HDD), solid state drive (SSD), etc. For example, the memory 120 is a ROM for storing programs and/or applications for performing operations of the review reliability verification device 100, and temporarily stores data according to the operation of the review reliability verification device 100. RAM can be provided for Additionally, the memory 120 may further include an Electrically Erasable and Programmable ROM (EEPROM) for storing various reference data.

The processor 130 may control the above-described configurations of the review reliability verification device 100. For example, the processor 130 may control the communication unit 110 to transmit a review request to the review input device 300. The processor 130 may be manufactured in the form of one or more hardware chips and mounted on the consulting providing device 100. For example, the processor 130 may be manufactured as a general-purpose processor (eg, CPU or Application Processor), or may be manufactured as a dedicated hardware chip for artificial intelligence.

The processor 130 may receive the customer's order information and contact information from the order input device 200 through the communication unit 110. The customer's order information may include menu information, quantity information, etc. The contact information may generally be a phone number, but may also be a member ID, which is information that can be linked to the orderer's phone number as a member ID of an ordering application, web page, review provision application, or web page. Since the processor 130 receives menu information, it can request a review of each menu rather than the entire order.

The processor 130 may control the communication unit 110 to transmit a review request for each menu included in the received order information to the received contact information. Referring to FIG. 1, a review request may be transmitted to the review input device 300. Since reviews can be entered only when an actual order is placed in this way, the review reliability verification device 100 can primarily prevent the writing of fake reviews.

When a review is received in response to a transmitted review request, the processor 130 may store the received review in the memory 120. As described above, the memory 120 may store meta information such as the business where the received review was ordered, the order menu, the order time, and the review input time. The processor 130 may perform verification of the reliability of the stored reviews when the number of stored reviews exceeds a preset number. When the number of stored reviews is small, there is little incentive to filter out fake reviews and there is a lack of sufficient samples to analyze the characteristics of fake reviews.

The processor 130 may select some reviews whose reliability is questionable among the stored reviews as reviews to be verified. Because it is inefficient to verify reliability for all reviews, the processor 130 selects reviews for which reliability is to be verified based on various criteria that will be described below. And the processor 130 can determine the reliability of each selected review subject to verification.

For example, if the number of reviews written on a specific day is abnormally large compared to the number of orders, the processor 130 may select reviews written on that specific day as reviews to be verified. Specifically, the processor 130 may count the number of daily review requests and the number of input reviews, respectively. A review is not entered for every review request, and the ratio of input reviews to review requests is statistically determined for each business. The processor 130 may calculate the average value of the ratio of input reviews to review requests during a specific period. In addition, when the ratio of input reviews to review requests for a specific day exceeds the calculated average value, the processor 130 may select the review for the specific day as the review subject to verification.

As another example, if the number of orders generated through a specific path on a specific day is abnormally large, the processor 130 may select reviews written on that specific day and through a specific route as reviews to be verified. Low-trust reviews do not only include reviews written with low ratings by malicious customers. There are also cases where businesses create fake orders to increase their ratings and write reviews with high ratings for those orders, thereby lowering trust. In this embodiment, it is possible to prevent not only malicious customers but also the business itself from manipulating reviews.

Specifically, the processor 130 may count the number of daily review requests for each path through which order information and contact information were obtained. For example, the processor 130 can count the number of orders placed through a kiosk, the number of orders placed through a delivery application, and the number of orders received through a POS at a store. And the processor 130 can calculate the average value of the number of daily review requests for each counted path. If the number of review requests triggered through a specific path on a specific day exceeds the calculated average value, the processor 130 may select a review written through a specific path on a specific day as a review to be verified.

As another example, the processor 130 may collect and analyze reviews determined to have low reliability, thereby setting a new standard for determining reliability. Specifically, the processor 130 may collect reviews whose reliability is determined to be 'low' and store them in the memory 120. Reliability may be expressed as 'high' or 'low', or may be expressed as a score. The processor 130 may analyze a plurality of feature information of reviews determined to be low in reliability, and determine what feature information the reviews determined to have low reliability have in common among the plurality of feature information. can do. That is, the processor 130 may determine the degree of similarity by analyzing a plurality of characteristic information of the stored reviews, and may set the characteristic information for which the determined similarity is greater than or equal to a preset value as a reference value for selecting a review to be verified.

Once a review subject to verification is selected, the processor 130 can verify the reliability of each review. For example, the processor 130 may determine reliability as low when the number of reviews written by a reviewer over a certain period of time is excessively large. Because the processor 130 can primarily filter to write a review only when there is an actual order, if the number of reviews written by the reviewer is excessively large, the order is to write a review rather than to eat the actual food. It is highly likely that it occurred. Specifically, the processor 130 may collect the writing time of all reviews of the author who wrote the selected review subject to verification. And, if the number of reviews written by the author during a preset time exceeds the preset number, the processor 130 may determine the reliability of the selected review target to be verified as low.

As another example, the processor 130 may verify reliability based on whether it is possible to move from the establishment where the first order was placed to the establishment where the second order was placed during the time between the occurrence time of the first order and the occurrence time of the second order. there is. This embodiment can be particularly useful when orders are entered through a kiosk or POS located in a business. Specifically, the processor 130 may collect the writing time of all reviews of the author who wrote the selected review target review and the location of the review target business. If the time interval for writing a review by the author divided by the location interval of businesses to compare the time interval is less than a preset value, the processor 130 may determine that the customer was actually unable to move between businesses. Accordingly, the processor 130 may determine the reliability of the selected review target to be low.

As another example, the processor 130 may determine a review with low reliability when the review rating for a specific menu at a business is far from the overall average. Specifically, the processor 130 may collect the ratings and number of reviews of all reviews for the business to be reviewed by the author who wrote the selected review to be verified. If the number of collected reviews is large enough to be statistically significant, the processor 130 may compare the average review rating of the author with the average value of all review ratings of the business being reviewed. And, as a result of the comparison, if the difference between the two average values is greater than a preset value, the processor 130 may determine the reliability of the selected review target to be verified as low.

Figure 3 is a flowchart illustrating a review reliability verification method according to an embodiment of the present invention. Referring to FIG. 3, the review reliability verification device 100 may receive the customer's order information and contact information (S310). The customer's order information may include information on the business where the order was placed, menu information included in the order, order occurrence time information, etc. Contact information may include phone number, ID, etc.

Subsequently, the review reliability verification device 100 may transmit a review request for each ordered menu to the received contact information (S320). Accordingly, a review request is sent to the contact information entered by the customer when ordering, which has the effect of allowing reviews to be written only for actual purchasing customers. In other words, the review reliability verification device 100 can increase review reliability by first eliminating an environment in which fake reviews can be written. Additionally, the review reliability verification device 100 is useful in that it can receive more detailed customer feedback by requesting a review for each menu included in the order.

When a review is input in response to a review request (S330-Y), the review reliability verification device 100 may store the input review (S340). And when a certain number of reviews or more are stored (S350-Y), the review reliability verification device 100 may select some of the stored reviews as reviews to be verified (S360). This is because it is inefficient to verify the reliability of all reviews, so we go through a process of pre-selecting reviews whose reliability is questionable. For example, if the number of reviews entered compared to the number of daily review requests exceeds the average value on a specific day, the review reliability verification device 100 may select a review written on a specific day as a review subject to verification. As another example, if the number of review requests triggered by a specific order route on a specific day exceeds the average value, the review reliability verification device 100 may select a review written through a specific order route on that specific day as a review subject to verification. there is.

Subsequently, the review reliability verification device 100 may determine the reliability of each review selected as a review subject to verification (S370). For example, the review reliability verification device 100 is selected when the author of the review subject to verification writes reviews exceeding a preset number over a certain period of time, that is, when excessive orders are generated and reviews are written over a certain period of time. The reliability of the target review can be judged as low. As another example, the review reliability verification device 100 determines whether the review target review author can continuously create orders offline and whether the order can sufficiently move the distance between the businesses where the order was continuously generated, thereby determining the reliability of the review target review. can also be verified. As another example, the review reliability verification device 100 may determine the reliability of the review subject to verification written by the author as low if the author of the review subject to verification continues to leave reviews with ratings that are different from the overall rating for the business. there is.

Additionally, the review reliability verification device 100 may collect reviews determined to have low reliability and store them in a database. This is to analyze the characteristics of reviews determined to be of low reliability and use them as a standard for selecting reviews subject to verification. Artificial intelligence models can also be used to analyze the characteristics of reviews. In this way, the review reliability verification device 100 can build an increasingly sophisticated reliability verification system as review data accumulates and reviews with low reliability are filtered out.

Meanwhile, the term “unit” or “module” used in this specification includes a unit consisting of hardware, software, or firmware, and can be used interchangeably with terms such as logic, logical block, part, or circuit, for example. there is. “Part” or “module” may be an integrated part, the minimum unit that performs one or more functions, or a part thereof. For example, a module may be comprised of an application specific integrated circuit (ASIC).

Various embodiments of the present invention may be implemented as software including instructions stored in a machine-readable storage media (e.g., a computer). The device is a device capable of calling commands stored from a storage medium and operating according to the called commands, and may include an electronic device (eg, the review reliability verification device 100) according to the disclosed embodiments. When the instruction is executed by a processor, the processor may perform the function corresponding to the instruction directly or using other components under the control of the processor. Instructions may contain code generated or executed by a compiler or interpreter. A storage medium that can be read by a device may be provided in the form of a non-transitory storage medium. Here, 'non-transitory' only means that the storage medium does not contain signals and is tangible, and does not distinguish whether the data is stored semi-permanently or temporarily in the storage medium.

According to one embodiment, methods according to various embodiments disclosed in this document may be included and provided in a computer program product. Computer program products are commodities and can be traded between sellers and buyers. The computer program product may be distributed on a machine-readable storage medium (e.g. compact disc read only memory (CD-ROM)) or online through an application store (e.g. Play Store™). In the case of online distribution, at least a portion of the computer program product may be at least temporarily stored or created temporarily in a storage medium such as the memory of a manufacturer's server, an application store's server, or a relay server.

Each component (e.g., module or program) according to various embodiments may be composed of a single or plural entity, and some of the above-described sub-components may be omitted, or other sub-components may be variously used. Further examples may be included. Alternatively or additionally, some components (e.g., modules or programs) may be integrated into a single entity and perform the same or similar functions performed by each corresponding component prior to integration. According to various embodiments, operations performed by a module, program, or other component may be executed sequentially, in parallel, iteratively, or heuristically, or at least some operations may be executed in a different order, omitted, or other operations may be added. You can.

100: Review reliability verification device
110: Department of Communications
120: memory
130: processor
200: Order input device

Claims (14)

A communication unit that communicates with an external device;
Memory to store customer information, business information, and reviews;
Receives the customer's order information and contact information from the order input device through the communication unit, controls the communication unit to transmit a review request for each menu included in the order information to the contact information, and responds to the sent review request. When a review is received, the received review is stored in the memory, and if the number of the stored reviews exceeds a preset number, some of the stored reviews are selected as reviews to be verified, and the reliability of each of the selected reviews to be verified is determined. Includes a processor that determines,
The processor,
The reviews determined to have low reliability are collected and stored in memory, the similarity is determined by analyzing a plurality of characteristic information of the stored reviews, and the characteristic information for which the determined similarity is greater than a preset value is used as a criterion for selecting the review target for verification. Set to , count the number of daily review requests and the number of input reviews, respectively, and calculate the average value of the ratio of input reviews to review requests based on the counted number of review requests and the number of input reviews, and calculate the average value of the ratio of input reviews to review requests on a specific day. If the ratio of input reviews to review requests exceeds the calculated average value, the reviews on the specific day are selected as reviews subject to verification,
Collects the writing time of all reviews of the author who wrote the selected review subject to verification, and determines the reliability of the selected review subject to verification as low if the number of reviews written by the author during a preset time exceeds the preset number,
Collect the writing time of all reviews of the author who wrote the selected verification subject review and the location of the business to be reviewed, and the time interval for writing the review by the author divided by the location interval of the business to compare the time interval is less than the preset value. If so, the reliability of the selected verification target review is determined to be low,
Collect the rating and number of reviews of all reviews for the business subject to review by the author who wrote the selected review subject to verification, and if the number of collected reviews is greater than a preset value, the average value of the review rating of the author and the business subject to review are collected. A review reliability verification device that compares the average value of the overall review ratings, and determines the reliability of the selected review subject to be verified as low if the difference between the two average values as a result of the comparison is greater than a preset value. delete A communication unit that communicates with an external device;
Memory to store customer information, business information, and reviews;
Receives the customer's order information and contact information from the order input device through the communication unit, controls the communication unit to transmit a review request for each menu included in the order information to the contact information, and responds to the sent review request. When a review is received, the received review is stored in the memory, and if the number of the stored reviews exceeds a preset number, some of the stored reviews are selected as reviews to be verified, and the reliability of each of the selected reviews to be verified is determined. Includes a processor that determines,
The processor,
The reviews determined to have low reliability are collected and stored in memory, the similarity is determined by analyzing a plurality of characteristic information of the stored reviews, and the characteristic information for which the determined similarity is greater than a preset value is used as a criterion for selecting the review target for verification. Set to , count the number of daily review requests for each route through which the order information and contact information was obtained, calculate the average value of the number of daily review requests for each route counted, and calculate the number of review requests triggered by a specific route on a specific day. If it exceeds the calculated average value, reviews written through a specific path on the specific day are selected as reviews subject to verification,
Collects the writing time of all reviews of the author who wrote the selected review subject to verification, and determines the reliability of the selected review subject to verification as low if the number of reviews written by the author during a preset time exceeds the preset number,
Collect the writing time of all reviews of the author who wrote the selected verification target review and the location of the business to be reviewed, and the time interval for writing the review by the author divided by the location interval of the business to compare the time interval is less than the preset value. If so, the reliability of the selected verification target review is determined to be low,
Collect the rating and number of reviews of all reviews for the business subject to review by the author who wrote the selected review subject to verification, and if the number of collected reviews is greater than a preset value, the average value of the review rating of the author and the business subject to review are collected. A review reliability verification device that compares the average value of the overall review ratings and, if the difference between the two average values as a result of the comparison is greater than a preset value, determines the reliability of the selected review target review as low. delete delete delete delete delete delete delete delete delete delete delete