TWM569027U - Online review system - Google Patents

Abstract

An online review system includes a first electronic device, a banking server, and a customer database. The bank server receives the application request from the first electronic device and includes an image related to a face and an identity information and is related to a bid for a financial product, according to the face extracted from the image data. One of the feature data and the identification data determines the target customer data, and when it is determined that the face feature data and the identification data respectively correspond to the reference identity identification data and the reference face feature data of the target customer data, A credit rating model generates a credit rating result based on the face of the image data, and executes the financial product bid when it is determined that the credit rating result is not less than a threshold level.

Description

Online review system

This new type is related to a system of review, especially an online review system.

Existing online banking allows its customers to apply for financial products online. However, when the identity of the customer is used by other users who are different from the customer (common fraudulent means include, for example, the use of fake ID cards), if allowed The other user's bid for a financial product, such as a loan, may result in the loss of the customer and the bank.

In addition, even if the identity of the customer is not fraudulent, the credit of the customer may affect the performance of the transaction, so the bank still needs to conduct a credit evaluation of the customer. The bank usually evaluates according to the records it stores or the records provided by the Link Center corresponding to the customer. However, there is still room for improvement in the credibility of the evaluation results based only on the records.

Thus, it is an object of the present invention to provide an online review system that overcomes at least one of the disadvantages of the prior art.

Thus, the new online review system includes a first electronic device operated by a user, a bank server, and a customer database. The bank server is provided by a bank and is communicable with the first electronic device and stores a pre-trained credit rating model. The customer database is connected to the bank server, and the customer database stores a plurality of customer data respectively corresponding to the plurality of customers of the bank, and each customer data has reference identity identification data and reference face feature data.

The first electronic device obtains image data related to a face and an identity information by the user operation, and transmits a bid request containing the image data and related to a financial product to the bank server.

Receiving the bid request from the first electronic device, the bank server extracts the face feature data of the face and the identity identification data of the identity information from the image data, and according to the face feature data and the One of the identification materials, one of the customer data is used as the target customer data.

The bank server determines whether the face feature data, the identity identification data respectively meets the reference face feature data of the target client profile and the reference identity identification data.

When the bank server determines that the face feature data and the identity identification data respectively correspond to the reference identity identification data and the reference face feature data of the target client profile, the credit rating model at least according to the image The face of the data produces a credit rating.

The bank server executes the bidding of the financial product when it is determined that the credit rating result is not less than a threshold level.

In some implementations, the bank server determines that the similarity between the facial feature data and the reference facial feature data of one of the customer data stored in the customer database is not less than one. When the first threshold is devalued, the one of the customer data is the target customer profile.

In some implementations, the bank server further stores a pre-trained face feature comparison model, and compares the face feature comparison model to obtain the face feature data and the The similarity between the reference facial features of the customer data.

In some implementations, each customer data stored in the customer database further has a bid history data, and the bid history data has at least one financial product that the customer has applied for by the customer data.

The bank server not only determines whether the facial feature data and the identification data respectively correspond to the reference identity identification data and the reference facial feature data of the target customer data, and further determines the bid history data according to the target customer data. Determining whether the financial item related to the bid request is the same as one of the financial items of the bid history data of the target customer data.

The bank server determines that the face feature data and the identity identification data respectively match the reference identity identification data and the reference face feature data of the target customer profile, and the financial product related to the bid request is not related to the When one of the financial products of the target history data of the target customer data is the same, the credit rating model generates the credit rating result based on the face of the image data.

In some implementations, the bank server can also communicate with a link server belonging to the central office. The bank server determines that the face feature data and the identity identification data respectively match the reference identity identification data and the reference face feature data of the target customer profile, and the financial product related to the bid request is not related to the When one of the financial products of the target customer data of the target customer data is the same, a data request containing the reference identification data of the target customer data is transmitted to the affiliate server. When receiving the data request from the bank server, the association server transmits the joint information related to the customer's credit record corresponding to the target customer data to the bank server. The bank server generates the credit rating result based on the face of the image data and the joint information by the credit rating model.

In some implementations, the joint information has at least one of a credit card payment record, a loan record, an education level, a job category, and a job title corresponding to the customer corresponding to the target customer profile.

In some implementations, the bank server further stores an external environmental indicator including at least one of a current Gross Domestic Product (GDP) and a boom signal. The bank server uses the credit rating model to generate the credit rating result based not only on the face of the image data and the joint information but also on the external environment indicator.

In some implementations, the system further includes a second electronic device in communication with the bank server. The bank server determines that the face feature data and the identity identification data do not respectively correspond to the reference identity identification data and the reference face feature data of the target customer profile, or the financial commodity related to the bid request and the When one of the financial products of the target history data of the target customer data is the same, a first notification message related to the manual photo is sent to the second electronic device for display. When the bank server determines that the credit rating result is less than the threshold level, transmitting a second notification message related to the manual note to the second electronic device for display.

In some implementations, the first electronic device obtains the image data by capturing a document including the face and the identity information.

In some implementations, the first electronic device obtains the first image data related to the identity information by capturing a document containing the identity information, and obtains related information by capturing the face of the user. The second image data of the human face, the image data includes the first image data and the second image data.

The effect of the present invention is that the bank server can determine the face profile data and the identity identification data are consistent with the target customer profile and the credit rating model is used according to the face of the image data. When the more credible credit rating result is not less than the threshold level, the financial product is allowed to be applied, so that not only can the effective use of the customer identity be fraudulently exploited, but also the external appearance of the face. There are great opportunities for characteristics related to economic characteristics, so it can effectively reduce the risk of banks.

Before the present invention is described in detail, it should be noted that in the following description, similar elements are denoted by the same reference numerals.

Referring to FIG. 1 and FIG. 2, the online online review system includes a bank server 1, a customer database 2, a first electronic device 3, and a second electronic device 4.

The bank server 1 is provided by a bank and can communicate with the first electronic device 3, the second electronic device 4, and a link server 6 belonging to the association center via a communication network 5. The bank server 1 stores a pre-trained credit rating model 11, a pre-trained facial feature comparison model 12, and an external environmental indicator 13, for example, including current GDP ( Gross Domestic Product, GDP) and the boom signal.

The customer database 2 is connected to the bank server 1, and stores a plurality of customer profiles 21, 21' corresponding to a plurality of customers of the bank, each customer profile 21 (21') having reference identity identification data 211, Referring to the face feature data 212 and the bid history data 213, the bid history data 213 has at least one financial product that the client has applied for by the customer profile 21 (21'). In this embodiment, the financial goods are, for example, credit cards, loans, insurance, and the like.

The first electronic device 3 is held by a user, such as a customer of the bank. In this embodiment, the first electronic device 3 is, for example, a smart phone, and is not limited thereto. In other embodiments, it may also be a tablet computer, a notebook computer, a personal computer, or the like.

In the present embodiment, the second electronic device 4 is held by, for example, a staff member of the bank. In this embodiment, the first electronic device 3 is, for example, a personal computer, and is not limited thereto. In other embodiments, it may also be a tablet computer, a notebook computer, a smart phone, or the like.

In the following, in the case where the user holds the first electronic device 3 and an identity card to facilitate online bidding for a credit card, how the above embodiment cooperates with the association server 6 to perform an online review method.

First, as described in step S201, the first electronic device 3 obtains image data related to a face and an identity information by the user operation. In this embodiment, the identity information is, for example, an identity card number, and is not limited thereto.

In this embodiment, the first electronic device 3 obtains the image data by capturing a document including the face and the identity information, but is not limited thereto. For example, the first electronic device 3 captures the front side of the identity card held by the user (including the headshot and the identity card number), and obtains image data related to the face and the identity card number. In other implementations, the first electronic device 3 obtains the first image data related to the identity information by capturing a document containing the identity information and obtains related information by capturing the face of the user. The second image data of the face, the image data includes the first image data and the second image data. For example, the first electronic device 3 captures the front side of the identity card held by the user to obtain Corresponding to the first image data of the identity card number, and obtaining the second image data related to the face by capturing the face of the user, it is worth mentioning that the image data obtained in this way is related. The face of the user is the face of the user before the appropriate appearance, which can make the results produced by the subsequent steps S205 and S207 of the embodiment have relatively better accuracy than the big picture on the identity card.

Next, as described in step S202, the first electronic device 3 transmits a bid request containing the image data and related to a financial product to the bank server 1. Continuing the previous example, the bid request is related to the credit card's bid.

When the bank server 1 receives the bid request from the first electronic device 3, the bank server 1 extracts the face feature data of the face and the identity information from the image data, as described in step S203. Identification information.

Next, in step S204, the bank server 1 uses one of the customer profile information 21, 21' (for example, the customer profile 21') as the target customer profile according to one of the face feature data and the identity identification data. twenty one'. In this embodiment, the bank server 1 determines between the face feature data and the reference face profile 211 of one of the pen client profiles 21, 21' stored in the client database 2. When the similarity is not less than a first threshold, the customer information 21, 21' is the target customer data (ie, the customer data 21'), and is not limited thereto. In other implementations, the bank server 1 can also be designed to treat the identity identification information 212 when the identity identification data is identical to the reference identity identification information 212 of one of the customer profiles 21, 21' The information 21, 21' is the one of the target customer data (ie customer data 21').

It should be noted that, in this embodiment, the bank server 1 is compared by the face feature comparison model 12 to obtain the face feature data and the pen customer data 21, 21'. The similarity between the reference face feature data 212 can effectively reduce the computing resources of the system, but not limited thereto. In other implementations, the face can be used without using the face feature comparison model 12. The feature matching software compares the face feature data with all of the reference face feature data 212. It should be noted that when the number of the reference facial feature data 212 is large, if the facial feature data is compared with the reference facial feature data 212 one by one, it will be very time consuming and consume computing resources. In this embodiment, the face feature comparison model 12 that is pre-trained is used for comparison, so that it is not necessary to compare one by one, and has the effect of saving time and saving computing resources.

In step S205, the bank server 1 determines whether the face feature data and the identity identification data respectively correspond to the reference identity identification data 211 and the reference face feature data 212 of the target client profile (ie, the client profile 21'). . If the bank server 1 determines that the face profile data and the identity identification data do not respectively correspond to the reference identity identification data 211 and the reference face profile data 212 of the target client profile (ie, the client profile 21'), the process Go to step S206. Otherwise, if the bank server 1 determines that the face feature data and the identity identification data respectively correspond to the target customer profile (ie, customer profile 21'), the reference identity identification data 211 and the reference person The face feature data 212, the flow proceeds to step S207.

In step S206, a first notification message related to the manual note is transmitted to the second electronic device 4 for display. In this way, the staff of the bank can be notified immediately that the documents held by the user may be forged for subsequent processing, thus reducing the risk of the customer being fraudulently used.

In step S207, the bank server 1 determines, according to the bid history data 213 of the target customer profile (ie, the customer profile 21'), whether the financial item related to the bid request and the target customer profile (ie, customer profile 21) ') One of the financial records of the bid history data 213 is the same. When the bank server 1 determines that the financial item related to the bid request is the same as one of the financial items of the bid history data 213 of the target customer data (ie, the customer data 21'), the process proceeds to step S206; otherwise, the process proceeds to step S206; otherwise, The process proceeds to step S208. In this way, the bank staff can be notified immediately that the customer corresponding to the target customer profile (ie, the customer profile 21') is about to repeat the same financial product, so that the subsequent processing can be reduced. The risk of fraudulent use of this customer identity. If the previous example shows that the credit card that the user wants to apply for is the same as the credit card indicated in the bid history data, there is a high probability that the user will fraudulently use the customer corresponding to the target customer profile 21'. The identity, if you immediately notify the bank's staff, can effectively reduce the chance of fraudulent use.

In addition, in other embodiments, step S207 may be omitted, that is, when the determination result in step S205 is YES, the flow proceeds directly to step S208.

In step S208, the bank server 1 transmits a data request containing the reference identification data 211 of the target customer profile (i.e., customer profile 21') to the association server 6.

When the association server 6 receives the data request from the bank server 2, the association server 6 transmits the corresponding information corresponding to the target customer profile (ie, the customer profile 21'), as described in step S209. The customer's credit history is linked to the bank server 1. In this embodiment, the joint information has a credit card payment record, a loan record, an education level, a job category, and a job title corresponding to the customer corresponding to the target customer profile (ie, customer profile 21').

When the bank server 1 receives the joint information from the association server 6, the human face and the joint information according to the image data are used by the credit rating model 11 as described in step S210. And the external environmental indicator 13, generating a credit rating result.

In step S211, the bank server 1 determines whether the credit rating result is not less than a threshold level. If the result is yes, the flow proceeds to step S212, otherwise, the flow proceeds to step S213.

Step S212, the bank server 1 executes the bidding of the financial item related to the bid request. Following the previous example, the bank server 1 executes the credit card application.

In step S213, the bank server 1 transmits a second notification message related to the manual note to the second electronic device 4 for display. In this way, the bank staff can be notified immediately that the credit rating of the user is too low, so that the subsequent processing can reduce the risk of the bank.

In summary, the bank server 1 determines that the face profile data and the identity identification data respectively match the reference identity identification data 211 and the reference face profile data of the target client profile (ie, customer profile 21'). And when the credit rating model 11 is based on the face of the image data, the joint information, and the more favorable credit rating result generated by the external environment indicator 13 is not less than the threshold level, Allowing the application of the financial product, so as not only can effectively reduce the possibility of being fraudulently used by the customer, and the external features presented by the face have great opportunities and economic characteristics, so it can effectively reduce the risk of the bank. .

However, the above is only the embodiment of the present invention, and when it is not possible to limit the scope of the present invention, all the simple equivalent changes and modifications according to the scope of the patent application and the contents of the patent specification are still This new patent covers the scope.

100‧‧‧ online review system

1‧‧‧Bank server

11‧‧‧Credit rating model

12‧‧‧Face feature comparison model

13‧‧‧ External environmental indicators

2‧‧‧Customer Database

21,21’‧‧‧Customer Information

211‧‧‧Reference identification information

212‧‧‧Reference face profile data

213‧‧‧Bid history

3‧‧‧First electronic device

4‧‧‧Second electronic device

5‧‧‧Communication network

6‧‧‧Link server

S201~S213‧‧‧Steps

Other features and advantages of the present invention will be apparent from the embodiments of the present invention, wherein: FIG. 1 is a block diagram illustrating an embodiment of the present invention; and FIG. A flow chart exemplarily illustrates an online review method performed by the embodiment.

Claims (10)

An online review system comprising: a first electronic device operated by a user; a banking server provided by a bank and capable of communicating with the first electronic device and storing a pre-trained credit rating And a customer database connected to the bank server, the customer database storing a plurality of customer data respectively corresponding to the plurality of customers of the bank, each customer data having reference identity identification data and reference face features The first electronic device obtains image data related to a face and an identity information by the user, and transmits a bid request containing the image data and related to a financial product to the bank server. When the banking server receives the bid request from the first electronic device, the face feature data of the face and the identity identification data of the identity information are extracted from the image data, and according to the face feature One of the information and the identification information, one of the customer data is used as the target customer data; The line server determines whether the face feature data, the identity identification data respectively corresponds to the reference face feature data of the target customer profile and the reference identity identification data; wherein the bank server determines the face feature data And when the identification data is consistent with the reference identity identification data and the reference facial feature data of the target customer data, the credit rating model generates a credit rating result according to at least the face of the image data. And wherein the bank server executes the bidding of the financial product when it is determined that the credit rating result is not less than a threshold level. The online review system of claim 1, wherein the bank server determines the reference facial feature data of the one of the facial feature data and the customer data stored in the customer database. When the similarity between the two is not less than a first threshold, the one of the customer data is the target customer data. The online review system of claim 2, wherein the bank server further stores a pre-trained face feature comparison model and compares the face feature comparison model to obtain The similarity between the facial feature data and the reference facial features of the customer data. The online review system of claim 1, wherein: each customer data stored in the customer database further has a bid history data, and the bid history data has the customer's application corresponding to the customer data. At least one financial product; the bank server not only determines whether the facial feature data and the identification data respectively correspond to the reference identity identification data and the reference facial feature data of the target customer data, and further according to the target customer data The bid history data determines whether the financial item related to the bid request is the same as one of the financial products of the bid history data of the target customer data; and the bank server determines the face feature data and the The identification data is consistent with the reference identity identification data and the reference facial feature data of the target customer data, and the financial product related to the bid request is not one of the financial products of the bid history data of the target customer data. At the same time, according to the credit rating model, at least according to the image resource The face of the material, the result of the credit rating. The online review system of claim 4, wherein: the bank server is further communicable with a link server belonging to the joint center; the bank server determines the face feature data and the identity The identification data is consistent with the reference identity identification data and the reference facial feature data of the target customer data, and the financial product related to the bid request is not the same as one of the financial products of the target history data of the target customer data. Transmitting, by the link server, a data request including the reference identification data of the target customer data to the association server; the association server transmitting the related data request related to the target client when receiving the data request from the bank server And the bank server generates the credit according to the face of the image data and the joint information by using the credit rating model; Rating results. The online review system of claim 5, wherein the joint information has at least one of a credit card payment record, a loan record, an education level, a job category, and a job title corresponding to the customer corresponding to the target customer profile. . The online review system of claim 6, wherein: the bank server further stores an external environmental indicator including a current gross domestic product (GDP) and a glory signal. At least one of the bank servers and the bank rating model generate the credit rating result based on the face and the joint information of the image data and the external environment indicator. The online review system of claim 7, further comprising a second electronic device that can communicate with the bank server, wherein: the bank server determines that the facial feature data and the identification data are not different The reference identity identification data and the reference facial feature data corresponding to the target customer data, or the financial product related to the bid request is the same as one of the financial products of the bid history data of the target customer data, Corresponding to the first notification message of the manual note to the second electronic device for display; and the bank server transmitting a second notification message related to the manual note when determining that the credit rating result is less than the threshold level To the second electronic device for display. The online review system of claim 1, wherein the first electronic device obtains the image data by capturing a document including the face and the identity information. The online review system of claim 1, wherein the first electronic device obtains the first image data related to the identity information by capturing a document containing the identity information and by photographing the user The face is configured to obtain second image data related to the face, the image data including the first image data and the second image data.