TWI673673B - Voice control trading system - Google Patents

Abstract

An intelligent voice transaction system includes a processing host. The processing host includes a storage unit, a processor, and a transmission unit. The processor is electrically connected to the storage unit and the transmission unit. The transmission unit is used for receiving a text signal. The processor is configured to execute a semantic recognition module stored in the storage unit. The semantic recognition module matches text signals with a plurality of command options stored in a database of the storage unit, thereby generating a plurality of confidence indexes corresponding to the command options. The processor performs a response according to the confidence index, including (a) when one of the confidence indices is greater than a preset value, executing an action corresponding to a command option corresponding to the confidence index greater than the preset value.

Description

Intelligent voice trading system

The invention relates to an intelligent voice transaction system.

Generally speaking, users have to send instructions to bank hosts through terminal devices such as mobile phones and computers to perform business services provided by the bank. Traditionally, the user manually operates to input instructions according to the options displayed on the terminal device, and transmits the instructions to the bank host. However, for some specific users (such as the elderly), the manual operation is too complicated to use the business services provided by the bank smoothly. Furthermore, when the user already has a business purpose that he wants to perform, he still needs to choose one by one to complete the input of instructions in accordance with the various options displayed on the terminal device, which is time-consuming and labor-intensive.

It can be seen that the above existing methods obviously still have inconveniences and defects, and need to be improved. In order to solve the above-mentioned problems, the related fields have made every effort to find a solution, but a suitable solution has not been developed for a long time.

One aspect of the present invention provides an intelligent voice trading system. The intelligent voice trading system includes a processing host. The processing host includes a first storage unit, a first processor, and a first transmission unit. The first processor is electrically connected to the first storage unit and the first transmission unit. The first transmission unit is used for receiving a text signal. The first processor is configured to execute a semantic recognition module stored in the first storage unit. The semantic recognition module matches the text signal with a plurality of command options stored in a database of the first storage unit, thereby generating a plurality of confidence indexes corresponding to the command options. The first processor performs a response according to the confidence index, including (a) when one of the confidence indexes is greater than a preset value, executing an action corresponding to a command option corresponding to the confidence index greater than the preset value.

In one embodiment, the intelligent voice transaction system further includes an electronic device. The electronic device includes a voice input unit, a second storage unit, a second processor, and a second transmission unit. The second processor is electrically connected to the voice input unit, the second storage unit, and the second transmission unit. The second processor is configured to execute a voice analysis module stored in the second storage unit to analyze a voice signal received by the voice input unit to generate a text signal. The second transmission unit is configured to establish a communication link with the first transmission unit to transmit a text signal.

In one embodiment, the communication link is a network.

In one embodiment, an application program is stored in the second storage unit. The application is executed by the second processor. When the speech analysis module generates a text signal that is the same as a preset text, the generated text signal is transmitted, so that the text signal is transmitted to the processing host.

In an embodiment, the electronic device further includes a display unit. The second processor is electrically connected to the display unit. After the first processor performs the response (a), an execution result is returned to the electronic device. The display unit displays the execution result.

In one embodiment, an application program is stored in the second storage unit. The application is executed by the second processor. The electronic device further includes a playback unit. The second processor is electrically connected to the playback unit. After the first processor performs the response (a), an execution result is returned to the electronic device. The application generates a voice dial signal corresponding to one of the execution results. The playback unit plays the voice playback signal.

In an embodiment, the response further includes (b) when the confidence index is less than a preset value, returning 3 to 5 corresponding command options with higher values in the confidence index to the electronic device.

In one embodiment, an application program is stored in the second storage unit. The application is executed by the second processor. The electronic device further includes a playback unit. The second processor is electrically connected to the playback unit. After the first processor performs the response (b), the application program generates a voice dial signal corresponding to one of the command options returned. The playback unit plays the voice playback signal.

In an embodiment, the electronic device further includes an input unit. The second processor is electrically connected to the input unit. After the first processor performs the response (b), it transmits an input signal inputted through the input unit to the processing host to select the returned command option, so that the first processor executes the action of the selected command option.

In one embodiment, an application program is stored in the second storage unit. The application is executed by the second processor. After the first processor performs the (b) response, when the speech analysis module generates a text signal identical to a preset text, the generated text signal is transmitted, so that a command option for return is selected to make the first processing The controller performs the action of the selected command option.

The above description will be described in detail in the following embodiments, and further explanation will be provided for the technical solution of the present invention.

In order to make the description of the present invention more detailed and complete, reference may be made to the accompanying drawings and various embodiments described below. The same numbers in the drawings represent the same or similar elements. On the other hand, well-known elements and steps have not been described in the embodiments, so as to avoid unnecessary limitation to the present invention.

Please refer to FIG. 1. FIG. 1 is an intelligent voice transaction system 100 according to an embodiment of the present disclosure. As shown in FIG. 1, the intelligent voice transaction system 100 includes a processing host 120. The processing host 120 includes a first storage unit 122, a first processor 123, and a first transmission unit 124. In terms of connection, the first processor 123 is electrically connected to the first storage unit 122 and the first transmission unit 124. It should be understood that the first transmission unit 124 is configured to receive a text signal, and the first processor 123 is configured to execute the semantic recognition module 122a stored in the first storage unit 122. The semantic recognition module 122a matches text signals with a plurality of command options stored in the database 122b of the first storage unit 122, thereby generating a plurality of confidence indexes corresponding to the plurality of command options. In one embodiment, when one of the plurality of confidence indices is greater than a preset value, the first processor 123 executes an action corresponding to the instruction option corresponding to the confidence index greater than the preset value. The so-called "semantic recognition module" may be any program known in the art for recognizing the similarity between a text signal and a specific command option, and the so-called "confidence index" is a text signal and a specific command option Similarity value. That is, when the confidence index is higher, it means that the text signal is more similar to a specific command option. Further, a preset value may be set. When the confidence index is greater than the preset value, the first processor 123 may directly execute the action of the specific instruction option. It is worth mentioning that although any program known in the art can be used as the semantic recognition module to identify the closest command option, the closest command option may not be the business purpose intended by the user. If the first processor 123 rashly executes this closest command option directly, it will cause confusion to the user. However, with the preset value threshold mechanism proposed in the present disclosure, the closest command option is not executed when the confidence index is lower than the preset value, which can greatly improve the accuracy of identification.

The source of the text signal received by the first transmission unit 124 is also disclosed herein. The user must operate an electronic device 110 to generate the text signal, and send the text signal to the processing host 120 through the communication link L1. Specifically, the intelligent voice transaction system 100 further includes an electronic device 110. The electronic device 110 includes a voice input unit 111, a second storage unit 112, a second processor 113, and a second transmission unit 114. In terms of connection, the second processor 113 is electrically connected to the voice input unit 111, the second storage unit 112, and the second transmission unit 114. It should be understood that the second processor 113 is configured to execute the voice analysis module 112a stored in the second storage unit 112 to analyze a voice signal received by the voice input unit 111 to generate the text signal. The second transmission unit 114 is configured to establish a communication link L1 with the first transmission unit 124 to transmit the text signal.

In one embodiment, the electronic device 110 may be a personal computer, a notebook computer, a tablet computer, or a mobile phone. In an embodiment, the voice input unit 111 may be a microphone or a microphone. In one embodiment, the first storage unit 122 and the second storage unit 112 may be hard disks, flash memory, or other recording media. In an embodiment, the first processor 123 and the second processor 113 may be a central processing unit, a microcontroller, or other circuits. In one embodiment, the first transmission unit 124 and the second transmission unit 114 may be wireless transceivers, network cards or other communication devices. In one embodiment, the communication link L1 may be a network.

It should be understood that, in order to achieve a more convenient use purpose, an application program 112b may be installed in the electronic device 110 to perform voice control signal transmission. Specifically, the application program 112b is stored in the second storage unit 112. The user must perform operations to cause the second processor 113 to execute the application program 112b. In the case where the application program 112b is executed by the second processor 113, when the voice analysis module 112a generates a text signal identical to a preset text, the generated text signal is transmitted. For example, the "default text" may be "send". When the user speaks "balance inquiry, transmission", the voice input unit 111 receives the user's utterance (ie, a voice signal), and analyzes it through the speech analysis module 112a, thereby generating a text signal "balance inquiry, transmission". In this way, because the generated text signal contains the same "send" as the set "default text", the generated text signal "balance inquiry" is transmitted to the processing host 120.

In one embodiment, the electronic device 110 further includes a display unit 115. In terms of connection, the second processor 113 is electrically connected to the display unit 115. As described above, when the confidence index generated by the semantic recognition module 122a by matching the text signal with the plurality of command options is greater than a preset value, the first processor 123 executes the corresponding command option action. Further, after the first processor 123 executes the corresponding command option action, an execution result may be returned to the electronic device 110, so that the display unit 115 may display the returned execution result. For example, when the text signal "balance inquiry" sent to the processing host 120 matches the command option with a confidence index greater than a preset value through the semantic recognition module 122a, and the first processor 123 executes the action of the command option (ie, performs After checking the balance, the returned execution result (such as the living balance in the account) can be displayed by the display unit 115, so that the user knows the content of the returned execution result. In one embodiment, the display unit 115 may be a liquid crystal display, a cathode ray tube display, an electronic paper display, or other display devices.

In one embodiment, the electronic device 110 further includes a playback unit 116. In connection relationship, the second processor 113 is electrically connected to the playback unit 116. As described above, an application 112b can be installed in the electronic device 110. In one embodiment, the application program 112b is stored in the second storage unit 112 and is used to generate a voice play signal. In detail, when the application program 112b is executed by the second processor 113, the application program 112b can generate a voice dial signal corresponding to the execution result of the return, and the playback unit 116 can dial this voice dial signal. For example, if the execution result of the postback is the living balance in the account, the application program 112b may generate a voice dial signal (such as the voice of the amount of the living balance) according to the execution result of the postback. Therefore, the playback unit 116 can play the voice play signal, so that the user can know the content of the execution result of the return without having to watch the display screen. In an embodiment, the playback unit 116 may be a speaker or a speaker.

As described above, the semantic recognition module 122a matches the text signal with a plurality of command options in the database 122b, thereby generating a plurality of confidence indexes corresponding to the plurality of command options. In one embodiment, when the plurality of confidence indices are all smaller than a preset value, the corresponding three or five corresponding command options with higher values in the plurality of confidence indices are returned to the electronic device 110. That is, 3 to 5 command options that are most similar to the text signal are returned to the electronic device 110. Therefore, although multiple confidence indexes are lower than the preset value and the closest command option is not executed, the user can know the most similar 3 to 5 command options, and then can think about the next action, such as to execute One of these 3 to 5 command options, or perform another operation. In addition, as a way for the user to know the most similar 3 to 5 command options, it can be a visual mode or an auditory mode. For example, in an embodiment in which the electronic device 110 includes a display unit 115, the display unit 115 can display the three to five command options, so that the user can know the command options through a visual manner. Alternatively, in the embodiment where the electronic device 110 includes the playback unit 116, the application program 112b stored in the second storage unit 112 generates a voice dial signal corresponding to one of the 3 to 5 command options (eg, a voice explaining each command option) . Then, the playback unit 116 can play the voice-play signal, so that the user can know the command options through hearing.

In one embodiment, the electronic device 110 further includes an input unit 117. In connection relationship, the second processor 113 is electrically connected to the input unit 117. In an embodiment, the input unit 117 may be a touch device, a keyboard, a mouse, or other input elements. As described above, after the user knows the most similar 3 to 5 command options, the user can execute one of the 3 to 5 command options. In detail, the user has to input an input signal through the input unit 117 and send it to the processing host 120 to select the returned command option. For example, the user clicks 3 to 5 command options presented on the display unit 115 through the input unit 117, so that the input signal (the signal of the clicked command option) is transmitted to the processing host 120, so that the first processor 123 can perform the action of the selected command option.

Alternatively, the application 112b stored in the second storage unit 112 may be used to perform voice control signal transmission to select the returned command option. Specifically, when the application program 112b is executed by the second processor 113, when the speech analysis module 112a generates a text signal that is the same as a preset text, the generated text signal is transmitted, so that a return is selected. Command options. For example, the "default text" may be "send", and the most similar 3 to 5 command options are, for example, "balance inquiry", "transfer inquiry", and "payment inquiry". When the user speaks "balance inquiry, transmission", the voice input unit 111 receives the user's utterance (ie, a voice signal) and analyzes it through the voice analysis module 112a, thereby generating a text signal "balance inquiry, transmission". In this way, because the generated text signal contains the same "send" as the set "default text", the generated text signal "balance inquiry" is transmitted to the processing host 120, so that the first processor 123 can execute all The action of the selected command option (balance inquiry).

In order to describe the operation mode of the intelligent voice trading system 100 in detail, it will be described below with reference to FIG. 2A and FIG. 2B. FIG. 2A and FIG. 2B are flowcharts of the operation method 200 of the intelligent voice transaction system. It should be understood that the steps mentioned in FIG. 2A and FIG. 2B can be adjusted according to actual needs, except for those in which the sequence is specifically described. They can also be performed simultaneously or partially at the same time. Steps or omit some steps.

First, please refer to FIG. 1, FIG. 2A, and FIG. 2B at the same time. In step 201, the user opens an application 112 b (for example, a bank application) of the electronic device 110. For example, the user may click the icon 112b of the application 112b presented on the display unit 115 through the input unit 117, so that the second processor 113 executes the application 112b stored in the second storage unit 112. Next, in step 202, the voice input unit 111 of the electronic device 110 receives a voice signal. Next, in step 203, the second processor 113 executes the voice analysis module 112a to analyze the voice signal, thereby generating a text signal. Then, in step 204, the text signal is transmitted to the processing host 120 through the communication link L1. Subsequently, in step 205, the first processor 123 executes the semantic recognition module 122a to match the text signal with a plurality of command options in the database 122b, thereby generating a plurality of confidence indexes corresponding to the plurality of command options.

If one of the plurality of confidence indices is greater than a preset value, in step 206, the first processor 123 executes an action corresponding to the instruction option corresponding to the confidence index greater than the preset value, and returns an execution result to the electronic device. 110. It should be understood that, in an embodiment in which the electronic device 110 includes a display unit 115, the display unit 115 may display the execution result of the postback. Alternatively, in the embodiment where the electronic device 110 includes the playback unit 116, the application program 112b generates a voice dial signal corresponding to the execution result of the return, and the playback unit 116 can dial this voice dial signal.

On the other hand, if the plurality of confidence indexes are all smaller than the preset value, in step 207, 3 to 5 corresponding command options with higher values among the plurality of confidence indexes are returned to the electronic device 110. Similarly, in an embodiment where the electronic device 110 includes a display unit 115, the display unit 115 may display 3 to 5 command options returned. Alternatively, in the embodiment in which the electronic device 110 includes the playback unit 116, the application 112b generates a voice dial signal corresponding to 3 to 5 command options returned, and the playback unit 116 can dial this voice dial signal. Next, in step 208, the user can transmit the voice control signal through the voice input unit 111, or input the signal through the input unit 117, and send the signal to the processing host 120 to select the returned command option.

In summary, the intelligent voice transaction system of the present invention provides a systematic and appropriate intelligent voice transaction method, which helps to improve the complicated procedures of manual operations and makes it easier for some specific users (such as the elderly). Operations to complete their business purpose. In addition, through the voice dialing method, the user does not need to watch the display screen to know the execution result of the postback, so that some specific users (such as the visually impaired) can also perform operations. In addition, with this system, the user's business purpose will be executed more quickly and correctly, and the operation efficiency will be improved, avoiding the time-consuming problem of selecting one by one in accordance with the displayed various levels of options.

Although the present invention has been disclosed as above in the embodiments, it is not intended to limit the present invention. Any person skilled in the art can make various modifications and retouches without departing from the spirit and scope of the present invention. Therefore, the protection of the present invention The scope shall be determined by the scope of the attached patent application.

In order to make the above and other objects, features, advantages, and embodiments of the present invention more comprehensible, the description of the attached symbols is as follows: 100 intelligent voice trading system 110 electronic device 111 voice input unit 112 storage unit 112a voice analysis module 112b Application 113 processor 114 transmission unit 115 display unit 116 playback unit 117 input unit 120 processing host 122 storage unit 122a semantic recognition module 122b database 123 processor 124 transmission unit 200 method 201 to 208 step L1 communication link

In order to make the above and other objects, features, advantages, and embodiments of the present invention more comprehensible, the description of the drawings is as follows: FIG. 1 is a diagram of an intelligent voice transaction system 100 according to an embodiment of the present disclosure. FIG. 2A and FIG. 2B are flowcharts of an operation method 200 of an intelligent voice transaction system according to an embodiment of the present disclosure.

Claims (9)

An intelligent voice transaction system includes a processing host including a first storage unit, a first processor, and a first transmission unit, wherein the first processor is electrically connected to the first storage unit and the first storage unit. A transmission unit, the first transmission unit is configured to receive a text signal, the first processor is configured to execute a semantic recognition module stored in the first storage unit, and the semantic recognition module performs the text signal and stored in the first Matching of a plurality of command options in a database of a storage unit, thereby generating a plurality of confidence indices corresponding to the command options; and an electronic device including a voice input unit, a second storage unit, and a second process And a second transmission unit, wherein the second processor is electrically connected to the voice input unit, the second storage unit, and the second transmission unit, and the second processor is configured to execute storage in the second storage A voice analysis module of the unit analyzes a voice signal received by the voice input unit to generate the text signal, and the second transmission The first unit is configured to establish a communication link with the first transmission unit to transmit the text signal. The first processor performs a response according to the confidence indexes, including: (a) when one of the confidence indexes is greater than At a preset value, an action corresponding to a command option corresponding to the confidence index greater than the preset value is executed. For example, the intelligent voice trading system under the scope of patent application No. 1 in which the communication link is a network. For example, in the intelligent voice trading system under the scope of patent application, an application program is stored in the second storage unit, and the application program is executed by the second processor. When the voice analysis module generates the same text as a preset text, When the text signal is transmitted, the generated text signal is transmitted, thereby transmitting the text signal to the processing host. For example, the intelligent voice transaction system of the first patent application scope, wherein the electronic device further includes a display unit, the second processor is electrically connected to the display unit, and after the first processor performs the response (a), returns An execution result is transmitted to the electronic device, and the display unit displays the execution result. For example, in the intelligent voice trading system under the scope of patent application, an application program is stored in the second storage unit, the application program is executed by the second processor, the electronic device further includes a playback unit, and the second process The device is electrically connected to the playback unit, and after the first processor performs the response (a), an execution result is returned to the electronic device, and the application generates a voice dial signal corresponding to the execution result. The playback unit Play the voice play signal. For example, if the intelligent voice trading system under the scope of patent application is applied for, the response further includes: (b) when the confidence indices are less than the preset value, returning 3 to 5 of the higher confidence indices. The corresponding command option is to the electronic device. For example, in the intelligent voice trading system with the scope of patent application No. 6, an application program is stored in the second storage unit, the application program is executed by the second processor, the electronic device further includes a playback unit, and the second process The device is electrically connected to the playback unit. After the first processor performs a response (b), the application generates a voice play signal corresponding to one of the command options that should be returned, and the playback unit plays the voice play signal. . For example, the intelligent voice trading system under the scope of patent application No. 6, wherein the electronic device further includes an input unit, the second processor is electrically connected to the input unit, and after the first processor performs the (b) response, transmits An input signal is input to the processing host by one of the input units to select the returned command option, so that the first processor executes the action of the selected command option. For example, in the intelligent voice trading system under the scope of patent application 6, an application program is stored in the second storage unit, the application program is executed by the second processor, and after the first processor performs (b) reaction, When the speech analysis module generates a text signal that is the same as a preset text, the generated text signal is transmitted, so that the returned command option is selected, so that the first processor executes the action of the selected command option. .