JP3608449B2 - Voice response method and apparatus, and storage medium storing voice response program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a voice response method and apparatus and a storage medium storing a voice response program, and more particularly to a voice response method and apparatus for connecting to a communication network and responding to a caller and a voice response program. The present invention relates to a storage medium.
[0002]
[Prior art]
FIG. 10 shows a configuration of a conventional response device.
The conventional response device includes a line interface unit 1, an incoming call detection unit 2, a call unit 4, a response unit 5, and a control unit 6.
In the response device shown in the figure, when an incoming call is received at the line interface unit 1, the incoming call detection unit 2 detects the incoming call, notifies the control unit 6, displays the incoming call, and the operator responds at the call unit 4.
[0003]
Alternatively, the response unit 5 is provided instead of the operator at night for unmanned operation and responds by voice guidance and push button dial signal recognition.
Further, in an electronic exchange system, a method of changing a response message by an originating ID provided by the ISDN network (Japanese Patent Laid-Open No. 6-237296), or a facsimile apparatus capable of searching for a user by the originating ID and determining service contents (Japanese Patent Laid-Open No. 9-65088).
[0004]
[Problems to be solved by the invention]
However, in the conventional response device, there is a problem that the response format on the receiving side at the time of unattended response is limited, and it is impossible to respond in the response format desired by the user. Moreover, there are many selection branches by PB, and there is a problem that it takes time to reach the service requested by the user.
[0005]
Further, when all incoming calls are transferred to the reception at night, there is a problem that the operation of the operator is large at night.
Further, there are a conventional method of changing a response message based on the calling ID (Japanese Patent Laid-Open No. 6-237296) and a facsimile apparatus (Japanese Patent Laid-Open No. 9-65088) capable of searching for a user based on the calling ID and determining service contents. In order to apply voice recognition to a large number of speakers, it is necessary for the user to input an ID number that identifies the speaker. In addition, it is necessary to specify a speaker by using a part of the speaker's utterance and collating with a recognition pattern for each of a large number of accumulated speakers.
[0006]
The present invention has been made in view of the above points. In the case where it is determined that the speaker is not suitable for speech recognition by improving the recognition rate of speech recognition, making the response format and response method variable for each speaker. Another object of the present invention is to provide a voice response method and apparatus capable of reducing the operation time by not attempting voice recognition, and a storage medium storing a voice response program.
[0007]
[Means for Solving the Problems]
FIG. 1 is a diagram for explaining the principle of the present invention.
The present invention (Claim 1) is a voice response method for replying to a caller by voice connected to a communication network.
For each caller, a response history including the result of past speech recognition success / failure, the number of responses, and the response format at the time of response is retained (step 1).
When there is an incoming call from the caller (step 2), the response history corresponding to the caller is searched (step 3),
In the response history corresponding to the retrieved caller, when there is a continuous history of successful voice recognition more than a certain number of times (step 4), a response by voice recognition is performed (step 5).
[0008]
The present invention (Claim 2) maintains a response history for each caller ID,
When there is an incoming call from the caller, obtain the caller ID of the caller,
Search the response history using the caller ID as the caller-specific key,
A response format is determined based on the retrieved response history.
In the present invention (Claim 3), as a response format,
A combination of voice recognition and voice guidance, a combination of push button dial signal recognition and voice guidance, a combination of voice guidance and call recording, voice guidance, an incoming call, and a response at the transfer destination are used.
[0009]
The present invention (Claim 4) allows the caller to specify the response format when the calling ID is not registered in the response history,
For an incoming call with a caller ID, response history information corresponding to the caller ID is acquired from the response history, and the response format is determined based on the response frequency history.
The present invention (Claim 5) changes the response format when there is no sound for a predetermined time at the time of response, or changes the response format due to the success / failure of voice recognition, or the response format by the operation of the caller One or a plurality of control methods are performed.
[0010]
FIG. 2 is a principle configuration diagram of the present invention.
The present invention (Claim 6) is a voice response device connected to a communication network for responding with voice to a caller,
Response format history storage means 11 for holding a response history including the result of past voice recognition success / failure for each caller, the number of responses, and the response format at the time of response;
Incoming call detection means 2 for detecting an incoming call from a caller;
In the incoming call detection means 2, when there is an incoming call, the history search means 6 for searching the response history corresponding to the caller from the response format history storage means 11,
In the response history corresponding to the caller searched by the history search means 6, there is a response means 5 that makes a response by voice recognition when there is a continuous history of voice recognition more than a predetermined number of times.
[0011]
According to the present invention (Claim 7), the response format history storage means 11
Keep a response history for each caller ID,
In the history search means 6,
Means for obtaining a caller ID of the caller when there is an incoming call from the caller, and searching for a response history using the caller ID as a caller-specific key;
In response means 5,
A means for determining a response format based on the response history searched by the history search means;
[0012]
In the present invention (claim 8), as a response format of the response history,
A combination of voice recognition and voice guidance, a combination of push button dial signal recognition and voice guidance, a combination of voice guidance and call recording, voice guidance, and an incoming call are transferred, and one of the responses at the transfer destination is held.
The present invention (Claim 9) further includes response format designation instruction means for causing the caller to designate a response format when the calling ID is not registered in the response history.
In response means 5,
For an incoming call with a caller ID, means for determining a response format based on the history of the number of responses searched by the history search means 6 is included.
[0013]
The present invention (Claim 10) further includes silent time detecting means for detecting silence for a predetermined time when responding to the caller,
In response means 5,
Means for changing the response format when the silent time detection means is silent for a predetermined time, means for changing the response format due to success / failure of voice recognition, or means for changing the response format by the operation of the caller Or a plurality of means are executed.
[0014]
The present invention (Claim 11) is a storage medium that stores a voice response program that is connected to a communication network and is mounted on a voice response device for responding to a caller with voice.
A response format history storage process for accumulating in the storage means a response history including the results of past voice recognition success / failure for each caller, the number of responses, and the response format at the time of response;
An incoming call detection process to detect incoming calls from callers;
In the incoming call detection process, when there is an incoming call, a history search process for searching a response history corresponding to the caller from the storage means;
In the response history corresponding to the caller searched by the history search process, there is a response process for performing a response by voice recognition when there is a continuous history of voice recognition more than a predetermined number of times.
[0015]
According to the present invention (Claim 12), in the response format history storage process,
The response history is stored in the storage means for each caller ID,
In the history search process,
Including a process of obtaining a caller ID of the caller when there is an incoming call from the caller, and searching a response history using the caller ID as a caller-specific key,
In the response process,
Including a process of determining a response format based on the response history retrieved by the history retrieval process.
[0016]
According to the present invention (claim 13), as a response format of the response history,
A combination of voice recognition and voice guidance, a combination of push button dial signal recognition and voice guidance, a combination of voice guidance and call recording, voice guidance, an incoming call, and a response at the transfer destination are used.
The present invention (Claim 14) further includes a response format designation instruction process for causing the caller to designate a response format when the calling ID is not registered as a response history in the storage means,
In the response process,
For an incoming call with a caller ID, a process of determining a response format based on a history of the number of responses retrieved by the history retrieval process is included.
[0017]
The present invention (Claim 15) further includes a silent time detection process for detecting silence for a predetermined time when responding to the caller,
The response process is
The process of changing the response format when there is silence for a predetermined time in the silent time detection process, the process of changing the response format due to the success / failure of voice recognition, or the process of changing the response format by the operation of the caller One or more processes are executed.
[0018]
As described above, according to the present invention, the success / failure history of the speaker's past speech recognition is determined. By activating the response by recognition, the recognition rate of voice recognition can be improved.
In addition, the response format and response method can be changed for each speaker.
[0019]
Further, when searching for the history of the speaker, the operability of the voice recognition user is obtained by using a search means using the calling ID (calling telephone number) transmitted via the telephone line as a key for specifying the speaker. Can be improved.
[0020]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 3 shows the configuration of the response device of the present invention.
The response device shown in FIG. 1 includes a line interface unit 1 connected to a communication network, an incoming call detection unit 2 that detects an incoming call from the communication network, an outgoing ID detection unit 3 that detects an outgoing ID sent from the communication network, A communication unit 4 for making a call with a person, a response unit 5 having a plurality of response formats capable of responding to incoming calls, a control unit 6 for controlling the device, and a response history in the response unit 5 are stored for each calling ID and response format It is comprised from the response format log | history storage part 11 by origination ID.
[0021]
The response format history storage unit 11 for each calling ID stores the calling ID, the result of past voice recognition success / failure for each caller, the number of responses, and the response format at the time of response as the response history.
The control unit 6 searches the response history from the response ID history storage unit 11 for each calling ID based on the calling ID detected by the calling ID detection unit 3 at the time of an incoming call, and responds in the response unit 5 from the obtained response history. Is specified.
[0022]
Next, the operation in the above configuration will be described.
FIG. 4 is a sequence chart of the operation of the response device according to the present invention.
Step 101) The incoming call detection unit 2 detects an incoming call from the communication network via the line interface unit 1, and notifies the control unit 6 of the incoming call.
Step 102) The calling ID detection unit 3 detects the calling ID sent from the communication network and notifies the control unit 6 of it.
[0023]
Step 103) The control unit 6 searches the response format history storage unit 11 for each outgoing ID based on the outgoing ID acquired from the outgoing ID detection unit 3, and acquires the response history corresponding to the outgoing ID.
Step 104) The control unit 6 passes the acquired response history to the response unit 5, and the response unit 5 determines a response format corresponding to the response history in the plurality of response formats 1 to n included in the response unit 5.
[0024]
Step 105) The response unit 5 transfers the determined response format to the call unit 4, and responds to the user from the call unit 4 via the communication network.
Accordingly, it is possible to provide a variable response format for each caller based on the response format for each caller ID stored in the response format history storage unit 11 for each caller ID.
[0025]
【Example】
Embodiments of the present invention will be described below with reference to the drawings.
[First embodiment]
FIG. 5 shows the configuration of the response device according to the first embodiment of the present invention. 3, the same parts as those in FIG. 3 are denoted by the same reference numerals, and the description thereof is omitted.
[0026]
The response device shown in FIG. 5 has a transfer destination telephone directory storage unit 15, a transmission unit 16, a speech path switch 17, and a response detection unit 18 added to the configuration of FIG. A pulse signal (PB / DP) recognition unit 8, a voice recognition unit 9, a call recording unit 10, and a voice guidance unit 19 are added.
In the control unit 6, based on the outgoing ID detected by the outgoing ID detection unit 3, the response format history storage unit 11 for each outgoing ID is referred to, the response format is acquired, and the response unit 5 Instruct various functions to output a response format to the communication unit 4.
[0027]
In the present embodiment, the ID ID response format history storage unit 11 includes a response format combining voice guidance and voice recognition, a response format combining voice guidance and PB / DP recognition, voice guidance and call recording as response formats. It is assumed that a combined response format, a voice guidance response format, or a response format in which an incoming call is transferred to a transfer destination and responded from the transfer destination is stored for each calling ID.
[0028]
Further, as a response format in the response unit 5, a response format combining voice guidance and voice recognition, a response format combining voice guidance and PB / DP recognition, a response format combining voice guidance and call recording, a response format of voice guidance Alternatively, the incoming call is transferred to the transfer destination, and a response format for responding from the transfer destination is designated.
[0029]
The response recognition unit 9 of the response unit 5 recognizes the voice data acquired from the call unit 4 based on the response format as disclosed in, for example, JP-A-10-190842 and JP-A-7-230295.
In transferring an incoming call, the control unit 6 reads the transfer destination from the transfer destination telephone directory storage unit 15 and transmits the transfer destination from the transmission unit 16 to the transfer destination. Furthermore, at the time of a transfer response, the response detection unit 18 detects a response from the user and notifies the control unit 6 of the response. The control unit 6 controls the speech path switch 17 to form a speech path between the incoming call and the transfer destination, and performs control to transfer the incoming call.
[0030]
FIG. 6 is a sequence chart showing the operation of the response device according to the first exemplary embodiment of the present invention.
Step 201) The incoming call detection unit 2 detects an incoming call from the communication network via the line interface unit 1, and notifies the control unit 6 of the incoming call.
Step 202) The calling ID detection unit 3 detects the calling ID sent from the communication network, and notifies the control unit 6 of it.
[0031]
Step 203) The control unit 6 searches the response ID history storage unit 11 for each outgoing ID based on the outgoing ID acquired from the outgoing ID detection unit 3, and acquires a response history corresponding to the outgoing ID.
Step 204) The control unit 6 refers to the response format corresponding to the caller ID in the response unit 5 from the response history, and according to the response format, each function of the response unit 5 (PB / DP recognition unit 8, voice recognition) From any one of the unit 9, the call recording unit 10, the voice guidance unit 19, or a combination of functions).
[0032]
Step 205) If the call is not an incoming call, the call unit 4 responds to the user via the communication network based on the response format selected in Step 204.
Step 206) In the case of an incoming call, the control unit 6 determines a response format to the transfer destination.
Step 207) Further, the control unit 6 extracts the transfer destination ID from the incoming call, and reads the transfer destination from the transfer destination telephone directory storage unit 15 based on the transfer destination ID.
[0033]
Step 208) The transmitting unit 16 makes a response corresponding to the response format determined in Step 206 to the read transfer destination.
Step 209) When the response detector 18 detects a response from the transfer destination, the controller 6 controls the speech path switch 17 so that the incoming call can be transferred.
Step 210) A communication path is formed between the incoming call and the transfer destination, and the incoming call is transferred from the transmission unit 16 to the transfer destination.
[0034]
[Second Embodiment]
In this embodiment, when the silent time continues for a certain time or more, and when the response history voice recognition has been successful, the number of responses is extracted, and when there is a predetermined number of consecutive IDs, Processing for performing a response by voice recognition will be described.
FIG. 7 shows the configuration of the response device according to the second embodiment of the present invention. The same components as those in FIG.
[0035]
The configuration shown in the figure is a configuration in which a silent time detection unit 14 is added to the configuration of FIG. 5. The silent time detecting unit 14 monitors the incoming call that has been answered, and determines that there is no sound if the silent time is longer than a certain time.
8 and 9 are flowcharts showing the operation of the response device according to the second embodiment of the present invention.
[0036]
First, the response device responds unattended (step 300). Here, when the calling ID detection unit 3 cannot detect the calling ID, the voice guidance is played from the response unit 5 through the calling unit 4 (step 301). When the originating ID can be detected, the control unit 6 reads the response history and response count for each originating ID from the response format history storage unit 11 for each originating ID, and when the response count is the first time, the response unit Then, the call unit 4 performs a manned response. At this time, the operator inputs the user data into the response format history storage unit 11 for each calling ID, and stores the contents of the response as a history (step 302). When the number of responses is two or more, the control unit 6 reads the response history for each outgoing ID from the response history storage for each outgoing ID 11 (step 303).
[0037]
Here, when the voice recognition has succeeded twice consecutively and the voice recognition is selected, the voice recognition unit 9 of the response unit 5 responds by voice recognition (step 305). If PB / DP recognition is selected in step 304, the PB / DP recognition unit 8 of the response unit 5 responds (step 306). If the speech recognition in step 305 is successful and not silent, the response history is updated (step 307). Thereafter, a service is provided (step 308). As a method for determining the success / failure of voice recognition, a method is used in which a user inputs a voice at the time of response, and after the voice is recognized, it is judged by a confirmation means using the user's PB or the like by voice guidance.
[0038]
If the response fails in voice recognition or if silence is determined, the PB / DP recognition unit 8 of the response unit 5 responds (step 306).
If the PB / DP recognition in step 306 is successful, the response history is updated (step 309). The response history is updated by writing the response format / order contents in the database. Thereafter, a service is provided (step 310).
[0039]
If the PB / DP recognition by the PB / DP recognition unit 8 fails, or if the silent time is longer than a certain time in the silence determination, the guidance asks whether to transfer to a manned person or record a call (step 311). If the user wishes to transfer, the user performs a transfer operation to update the response history in the response format history storage unit 11 for each calling ID (step 314), and perform transfer and manned response (step 315). If the user does not wish to transfer, the response history in the response format history storage unit 11 for each calling ID is updated (step 312) and the call is recorded (step 313).
[0040]
In addition, for users who have a success history of a certain number of consecutive responses in voice recognition, a voice recognition response is made from the beginning. Asks whether or not he / she wants a response by recognition, and the user who responded in the form of a combination of PB / DP recognition and voice guidance last time inquires whether or not he / she wants a response by voice recognition at the beginning of the response If the user desires, a response by voice recognition is performed.
[0041]
In the above operation, the determination of whether or not the response using the voice recognition has succeeded twice or more in succession is not limited to two times, and the voice recognition has been continuously performed a predetermined number of times or more in the past. The determination may be made based on whether or not there is a successful history.
Moreover, although the above embodiment has been described based on the configurations of FIGS. 3, 5, and 7, a disk device that is constructed as a program and that is connected to a computer that is used as the response device. It is possible to easily realize the present invention by storing it in a portable storage medium such as a floppy disk or a CD-ROM and installing it when implementing the present invention.
[0042]
The present invention is not limited to the above-described embodiments, and various modifications and applications can be made within the scope of the claims.
[0043]
【The invention's effect】
As described above, according to the present invention, the caller's past speech recognition success / failure history is determined. By starting the response according to, the recognition rate of voice recognition can be improved.
[0044]
Further, the present invention does not change the recognition dictionary as in the prior art, but by changing the response format and response method for each caller, when it is determined that the speaker is not suitable for speech recognition, In the subsequent processing, the operation time is shortened by not trying speech recognition.
Also, by detecting the caller identification information sent from the communication network, responding by selecting the response format based on the caller identification information, and storing the response history for each caller identification information, It is possible to provide a response format suitable for the person.
[0045]
In addition, when searching for a caller's response history, a caller's specific key for searching for a caller ID sent via a telephone line without causing the caller to input a caller ID as in the prior art. As a result, it is possible to improve the operability of the voice recognition user.
Also, when answering, the incoming call is forwarded using a response format such as voice recognition and voice guidance combination, push button dial signal recognition and voice guidance combination, voice guidance and call recording combination, voice guidance, etc. By responding at the forwarding destination, the burden on the caller can be reduced.
[0046]
In addition, a response format is specified for an incoming call without caller identification information, and an accumulated response history is obtained for an incoming call with caller identification information. It can be specified. In addition, it is possible to change the response format in the case of silence detection at the time of response or by the operation of the caller.
[Brief description of the drawings]
FIG. 1 is a diagram for explaining the principle of the present invention.
FIG. 2 is a principle configuration diagram of the present invention.
FIG. 3 is a configuration diagram of a response device according to the present invention.
FIG. 4 is a sequence chart of the operation of the response device according to the present invention.
FIG. 5 is a configuration diagram of a response device according to the first exemplary embodiment of the present invention.
FIG. 6 is a sequence chart showing the operation of the response device according to the first exemplary embodiment of the present invention.
FIG. 7 is a configuration diagram of a response device according to a second embodiment of the present invention.
FIG. 8 is a flowchart (No. 1) showing an operation of the response device according to the second exemplary embodiment of the present invention.
FIG. 9 is a flowchart (part 2) illustrating the operation of the response device according to the second exemplary embodiment of the present invention.
FIG. 10 is a configuration diagram of a conventional response device.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Line interface part 2 Incoming call detection means, incoming call detection part 3 Calling ID detection part 4 Calling part 5 Response means, response part 6 History search means, control part 8 PB / DP recognition part 9 Voice recognition part 10 Call recording part 11 Calling ID Separate response format history storage means, ID-specific response format history storage unit 15 Forwarding destination telephone directory storage unit 16 Transmission unit 17 Communication path switch 18 Response detection unit

Claims (15)

In a voice response method for connecting to a communication network and responding to a caller with voice,
For each caller, keep a response history including the results of past voice recognition success / failure, number of responses, response format at the time of response,
When there is an incoming call from the caller, search the response history corresponding to the caller,
A voice response method characterized by performing a response by voice recognition when there is a success history of voice recognition more than a predetermined number of times in the response history corresponding to the searched sender. The response history is kept for each caller ID,
When there is an incoming call from the caller, obtain the caller ID of the caller,
Search the response history using the caller ID as a caller specific key,
The voice response method according to claim 1, wherein a response format is determined based on the response history. As the response format,
3. The combination of voice recognition and voice guidance, a combination of push button dial signal recognition and voice guidance, a combination of voice guidance and call recording, voice guidance, an incoming call, and using one of the responses at the transfer destination. Voice response method. If the calling ID is not registered in the response history, let the caller specify the response format,
2. The voice response method according to claim 1, wherein for an incoming call having a caller ID, response history information corresponding to the caller ID is acquired from the response history, and a response format is determined based on a history of the number of responses. Any of the control methods of changing the response format when the response is silent for a predetermined time, changing the response format by success / failure of voice recognition, or changing the response format by the operation of the caller The voice response method according to claim 1, wherein a plurality of controls are performed. A voice response device connected to a communication network for responding to a caller with voice,
Response format history storage means for holding a response history including the result of success / failure of past voice recognition for each caller, the number of responses, and the response format at the time of response;
An incoming call detection means for detecting an incoming call from a caller;
In the incoming call detection means, when there is an incoming call, a history search means for searching a response history corresponding to the caller from the response format history storage means;
In the response history corresponding to the caller searched by the history search means, there is a response means for making a response by voice recognition when there is a continuous history of voice recognition more than a predetermined number of times. Voice response device. The response format history storage means includes:
The response history is kept for each caller ID,
The history search means includes:
When there is an incoming call from the caller, the caller ID of the caller is acquired, and the caller ID is used as a caller-specific key to search the response history,
The response means includes
7. The voice response device according to claim 6, further comprising means for selecting a response format based on the response history searched by the history search means. As the response format,
8. A combination of voice recognition and voice guidance, a combination of push button dial signal recognition and voice guidance, a combination of voice guidance and call recording, a voice guidance, and an incoming call are transferred, and any one of responses at the transfer destination is used. Voice response device. In the case where the calling ID is not registered in the response history, it further has a response format designation instruction means for causing the caller to specify a response format,
The response means includes
7. The voice response apparatus according to claim 6, further comprising means for determining a response format based on a history of the number of responses searched by the history search means based on the call ID for an incoming call having a call origination ID. Silence time detection means for detecting silence for a predetermined time when responding to the caller,
The response means includes
When the silent time detection means is silent for a predetermined time, the response format is changed, the response format is changed by the success / failure of voice recognition, or the response format is changed by the operation of the caller. The voice response device according to claim 6, wherein one or a plurality of means is executed. A storage medium connected to a communication network and storing a voice response program installed in a voice response device for responding to a caller with voice,
A response format history storage process for accumulating in the storage means a response history including the results of past voice recognition success / failure for each caller, the number of responses, and the response format at the time of response;
An incoming call detection process to detect incoming calls from callers;
In the incoming call detection process, when there is an incoming call, a history search process for searching a response history corresponding to the caller from the storage means;
The response history corresponding to the caller searched by the history search process has a response process of performing a response by voice recognition when there is a continuous history of successful voice recognition more than a certain number of times. A storage medium that stores a voice response program. The response format history storage process includes:
The response history is stored in the storage means for each caller ID,
The history search process includes:
When an incoming call is received from the caller, the caller ID of the caller is acquired, and the response history is searched using the caller ID as a caller-specific key,
The response process is:
12. A storage medium storing a voice response program according to claim 11, further comprising means for selecting a response format based on the response history searched by the history search process. As the response format,
The combination of voice recognition and voice guidance, a combination of push button dial signal recognition and voice guidance, a combination of voice guidance and call recording, voice guidance, an incoming call, and using one of the responses at the transfer destination. A storage medium storing a voice response program. If the calling ID is not registered as a response history in the storage means, it further includes a response format designation instruction process for allowing the caller to designate a response format;
The response process is:
12. A storage medium storing a voice response program according to claim 11, further comprising a process of determining a response format based on a history of the number of responses retrieved by the history retrieval process for an incoming call having a caller ID. A silence time detection process for detecting silence for a predetermined time when responding to the caller;
The response process is:
When there is no sound for a predetermined time in the silent time detection process, the response format is changed, the response format is changed by the success / failure of voice recognition, or the response format is changed by the operation of the caller. The storage medium storing the voice response program according to claim 11, wherein one or more of the processes are executed.

Images