JPH10254935A - Answer sequence production method and device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily and quickly produce a desired answer sequence with no mistake by producing a new answer sequence based on the necessary one of prescribed number of functions and setting the detailed data to each function for execution of these functions included in the new answer sequence. SOLUTION: An answer sequence is newly produced in response to each voice answer job to be built in a voice answer system by means of the prescribed number of functions which are common to the answer sequences prepared in a function memory 3. This produced answer sequence including plural functions is written and set into an answer sequence buffer 7 via an input device 8. Then the detailed data peculiar to the answer sequence are set in each function included in the answer sequence. Thus, a necessary answer sequence is easily and quickly produced with no mistake.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice response system for identifying information including voice input from the outside and performing various voice response services specified by the information. The present invention relates to a response sequence creation method and a response sequence creation device to be executed.

[0002]

2. Description of the Related Art For example, a telephone order in a mail order is automatically accepted, various information corresponding to a telephone inquiry is automatically answered, and various opinion polls are automatically executed. In a voice response system, it is necessary to voice-recognize the name, address, product name, and quantity reported by a customer and perform necessary data processing. Also,
In order to prompt the customer to confirm by repeating the name and address given by the customer, various words are synthesized and sent back. further,
A voice message prompting the customer to input the next number or word is synthesized and output.

[0003] Information input from the outside includes:
In addition to the above-mentioned voice, customer push button (PB)
There is a numerical value consisting of a DTMF signal by operation and a code such as “*” or “#” (hereinafter, these are abbreviated as PB values).

Therefore, for example, in a voice response service of "mail order", a response sequence (program) indicating exchange of various voices and PB values exchanged with a customer.
Must be incorporated in the voice response system in advance.

[0005] In each step, a response sequence necessary for executing one voice response service includes a step of recognizing a voice inputted by a customer and converting the voice into one piece of text data or numerical data. It is determined whether the data is the requested data.
I say, "Speak." If the data is not the requested data, a re-input word of "input error, please input again" is synthesized and sent back.

Further, if there is no input even after a certain permissible time has passed, a re-input voice is transmitted, and if the number of re-input voices exceeds the permissible number, this voice response task is forcibly terminated.

In general, a voice response system includes:
A voice data memory is provided. In the voice data memory, voice data of typical words used in a response sequence of each voice response service and voice data of each sound are stored as digital data. I have. Then, a procedure for reading necessary voice data from a voice data memory, converting the voice data into an analog voice signal, and outputting the voice is incorporated in a necessary portion of the response sequence for voice synthesis.

In order to execute a new voice response service for the voice response system, a new response sequence for realizing the voice response service is created and set in the voice response system.

[0009]

However, as described above, the response sequence includes a complicated procedure such as a process of reading voice data from the voice data memory and outputting the voice, and setting of an allowable time. Creating a new response sequence for realizing a response task required specialized knowledge of program creation and a great deal of time and effort. Further, since it is necessary to write a complicated processing step, there is a concern that the rate of occurrence of human error in program creation increases.

The present invention has been made in view of such circumstances, and by constructing a response sequence for realizing a voice response service with a plurality of functions commonly used for each response sequence, the response sequence Creation work can be summarized into two simple tasks: creation of a simple whole sequence using each function and data setting work to set detailed data for each function, even if there is no specialized knowledge on program creation It is an object of the present invention to provide a response sequence creation method and a response sequence creation device that can easily create a required response sequence without error and in a short time even by a person.

[0011]

According to a first aspect of the present invention, there is provided a method for creating a response sequence in a voice response system which is executed in accordance with a response sequence. A predetermined number of functions are stored and held, and when a new response sequence is created, a corresponding response sequence is created from the stored and held number of functions using necessary functions, and each function included in the created response sequence is executed. Is set in the corresponding function.

In the present invention, a data setting screen for setting various detailed data for executing the corresponding function for each function is prepared in advance, and the data setting screen corresponding to each function included in the created response sequence is prepared. The screen is displayed, and detailed data of the function is set using the data setting screen.

In the present invention, the detailed data of each function includes information for specifying a function to be executed next. According to the fourth aspect, the predetermined number of functions include a function of performing continuation processing even after information including voice is interrupted.

Further, in the apparatus for creating a response sequence in a voice response system which executes in accordance with the response sequence, a function memory for storing and holding a predetermined number of functions required to construct each response sequence, A screen memory for storing a data setting screen for setting various detailed data for executing the corresponding function for each,
Data setting screen display means for sequentially reading out data setting screens of respective functions constituting a response sequence created using a necessary function from a predetermined number of functions stored in the function memory from the screen memory, and sequentially outputting the data setting screens; When a data setting screen of each function is displayed, detailed data setting means for setting detailed data designated on the display screen to the corresponding function is provided.

In the method for creating a response sequence configured as described above, when newly creating each response sequence corresponding to each voice response task to be incorporated in the voice response system, a response sequence in which detailed procedures are incorporated from the beginning. , Instead of creating a predetermined number of functions that are common to each prepared response sequence. Each of these functions is a rough function, and it is necessary to set various detailed data in order to actually realize this function.

After a response sequence is created by incorporating a plurality of functions, detailed data unique to the response sequence is set for each function incorporated in the response sequence.

As described above, the response sequence can be summarized into two simple operations, ie, the operation of creating a simple overall sequence using a plurality of functions and the operation of setting detailed data for each function. Can be facilitated.

Since the data setting screen of each function is displayed and the detailed data of the function is set using the data setting screen, the setting operation of the detailed data for the function is simplified.

Further, since the detailed data of each function includes the information for designating the function to be executed next, it is not necessary to separately set information on the execution order of each function as a response sequence.

Further, by preparing a function for performing the continuation processing even after the information including the voice is interrupted, a predetermined processing operation can be performed even after the voice is interrupted.

Further, in the response sequence creating apparatus configured as described above, the response sequence creating operation can be facilitated similarly to the response sequence creating method according to the first and second aspects.

[0022]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a schematic configuration of a voice response system in which a response sequence creation device adopting the response sequence creation method of the present invention is incorporated.

This voice response system is composed of a kind of information processing device such as a computer.
It comprises a response sequence creation device 1 and a response device main body 2. The response device main body 2 executes, for example, each voice response service specified by information including voice input from the external telephone 3 in accordance with a response sequence corresponding to the relevant service. on the other hand,
The response sequence creation device 1 creates a response sequence for executing a voice response task performed by the response device body 2.

In the answering device main body 2, an external telephone 3
Of the information input from the, the voice is recognized by the voice input / output unit 15, converted into numerical data or text data, and input to the voice response task execution unit 16. Further, the PB value of the information input from the telephone 3 is converted into a numerical value or a code such as “*” or “#” by a PB receiving unit 17 having a function of a modem and input to the voice response task execution unit 16. You.

In the response sequence file 18, a plurality of response sequences corresponding to each voice response service for executing a plurality of executable voice response services in the response device main body 2 are stored.

The voice response service execution unit 16 reads a response sequence corresponding to the voice response service specified by the information input from the external telephone 3 from the response sequence file 18, and uses the telephone 3 according to the response sequence. Perform voice response tasks with existing customers. Then, the voice response task execution unit 16 transmits the task result to the external host computer 19 as necessary.

In the response sequence creating device 1,
1, a functional memory 3, an image memory 4, an audio data file 5, and the like are formed in, for example, an HDD (hard disk drive), and a response sequence file 6 and a response sequence are stored in a main memory such as a RAM. A buffer 7 is formed. Further, in the response sequence creating device 1, input devices 8 such as a keyboard, a CRT display device 9, hardware members such as a GUI (graphical user interface) 10, and data formed as a program module on an application program. A setting screen display output unit 11, a detailed data setting unit 12, a response schedule confirmation unit 13, and a file transfer unit 14 are provided.

In the function memory 3, as shown in FIG.
For example, for each of the 12 basic functions 20 constituting each response sequence for executing each of the voice response tasks that can be set in this voice response system, details 21 of the corresponding function and the actual execution of the corresponding function The setting screen number 22 of the data setting screen used when setting the detailed data of is registered.

This basic function is, as shown in the figure,
"Specified voice reproduction""Synthetic voice reproduction""Branch""PBreception""FAXreception""FAXtransmission""DBsearch""DB
Basic functions for executing voice response tasks such as registration, calling, transfer, and disconnection are shown.

Note that this basic function alone is not sufficient, and it is necessary to set individual detailed data for each function for each voice response task. For example, the function of "reproduce specified voice" is "reproduce specified voice", but it is necessary to set actual voice data such as "Thank you" as detailed data.

The screen memory 4 stores a data setting screen for setting detailed data of a corresponding function for each setting screen number 22 of each function 20 registered in the function memory 3. For example, FIG. 3A shows the data setting screen 23 of the function “PB reception”.

As detailed data to be set for the function 20 of “PB reception”, a parameter number 24 for specifying the function 20 when incorporated in a response sequence,
A comment 25 consisting of voice data indicating a response word when a PB value is received, an allowable number of digits 26 of the PB value, and a function 2 to be executed after the reception process for the PB value is completed
The next parameter number 27 specifying 0, the timeout period 28 when no PB value is input, the error message number 29 that tells the customer when a timeout or an incorrect PB value is input, the allowable output of error message There are a number of times 30, a parameter number 31 for specifying the function 20 to be executed when the error message exceeds the allowable output number, and a parameter number 32 for specifying the function 20 to be executed when the disconnection signal is detected.

As detailed data to be set for the "line disconnection" function 20, as shown in FIG. 3B, a parameter number 24 and a comment 25 including voice data of "line disconnection" are provided. There is.

In the voice data memory 5 of FIG. 1, voice data of typical words used in the response sequence of each voice response service and voice data of each sound are stored as digital data.

In the response sequence buffer 7,
The response sequence currently being created is temporarily stored. Further, in the response sequence file 6, a plurality of areas 6a for storing the response sequences created by the response sequence creating device 1 are formed.

Next, when the response sequence creation device 1 is used to create a new voice response service response sequence, the work procedure of the work performed by the creator, and the response sequence creation device 1 itself is an application program. Will be described with reference to the flowchart shown in FIG.

In S (step) 1, the creator of the response sequence firstly registers each function 20 registered in the function memory 3 with a general response sequence of the voice service to be newly realized in the voice response system. Create using
The response sequence is created by the creator by writing it in a notebook or the like.

FIG. 5 shows, for example, mail order (shopping).
4 is a schematic response sequence created on a notebook by the creator to execute the voice response task. First, a function of receiving a call (P1), a function of receiving a PB value indicating various types of voice response services, and a function of determining (P2) a function of determining the type of voice response services to be executed according to the PB value. (P
3) If it is determined that the voice response service is a mail order (shopping) service (P4), a designated voice reproduction function indicating the start of the corresponding service is executed (P5).

Next, the voice of the address and name from the customer is received (P6), voice recognition is performed (P7), and the confirmation voice of "I know" for the recognized address and name is synthesized and output. (P8). When a reply to this confirmation voice is received by voice (P9), voice recognition is performed (P10), and the function of the disconnection process is performed (P11). After that, the received address,
The name data is processed in the memory (P12).

When the creator prepares a general response sequence shown in FIG. 5 composed of a combination of a plurality of functions 20, the writer writes the response sequence into the response sequence buffer 7 via the input device 8 for the time being. Specifically, a function number for specifying each function is sequentially keyed in. Therefore, in this state, the detailed data shown in FIG. 3 described above is not set for each function 20 constituting the response sequence written in the response sequence buffer 7.

When a general response sequence including a plurality of functions 20 for which detailed data is not set is written into the response sequence buffer 7, the data setting screen display output unit 11 is activated in S2 of FIG. 5, one function 20 in the general response sequence shown in FIG. 5 stored in the response sequence buffer 7 is read out, the data setting screen 23 of the corresponding function 20 is read out from the screen memory 4, and displayed on the CRT display device 9 (S3). ).

Then, the creator specifies the above-described detailed data on the data setting screens 23 displayed on the CRT display device 9 using the GUI 10 on the screens, for example, as shown in FIGS. 3A and 3B. (S4). Finally, the next parameter number 27 of the function 20 to which the next transition is performed and its own parameter number 24 are designated. Then, the detailed data setting unit 12 is activated, and sets the detailed data specified by the creator on the display screen in the corresponding function 20 of the response sequence buffer 7 (S5). In addition, comment 2 of detailed data
For 5, the corresponding audio data is read from the audio data memory 5 and set.

When the detailed data setting operation for one function 20 is completed, if there is a function 20 in the response sequence buffer 7 for which detailed data has not yet been set in S6, the flow returns to S2, and the details in the response sequence are returned. The next one function 20 for which data has not yet been set is read, and the setting processing of the detailed data is restarted.

When the detailed data setting operation for all the functions 20 constituting the response sequence is completed in S6, the response sequence confirmation unit 13 is activated in S7, and the created response sequence is stored in the response sequence buffer. 7 and displayed as a confirmation screen 33 on the display screen of the CRT display device 9 as shown in FIG.

When the creator confirms that the response sequence on the displayed confirmation screen 33 is appropriate (S
8) The created response sequence is written to an empty area 6a of the response sequence file 6 where no response sequence has been set (S9).

In S10, if there is another response sequence of the voice response service to be created, the process returns to S1.
The creation of a general response sequence using each function 20 is started.

In step S10, the creation of all necessary response sequences is completed.
When the writing operation to the response sequence file is completed, the file transfer unit 14 is activated in S11 to transfer all the response sequences stored in the response sequence file 6 to the response sequence file 18 of the response device body 2 (step S11). S12).

Next, the creator operates the response device main body 2 in S13 to confirm that each response sequence is correctly set in the response sequence file 18. When the above operations are completed, the creation and setting of a response sequence for a new voice response service for the voice response system is completed.

When a PB value or voice is input via the external telephone 3 in S14, a response sequence corresponding to the voice response task specified by the PB value or voice is executed in S15. I do. In the course of executing this response sequence, if it becomes necessary to synthesize new voice data for confirming the input name or the like, the voice data memory 5
The voice data is read directly from.

Next, a description will be given of features of the response sequence creating method and the response sequence creating apparatus thus configured. First, when newly creating each response sequence corresponding to each voice response task incorporated in the voice response system,
The creator does not create a response sequence in which detailed procedures are built in from the beginning, but uses a predetermined number of functions 20 common to each response sequence prepared in advance in the function memory 3 to schematically show in FIG. Create a response sequence for Then, a response sequence is first created by incorporating a plurality of functions and set in the response sequence buffer 7. next,
Detailed data unique to the response sequence is set for each function 20 incorporated in the response sequence.

As described above, the response sequence for executing the voice response task is a simple operation of creating a simple overall sequence using each function 20 and a data setting operation of setting detailed data for each function 20. Since the operations can be integrated into two operations, the operation of creating the response sequence can be facilitated.

The data setting screen 23 of each function 20 is displayed on the CRT display device 9 and the data setting screen 23 is displayed.
Since the detailed data of the corresponding function 20 is set using the GUI 10, the operation of setting the detailed data for the function 20 is simplified.

Further, detailed data of each function 20 is shown in FIG.
As shown in (a), since the next parameter number 27 specifying the function 20 to be executed next is included, it is not necessary to separately set information on the execution order of each function as a response sequence.

Next, in the function 20, as shown in the data setting screen 23 of FIG. 3A, the next parameter number 32 at the time of detecting the disconnection signal is set as detailed data. The reason will be described.

In a general voice response system, when it is detected that the other party of the telephone has disconnected the communication line, the own side also disconnects the communication line. However, in some cases, it is necessary to perform some processing on the self side after the disconnection of the other party is detected. Therefore, as shown in FIG.
By providing the next parameter number 32 indicating the transition destination in the function at the time of disconnection as one of the detailed data in 0, processing after disconnection detection can be set.

For example, a customer inputs a telephone number, an age, and a gender by a PB value, and finally inputs a name by voice.
The voice response task of registering in the database of No. 9 is verified.

In this case, there is a concern that the customer immediately disconnects the telephone line after inputting the name. for that reason,
If the answering device main body 2 disconnects the communication line for the customer after the disconnection signal is detected, and terminates the voice response process for the customer forcibly, all the data of the customer input up to that point is wasted. There is a concern that it will. Further, each time it is attempted to register, there is a concern that the processing speed of the entire voice response task may be reduced because it is necessary to register each of the four data items individually in the database.

Therefore, as shown in FIG. 7, by performing processing for registering these four data after detecting the disconnection signal, all the input data is registered in the database without lowering the processing speed. It becomes possible.

[0059]

As described above, in the response sequence creating method and response sequence creating apparatus of the present invention, the response sequence for realizing the voice response service is provided by a plurality of functions commonly used for each response sequence. Building.

Therefore, the operation of creating a response sequence can be summarized into two simple operations: an operation of creating a simple whole sequence using each function and a data operation of setting detailed data for each function. Even a creator who does not have specialized knowledge of program creation can easily create a required response sequence without error and in a short time.

Further, since the data setting screen of each function is displayed and the detailed data of the corresponding function is set using this data setting screen, the work of setting the detailed data for the function is simplified, and the response sequence Creation work efficiency can be further improved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a voice response system incorporating a method and a device for generating a response sequence according to an embodiment of the present invention;

FIG. 2 is a diagram showing the contents stored in a functional memory in the response sequence creating device according to the embodiment;

FIG. 3 is a diagram showing the contents stored in a screen memory in the response sequence creation device of the embodiment.

FIG. 4 is a diagram showing a response sequence creation operation procedure in the response sequence creation method according to the embodiment;

FIG. 5 is an exemplary view showing a schematic response sequence created by a creator according to the response sequence creating method of the embodiment;

FIG. 6 is an exemplary view showing a confirmation screen in the response sequence creation device of the embodiment.

FIG. 7 is a flowchart illustrating a response process after disconnection of a communication line in the response sequence creation device according to the embodiment;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Response sequence creation device 2 ... Response device main body 3 ... Function memory 4 ... Screen memory 5 ... Voice data memory 6, 18 ... Response sequence file 7 ... Response sequence buffer 9 ... CRT display device 10 ... GUI 11 ... Data setting screen Display output unit 12 ... Detailed data setting unit 13 ... Response sequence confirmation unit 14 ... File transfer unit 15 ... Speech input / output unit 16 ... Speech response service execution unit 17 ... PB reception unit 19 ... Host computer 20 ... Function 23 ... Data setting screen 33… Confirmation screen

Claims (5)

[Claims] 1. A method for creating a response sequence in a voice response system that executes according to a response sequence, wherein a predetermined number of functions necessary for constructing each response sequence are stored and stored. A response sequence is created by using necessary functions from the predetermined number of functions stored and held, and detailed data for executing each function included in the created response sequence is set in the corresponding function. How to create a response sequence. 2. A data setting screen for setting various detailed data for executing the function corresponding to each function is prepared in advance, and the data setting screen corresponding to each function included in the created response sequence is prepared. 2. The method according to claim 1, further comprising the step of displaying and setting detailed data of the function using the data setting screen. 3. The detailed data of each function includes information designating a function to be executed next.
Or a method for creating a response sequence according to 2. 4. The response sequence according to claim 1, wherein the predetermined number of functions include a function of performing continuation processing even after the information including the voice is interrupted. How to create 5. An apparatus for creating a response sequence in a voice response system that executes according to a response sequence, comprising: a function memory for storing and retaining a predetermined number of functions necessary for constructing each of the response sequences; A screen memory for storing a data setting screen for setting various detailed data for executing a function, and a response sequence created using necessary functions among a predetermined number of functions stored in the function memory Data setting screen display means for sequentially reading the data setting screen of each function from the screen memory and sequentially outputting the data setting screen; and displaying the detailed data designated on this display screen while the data setting screen of each function is displayed. An apparatus for creating a response sequence, comprising: a detailed data setting means for setting the function.