TWM618718U - Intelligent inquiry system for inquiring check credit status - Google Patents

Abstract

An intelligent query system includes: a recognition module that converts a voice input information into a digital information to be recognized, and judges whether it is a checking account to output the digital information to be recognized; an interactive voice response module that stores multiple Each is related to a check status query result and corresponds to a voice response model in a preset mode, and outputs the digital information to be recognized, and outputs the voice input information to the recognition module; a bank-side server module, according to The to-be-identified digital information obtains a corresponding accounting information from a lookup table including a plurality of accounting information; and a data processing module generates an accounting mode based on the corresponding accounting information and outputs it to the interactive voice response module Group to compare the account mode with the preset modes to send out a voice reply message.

Description

Intelligent query system for querying the status of cheques

This new model relates to an inquiry system, especially an intelligent inquiry system for inquiring about the state of the cheque.

If the current bank customer wants to confirm the status of the cheque (for example, the account status corresponding to the cheque account, the number of refunds, the number of times the account has been registered, etc.), the customer can only call The bank's telephone incoming branch staff (or customer service staff) will make an inquiry.

However, there are many customers who find branch staff for telephone assistance. As a result, on the one hand, it will increase the workload of relevant branch staff on the bank side and consume more manpower; on the other hand, if customers choose to directly seek assistance from branch staff, Each customer will have to wait a long time on the phone before being answered by the branch staff. This will cause customers to spend more time waiting, and it is not convenient for customers to check the status of the cheque in real time. According to this, the existing inquiry mechanism of the status of the check ticket is indeed necessary for improvement.

Therefore, the purpose of the present invention is to provide an intelligent query system that can immediately query the status of the cheque without the assistance of branch staff.

Therefore, the intelligent inquiry system for inquiring about the status of the cheque includes an identification module, an interactive voice response module, a bank-side server module, and a data processing module.

The recognition module receives a customer's voice input information related to the check status query, and stores a plurality of pre-trained account classification models. The recognition module converts the voice input information into a digital information to be recognized , And identify the digital information to be recognized according to the account classification models to determine whether the digital information to be recognized is a checking account number, and output the digital information to be recognized when the result of the determination is yes.

The interactive voice response module stores a plurality of pre-trained voice response models, and is communicatively connected to the recognition module to receive and output the digital information to be recognized, and is used to receive the voice input information and output to the recognition module, Each voice response model is related to a check status query result and corresponds to a preset mode.

The bank-side server module is communicatively connected to the interactive voice response module to receive the digital information to be recognized, and stores a look-up table that records multiple check accounts and a check status corresponding to each check account. The bank-side server module obtains and outputs corresponding accounting information from the accounting information in the look-up table according to the digital information to be identified.

The data processing module is communicatively connected to the interactive voice response module and the bank-side server module, receives the corresponding accounting information from the bank-side server module, and generates an accounting mode based on the corresponding accounting information, and Output the account mode to the interactive voice response module, so that the interactive voice response module compares the account mode with the preset modes to obtain a corresponding one of the preset modes, And according to the voice response model corresponding to the corresponding one of the preset modes, a voice response message is generated and sent to the customer.

The effect of this new model is: through the intelligent query system, customers can immediately query the status of the check in a self-service manner, and do not need to wait for the branch staff to answer the status of the check. In this way, the intelligent query The system is more convenient for customers to use, and can reduce the workload of branch staff and save the bank's human resources.

Referring to Figure 1, the embodiment of the intelligent query system for querying the status of the check status of the present invention includes a recognition module 1, an interactive voice response (IVR) module 2, and a bank-side server module 3 , And a data processing module 4.

The recognition module 1 receives a customer's voice input information Vi related to the check status query, and stores a plurality of pre-trained account classification models. The recognition module 1 converts the voice input information Vi into a to-be-recognized digital information Ti, and recognizes the to-be-recognized digital information Ti according to the account classification models to determine whether the to-be-recognized digital information Ti is a checking account number. And when the judgment result is yes, output the to-be-identified digital information Ti.

In this embodiment, the recognition module 1 includes a speech recognition unit 11 and a semantic recognition unit 12. The speech recognition unit 11 is an automatic speech recognition (ASR) unit for performing speech to text recognition (Speech To Text, STT). The semantic recognition unit 12 is a natural language processing unit (NLP).

The voice recognition unit 11 receives the voice input information Vi, and converts the voice input information Vi into the to-be-recognized digital information Ti and outputs it.

The semantic recognition unit 12 stores the account classification models, and is communicatively connected to the voice recognition unit 11 to receive the digital information Ti to be recognized, and recognize the digital information Ti to be recognized according to the account classification models to determine Whether the to-be-identified digital information Ti is the checking account number. When the judgment result is yes, the semantic recognition unit 12 outputs the to-be-recognized digital information Ti. In addition, the semantic recognition unit 12 also stores a pre-trained text query information Ta that instructs the customer to provide again. When the semantic recognition unit 12 determines whether the to-be-recognized digital information Ti is the checking account and the judgment result is no, the semantic recognition unit 12 outputs the text query information Ta to the interactive voice response module 2.

The interactive voice response module 2 stores a plurality of pre-trained voice response models, and is communicatively connected to the voice recognition unit 11 and the semantic recognition unit 12 of the recognition module 1. The interactive voice response module 2 is configured to receive the voice input information Vi and output it to the voice recognition unit 11, and receive the to-be-recognized digital information Ti or the text query information Ta from the semantic recognition unit 12. When receiving the to-be-recognized digital information Ti, the interactive voice response module 2 outputs the to-be-recognized digital information Ti to the bank-side server module 3. When receiving the text query message Ta, the interactive voice response module 2 generates and sends a voice query message Va to the client according to the text query message Ta. Each voice response model is related to a check status query result and corresponds to a preset mode.

The bank-side server module 3 is communicatively connected to the interactive voice response module 2 to receive the to-be-identified digital information Ti, and stores a look-up table. The look-up table records multiple checking accounts and a piece of accounting information related to the state of the check letter corresponding to each checking account. The bank-side server module 3 obtains and outputs corresponding accounting information Ai from the accounting information in the look-up table according to the digital information Ti to be identified. In this embodiment, the accounting information corresponding to each checking account includes the account opening date of an account, the current status of the account (for example, normal, refusing to settle, etc.), the number of refunds of the account and the reason for the refund , And the number of times and reasons for the account, but not limited to this.

The data processing module 4 is communicatively connected to the interactive voice response module 2 and the bank-side server module 3, and receives the corresponding accounting information Ai from the bank-side server module 3. The data processing module 4 generates an accounting mode Am based on the corresponding accounting information Ai, and outputs the accounting mode Am to the interactive voice response module 2, so that the interactive voice response module 2 also returns the account The mode Am is compared with the preset modes to obtain a corresponding one of the preset modes, and a voice response is generated and issued according to the voice response model corresponding to the corresponding one of the preset modes Information Vr to the customer.

For example, when a customer wants to use the intelligent query system of the present invention to query the status of a check (that is, a check to be queried), the customer first dials the customer service line of the bank company corresponding to the check to enter the line and start Voice dialogue, so that the interactive voice response module 2 receives the voice input information Vi (that is, what the customer says, such as the check’s 12-code account number 123-AB-456789-0, where the symbol AB corresponds to the check A set of fixed coded numbers for the bank company). The interactive voice response module 2 transmits the voice input information Vi to the voice recognition unit 11, so that the voice recognition unit 11 converts the voice input information Vi into the to-be-recognized digital information Ti, and outputs it to the semantic recognition unit 12. Then, the semantic recognition unit 12 determines whether the digital information Ti to be recognized is the checking account number according to the account classification models. For example, it is judged whether the digital information Ti to be identified is 12 codes, and whether the 4th and 5th codes conform to the group of fixed codes AB? If the to-be-identified digital information Ti is 12 codes, and the fourth and fifth codes match the fixed-coded numbers AB, then the to-be-identified digital information Ti is the check account number; otherwise, if the to-be-identified digital information Ti is not When the number is 12, or the 4th and 5th codes do not match the number AB of the set of fixed codes, the digital information Ti to be identified does not belong to the checking account).

When the judgment result is that the to-be-recognized digital information Ti is the checking account number, the semantic recognition unit 12 transmits the to-be-recognized digital information Ti to the bank-side server module 3 via the interactive voice response module 2. The bank-side server module 3 compares the to-be-identified digital information Ti with the checking account numbers in the look-up table to obtain a counterpart of the same account number as the to-be-identified digital information Ti from the checking accounts. And a piece of accounting information corresponding to the counterpart in the checking account is used as the corresponding accounting information Ai and output to the data processing module 4. The corresponding accounting information Ai includes, for example, the account opening date (97/06/10) of an account corresponding to the digital information Ti to be identified, the current status of the account (normal), and the number of refunds of the account (2 times) And the reason for the refund (due to insufficient deposit), and the number of times the account has been registered (2 times) and the reason for the registration (due to insufficient deposit), etc. Then, the data processing module 4 is classified into different modes according to the corresponding accounting information Ai to generate an accounting mode Am and output to the interactive voice response module 2 (for example, the corresponding accounting information Ai is classified The category is mode 1, that is, the accounting mode Am is "mode 1"). Finally, the interactive voice response module 2 compares the account mode Am with the preset modes to obtain a corresponding one of the preset modes, and according to the corresponding one of the preset modes The corresponding voice response model generates and sends out the voice response message Vr. For example, the interactive voice response module 2 stores a first voice response model (e.g., the checking account that you note has been refunded twice in the past year and has been marked twice in the past year), and a second voice response model (e.g. , The checking account you have noted, no recent transactions), and a third voice response model (for example, the checking account you have noted is listed as a rejected current account), and these first to third voice response models Corresponding to a preset mode one, a preset mode two, and a preset mode three respectively. In this embodiment, since the account mode Am is "mode one" and corresponds to the preset mode 1, the interactive voice response module 2 generates and sends out the voice response message Vr according to the first voice response model. The customer (that is, the content that the interactive voice response module 2 broadcasts to the customer is "The checking account you have noted has been refunded twice in the past year and has been marked twice in the past year").

When the judgment result is that the digital information Ti to be recognized is not the checking account number, the semantic recognition unit 12 outputs the text query information Ta (for example, please provide the account number again) to the interactive voice response module 2. Then, the interactive voice response module 2 generates and sends out the voice query information Va to the customer according to the text query information Ta (that is, the content that the interactive voice response module 2 broadcasts to the customer is "please provide the account again") , So that the customer re-provides the account number and the intelligent inquiry system re-inquires.

In summary, with the intelligent query system, customers can immediately query the status of the check in a self-service manner. There is no need to wait for the branch staff to answer the phone to query the status of the check as in the prior art. , The intelligent query system of the present invention is more convenient for customers to use, and can reduce the workload of branch staff and save the human resources of the bank.

However, the above are only examples of the present model, and should not be used to limit the scope of implementation of the present model, all simple equivalent changes and modifications made in accordance with the patent scope of the present model application and the contents of the patent specification still belong to This new patent covers the scope.

1: Identification module 11: Voice recognition unit 12: Semantic Recognition Unit 2: Interactive voice response module 3: Bank-side server module 4: Data processing module Ai: Corresponding accounting information Am: Accounting model Ta: text asking for information Ti: Digital information to be recognized Va: Voice asking for information Vi: Voice input information Vr: voice reply message

The other features and effects of the present invention will be clearly presented in the embodiments with reference to the drawings, among which: Fig. 1 is a block diagram illustrating an embodiment of the intelligent query system for querying the status of a check of the present invention.

1: Identification module

11: Voice recognition unit

12: Semantic Recognition Unit

2: Interactive voice response module

3: Bank-side server module

4: Data processing module

Ai: Corresponding accounting information

Am: Accounting model

Ta: text asking for information

Ti: Digital information to be recognized

Va: Voice asking for information

Vi: Voice input information

Vr: voice reply message

Claims (7)

An intelligent query system for querying the status of the cheque, including: A recognition module that receives a voice input information related to the check status query from a customer, and stores a plurality of pre-trained account classification models. The recognition module converts the voice input information into a number to be recognized And identify the digital information to be recognized according to the account classification models to determine whether the digital information to be recognized is a checking account number, and output the digital information to be recognized when the judgment result is yes; An interactive voice response module that stores a plurality of pre-trained voice response models, and is communicatively connected to the recognition module to receive and output the digital information to be recognized, and is used to receive the voice input information and output to the recognition module , Each voice response model is related to a check status query result and corresponds to a preset mode; A bank-side server module, which is communicatively connected to the interactive voice response module to receive the digital information to be recognized, and stores a look-up table that records multiple check accounts and a related ticket letter corresponding to each check account. State accounting information, the bank-side server module obtains and outputs corresponding accounting information from the accounting information in the look-up table according to the digital information to be identified; and A data processing module is communicatively connected to the interactive voice response module and the bank-side server module, receives the corresponding accounting information from the bank-side server module, and generates an accounting mode based on the corresponding accounting information, And output the account mode to the interactive voice response module, so that the interactive voice response module compares the account mode with the preset modes to obtain a corresponding one of the preset modes , And generate and send a voice response message to the customer according to the voice response model corresponding to the corresponding one of the preset modes. The intelligent query system according to claim 1, wherein the identification module includes A voice recognition unit communicatively connected to the interactive voice response module to receive the voice input information, and convert the voice input information into the digital information to be recognized and output, and A semantic recognition unit stores the account classification models, and is communicatively connected to the voice recognition unit to receive the digital information to be recognized, and recognize the digital information to be recognized according to the account classification models to determine the to-be-recognized digital information Whether the digital information is the checking account number, and when the judgment result is yes, output the digital information to be identified. The intelligent query system according to claim 2, wherein the semantic recognition unit also stores a pre-trained text query message that instructs the customer to re-provide, and when the semantic recognition unit determines whether the digital information to be recognized is the When the judgment result of the checking account is no, the semantic recognition unit outputs the text query information to the interactive voice response module, so that the interactive voice response module also generates and sends a voice query message to the interactive voice response module based on the text query information customer. The intelligent query system according to claim 2, wherein the semantic recognition unit is a natural language processing unit. The intelligent query system according to claim 1, wherein the accounting information corresponding to each checking account includes an account opening date and current status. The intelligent query system according to claim 1, wherein the accounting information corresponding to each checking account includes the number of times an account has been refunded and the reason for the refund. The intelligent query system according to claim 1, wherein the accounting information corresponding to each checking account includes the number of times an account has been noted and the reason for the note.

Images