JP2006202205A - Speech dialog system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a speech dialog system wherein a user is hardly bothered by a registeration operation. <P>SOLUTION: This speech dialog system 1 includes a speech dialog device 11 and a dialog controller 13 for controlling a dialog between an originating-side user and a terminating-side user. The originating-side user operates a dialog start part of the speech dialog device 11 in order to start a dialog. After the operation of the dialog start part, at least a speech representing search contents is inputted through a microphone by the originating-side user. Additional information required for searching the terminating-side user is acquired by an acquisition part, and then speech data and additional information are sent to a network 14 from a transmitting part. In the speech dialog controller 13, a retrieval part retrieves a database by using received speech data and additional information as a retrieval key to specify the terminating-side user after receiving the speech data and additional information through the network 14. A grouping part groups the terminating-side user specified by the retrieval part, to be held. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a voice dialogue system, and more specifically, a dialogue for controlling a dialogue between a voice dialogue device and a first user who uses the voice dialogue device and a second user who uses another device. The present invention relates to a voice interaction system including a control device.

In recent years, a push-to-talk (hereinafter referred to as “PTT (Push To Talk)”) function capable of communicating with one or more other parties by simply pressing a conversation button has begun to be installed in a voice interactive apparatus such as a mobile phone. With such a PTT function, a voice uttered from a certain user reaches a voice interactive device carried by all persons grouped in advance via a digital network. In PTT, the voice started as described above is exchanged by half-duplex communication.
JP 2003-271193 A

  However, since the conversation partners need to be grouped and registered in advance in the voice interaction apparatus as described above, the user of the voice interaction apparatus is troublesome even if it is a partner who only interacts once. There is a problem of having to operate.

  Furthermore, there is a problem that the capacity of the storage device included in the voice interactive apparatus may be wasted because it is necessary to register with the voice interactive apparatus even if the other party interacts only once.

  Therefore, an object of the present invention is to provide a voice dialogue system in which a user does not feel the troublesome registration operation.

  In order to achieve the above object, a first aspect of the present invention is directed to a voice interaction system. The voice dialogue system includes a voice dialogue device and a dialogue control device that controls a dialogue between a first user who uses the voice dialogue device and a second user who uses another device. In addition, the voice interaction device is configured such that a dialogue start unit operated by the first user to start the dialogue, and after the dialogue start unit is operated, at least a voice representing a search content is input by the first user, A microphone that outputs audio data representing the input search content, an acquisition unit that acquires additional information necessary for searching for the second user together with the audio data output from the microphone, and acquired by the acquisition unit A user-side transmitter that transmits the audio data and the additional information to the network. The voice conversation control device receives a voice data and additional information transmitted from the voice dialogue device via the network, and uses the voice data and additional information received by the receiver as search keys. The storage unit storing the database configured to be searchable by the two users, and the database stored in the storage unit using the audio data and the additional information received by the receiving unit, the first user searches At least until the dialogue between the first user and the second user is completed by grouping the second user specified by the search unit and the second user specified by the search unit into a group, A grouping unit for holding second users registered in the group.

  In addition, when the voice interactive apparatus has a push-to-talk function, the dialog start unit is typically a button operated by the first user.

  Further, when the voice interactive device can be mounted on the vehicle, the additional information is at least one selected from the group including the current position, traveling direction, moving speed, destination, and route to the destination of the voice interactive device. Including. The search unit refers to the additional information to search for a second user near the current position of the voice interactive apparatus, near the destination, and / or near the route to the destination.

  Also, the database preferably includes status information indicating the current status of other equipment. In this case, the search unit searches for the second user with reference to the situation information included in the database.

  The status information typically indicates whether another device is in use. In this case, the search unit searches for a second user of a device that is not currently in use with reference to the situation information included in the database.

  The voice interaction control device preferably includes an audio storage unit that buffers the audio data received by the receiving unit, a device used by the second user grouped by the grouping unit, and the voice interaction device. After the connection is established, the control device side transmission unit that transmits the audio data buffered in the audio storage unit to the network in order to transmit to the second user side device grouped by the grouping unit Is further provided.

  Further, the voice interaction device preferably has a release request signal for requesting release of the connection in response to the operation of the first user to any of the devices that are currently connected to the voice interaction device. Is further provided. Here, the user side transmission unit further sends a release request signal from the release request unit to the network, and the reception unit further receives a release request signal transmitted from the voice interactive apparatus via the network. The dialogue control unit further releases the connection with the device specified by the release request signal received by the receiving unit.

  Further, the voice interaction device preferably has a maintenance request for requesting only one of the devices that are currently connected to the voice interaction device to maintain the connection in response to the operation of the first user. A maintenance request unit for outputting a signal is further provided. Here, the user side transmission unit further sends a maintenance request signal from the maintenance request unit to the network, and the reception unit further receives a maintenance request signal transmitted from the voice interactive apparatus via the network. Further, the dialogue control unit further releases a connection with a device other than the device specified by the maintenance request signal received by the receiving unit.

  The second aspect of the present invention is directed to a dialog control device that controls a dialog between a first user who uses a voice dialog device and a second user who uses another device. The apparatus includes: a dialog start unit operated by a first user to start a dialog; and a search that is input with at least a voice representing a search content input by the first user after the dialog start unit is operated A microphone that outputs audio data representing the contents, an acquisition unit that acquires additional information necessary for searching for the second user together with the audio data output from the microphone, and the audio data acquired by the acquisition unit and the addition A transmission unit for transmitting information to the network. The voice conversation control device receives a voice data and additional information transmitted from the voice dialogue device via the network, and uses the voice data and additional information received by the receiver as search keys. The storage unit storing the database configured to be searchable by the two users, and the database stored in the storage unit using the audio data and the additional information received by the receiving unit, the first user searches At least until the dialogue between the first user and the second user is completed by grouping the second user specified by the search unit and the second user specified by the search unit into a group, A grouping unit for holding second users registered in the group.

  The third aspect of the present invention is directed to a dialog control method for controlling a dialog between a first user who uses a voice dialog device and a second user who uses another device. The method is responsive to an operation for initiating a dialogue by a first user, wherein a dialogue start step executed by the voice interaction device is executed on the side of the voice interaction device, and at least after the dialogue start step is operated A voice input step for outputting the voice data representing the search content inputted by the first user and a voice input step for outputting the voice data representing the inputted search content, together with the voice data outputted in the voice input step. The acquisition step for acquiring the additional information necessary for searching for the second user, and the voice data and the additional information acquired by the acquisition step executed on the voice interaction device side A transmission step of transmitting to the network,

  A reception step for receiving voice data and additional information transmitted from the voice interaction device via the network, and a database stored in the voice interaction control device are received by the reception step. A search step for searching the voice data and the additional information as a search key to identify a second user who can interact with the first user at this time; A grouping step for holding the second users registered in the group at least until the second user is grouped into one group and the dialogue between the first and second users ends.

  The fourth aspect of the present invention is directed to a computer program for controlling a dialogue between a first user who uses a voice interaction device and a second user who uses another device. The dialogue device includes a dialogue start unit that is operated by the first user to start the dialogue, and after the dialogue start unit is operated, at least a voice representing a search content is input and input by the first user. A microphone that outputs audio data representing the search contents, an acquisition unit that acquires additional information necessary for searching for the second user together with the audio data output from the microphone, and the audio data acquired by the acquisition unit A transmission unit that transmits the additional information to the network. Here, the computer program is executed on the side of the voice interaction control device, and receives the voice data and additional information transmitted from the voice interaction device via the network, and is stored in the voice interaction control device. The database is searched by using the voice data and the additional information received in the receiving step as search keys, and the search step for specifying the second user who can interact with the first user this time is executed on the side of the voice dialog control device. The group that holds the second users registered in the group at least until the second user specified by the search step is grouped and the dialogue between the first and second users ends. And a conversion step.

  As described above, according to each aspect of the present invention, the voice interaction control device on the network responds to the voice data from the voice interaction device, and acts on behalf of the conversation partner specified by the search on behalf of the voice interaction device. Keep information. Therefore, the first user does not need to group and register the second user as the conversation partner in advance in the voice interaction apparatus. Furthermore, since it is not necessary to register the second user in the voice interactive apparatus, it is not necessary to waste the capacity of the storage device provided in the voice interactive apparatus.

  The above and other objects, features, aspects and advantages of the present invention will become more apparent when the detailed description of the present invention described below is understood in conjunction with the accompanying drawings.

(Embodiment)
FIG. 1 is a schematic diagram showing an overall configuration of a voice interaction system 1 according to an embodiment of the present invention. In FIG. 1, the voice dialogue system 1 includes at least a voice dialogue device (transmitting side) 11, one or more voice dialogue devices (incoming side) 12, and a voice dialogue control device 13. The voice interaction devices 11 and 12 and the voice interaction control device 13 are connected to be able to communicate with each other via a network 14 such as the Internet and / or a cellular network. FIG. 1 shows two voice interaction devices 12 a and 12 b as an example of one or more voice interaction devices (incoming side) 12.

  The voice interaction device 11 is, for example, an in-vehicle terminal device connected to a mobile phone 15 (see FIG. 2) having a hands-free function, and has a PTT (Push-to-Talk) function. As an example of such an in-vehicle terminal device, there is a navigation device that can be installed in a vehicle.

  Each of the voice interactive devices 12a and 12b is illustratively a POS (Point Of Sale) terminal device installed in a facility such as a restaurant or a store. As a specific example, the voice interaction device 12a is installed in the store A, and the voice interaction device 12b is installed in another store B. Further, these voice interaction devices 12a and 12b accept a call transmitted from the voice interaction device 11 by the PTT function, and then perform voice communication with the voice interaction device 11.

  The voice interaction control device 13 is typically incorporated in a server for performing call control or a packet switch in a cellular network, and the user of the voice interaction device 11 (hereinafter referred to as the originating user) and each of the voice interaction devices 12a and 12b. It is a device that controls one-to-many voice conversation (PTT) with a user (hereinafter called a called-side user).

  Next, the voice interaction apparatus 11 will be described. FIG. 2 is a block diagram showing a detailed configuration of the voice interaction apparatus 11 shown in FIG. In FIG. 2, the voice interaction apparatus 11 includes a communication unit 111, a microphone 112, a speaker 113, an interaction start unit 114, a CODEC 115, an additional information acquisition unit 116, a control unit 117, and a message generation unit 118.

  The communication unit 111 is connected to the network 14, sends a caller-side message (details will be described later) generated by the voice interaction device 11 to the network 14, and a callee-side message (via the network 14) Details will be received later.

  The microphone 112 receives at least a voice that can be searched by the voice dialog control device 13 for a dialog partner that the user of the voice dialog device 11 wants to search (hereinafter referred to as search content). Subsequent user's spoken words are input to the microphone 112. The microphone 112 converts such input sound into an analog sound signal and outputs it to the CODEC 115.

  The speaker 113 outputs a spoken word from the conversation partner in accordance with an analog voice signal sent from the CODEC 115 described later.

  The dialogue start unit 114 is typically a PTT button. After operating the dialog start unit 114, the user speaks toward the microphone 112 while operating the dialog start unit 114 more specifically. As a result, the analog sound information representing the input sound is output from the microphone 112 to the CODEC 115.

  The CODEC 115 converts the analog audio signal from the microphone 112 into a digital audio signal and outputs the digital audio signal to the control unit 117. The CODEC 115 further acquires the digital audio signal received by the communication unit 111 via the control unit 117, converts the acquired digital audio signal into an analog audio signal, and outputs the analog audio signal to the speaker 113.

  The additional information acquisition unit 116 acquires position information that can specify the current position of the user and / or personal information that can specify the user as additional information.

  The position information is the current position itself, the name of the road that is currently traveling, the name of the intersection where the current position is located, the position of the destination, the user's moving speed, the user's traveling direction, or the route to the destination, or A combination of two or more selected from these groups. Note that the position information as described above can be typically acquired by the additional information acquisition unit 116 from a known navigation system.

  The personal information is the user's name, age, address, telephone number, e-mail address, nickname or preference information, or a combination of any two or more selected from these groups. Such personal information is built into the voice interaction device 11 itself or stored in a storage device (not shown) connected to the voice interaction device 11. It should be noted that part of such personal information is preferably controlled by the voice interaction device 11 so that it is not transmitted to the unspecified voice interaction device 12. Specifically, only items (for example, nicknames) permitted in advance by the user are controlled to be transmitted to the unspecified voice interaction device 12.

  The control unit 117 is composed of, for example, a CPU, a ROM, and a RAM, and controls each component of the voice interaction device 11.

  The message generation unit 118 first acquires the digital audio signal output from the CODEC 115 through the control unit 117. Further, the message generation unit 118 acquires the additional information output from the additional information acquisition unit 116 through the control unit 117. The message generation unit 118 basically creates a caller-side message including at least the acquired digital audio signal and / or additional information, and outputs it to the communication unit 111. Specifically, the originator message includes a digital audio signal representing the search content.

  Next, the voice interaction device 12 will be described. Each voice interaction device 12 is connected to the network 14 and is a device capable of performing a PTT-based interaction. In addition, the voice interaction device 12 sends the internally generated incoming message to the network 14 in the same manner as the voice interaction device 11. Here, the incoming message includes a digital voice signal representing the spoken language of the user of each voice interaction device 12. Further, the voice interaction device 12 receives a digital voice signal included in the outgoing message sent by the voice interaction device 11 via the network 14. The voice interaction device 12 processes the received digital voice signal in the same manner as the voice interaction device 11, and outputs the voice represented thereby.

  Next, the voice interaction control device 13 will be described. FIG. 3 is a block diagram showing a detailed configuration of the voice interaction control device 13 shown in FIG. In FIG. 3, the voice conversation control device 13 includes a communication unit 131, a message decoding unit 132, and a database storage unit for controlling voice communication performed between the voice dialogue device 11 and one or more voice dialogue devices 12. 133, a search unit 134, a grouping unit 135, a dialogue control unit 136, a message storage unit 137, and a control unit 138.

  The communication unit 131 is connected to the network 14 and receives the outgoing message sent from the voice interactive device 11 and the incoming message sent from the voice interactive device 12 via the network 14.

  Further, the communication unit 131 sends the digital voice signal sent from the voice dialogue apparatus 11 together with the connection establishment request to the network 14 in accordance with the processing of the dialogue control unit 136 described later.

  The message decryption unit 132 acquires the transmission side message received by the transmission / reception unit 131 and extracts the digital voice information and additional information included in the acquired message. The extracted digital audio information and additional information are sent to the control unit 138.

  The database storage unit 133 stores a database including various types of information for each voice interactive device 12 connected to the network 14. Here, FIG. 4 is a schematic diagram showing the structure of records constituting the database. In FIG. 4, a record is exemplarily assigned to one voice interactive device 12 and includes identification information, a store name, an address, location information, store information, and status information.

  The identification information is information that can uniquely identify the target voice interaction device 12.

  The store name is the name of the store when the target voice interaction device 12 is installed in the store.

  The address is information necessary for a grouping process (details will be described later) executed before the voice dialogue apparatus 11 performs a voice dialogue with the voice dialogue apparatus 12 using the PTT. Examples of such information include a group ID, an IP (Internet Protocol) address, a URI (Uniform Resource Identifier), or a telephone number assigned to the target voice interaction device 12.

  The position information is information indicating the position where the target voice interactive device 12 is installed.

  The store information is information representing a service that can be provided to the user of the voice interaction device 11 when the target voice interaction device 12 is installed in the store. Such store information preferably includes outline information that roughly represents a service that the store can provide and detailed information that more specifically represents the service. The summary information includes, for example, “restaurant”, and the detailed information includes, for example, “Chinese restaurant” or “French restaurant”. Note that the summary information and the detailed information are not particularly shown in FIG.

  The status information is information indicating whether or not the target voice dialogue apparatus 12 is currently carrying out a voice dialogue with PTT. Specifically, when a connection is established between a target voice interaction device 12 and a certain voice interaction device 11, that is, when a voice conversation starts, for example, “in use”, the usage status is a dialogue control unit. 136 is set. On the other hand, when the connection is released, the usage status is set by the dialog control unit 136 as “available”, for example. In addition, the situation information may indicate that the target voice interactive device 12 is not turned on.

  The database consists of the above records. In the present embodiment, for convenience, one record is assigned to one voice interactive device 12, but one record may be assigned to a plurality of voice interactive devices 12 in some cases.

  Here, FIG. 3 will be referred to again. The search unit 134 searches the database stored in the database storage unit 133 using the digital audio information extracted by the message decoding unit 132 as a keyword, that is, the first sent information, that is, the search content. Preferably, the search unit 134 further searches the database using additional information sent together with the search content. Based on the identification information obtained by such a search, the search unit 134 identifies a called user that the calling user can interact with at this time.

  The grouping unit 135 groups and holds the identification information searched by the search unit 134 (that is, the incoming side user with whom the outgoing side user can interact this time) in addition to the search content and additional information included in the outgoing side message. .

  The dialogue control unit 136 establishes a connection for performing a push-to-talk between the calling side user and the called side user grouped by the grouping unit 135. The dialogue control unit 136 uses a call control protocol such as SIP (Session Initiation Protocol), for example, to establish a connection between the voice dialogue device 11 that has sent the search content and the voice dialogue device 12 that has been retrieved from the database. After the connection is established, the voice interaction device 11 and the voice interaction device 12 can directly perform PTT-based voice communication without going through the voice dialog control device 13, but the dialog control unit 136 monitors this voice communication. You can also. When the connection is established, the dialog control unit 136 sets the status information of the record assigned to the target voice dialog device 12 to “in use”. When the connection is released, the dialog control unit 136 sets the target status information. Set to “Available”.

  The voice accumulating unit 137 accumulates the search content extracted from the caller-side message sent from the voice interaction apparatus 11.

  The control unit 138 includes, for example, a CPU, a ROM, and a RAM, and controls each component of the voice interaction control device 13 according to a computer program stored in advance in the ROM.

  Next, the operation of the voice interaction system according to the present embodiment will be described with reference to the drawings. FIG. 5 is a sequence diagram showing respective operations of the voice interaction device 11, the voice interaction control device 13, and the voice interaction device 12 until the voice conversation of the calling user and the called user starts in the voice interaction system according to the present embodiment. FIG.

  First, in the voice interaction device 11, the message generator 20 generates a caller message (sequence S11). As described above, the first caller message includes at least digital audio information representing the search content and additional information. The outgoing message generated in step S11 is transmitted to the network 14 via the communication unit 111 and the mobile phone 15, and is transmitted to the voice interaction control device 13 (sequence S12).

  In the spoken dialogue control apparatus 2, the message decryption unit 132 decrypts the caller message received this time, and obtains search contents and additional information included therein (sequence S13). The search content and additional information are passed to the control unit 138, and the control unit 138 stores the search content in the message storage unit 137 and passes the search content and additional information to the search unit 134.

  Next, the search unit 134 searches the database stored in the database storage unit 133 by using the received search content and additional information (sequences S14 and S15), and thereby, the candidate that the calling-side user performs a voice conversation this time , The callee user (that is, the voice interaction device 12). For example, if the search content is to eat a Chinese restaurant and the additional information is, for example, the current location of the calling user, the search unit 134 checks the store information and location information in the database, and At least one record assigned to the Chinese restaurant around the location is searched. Since broadcast communication is possible in a PTT-based voice conversation, preferably a plurality of records are searched. More preferably, in the search process, the search unit 134 checks the status information, searches for a record in which the status information is set to “available”, and searches for a record in which the status information is set to “unavailable”. do not do.

  Here, FIG. 6 is a flowchart showing a detailed processing procedure of search processing by the search unit 134. In the following, the search process will be described on the assumption that the search content is “I want to eat Chinese food” and the additional information is the current location of the user on the transmission side.

  First, the search unit 134 compares the “location information” field in the database with the additional information, and searches all stores near the current location of the sender user (step S40). When one or more records are found in this search (Yes in step S41), the search unit 134 moves the process to step S42. If no record is found (No in S41), the search unit 134 moves the process to step S47, completes the extraction (zero), and ends the process.

  In step S42, the contents of the search and the “store information” field of the record found in step S41 are compared, and a record of a store that satisfies the condition of “Chinese cuisine” as the search request is searched. If there is one or more records (Yes in step S43), the search unit 134 moves the process to step S44. Conversely, if no record is found (No in step S43), the process proceeds to step S47 described above.

  Next, the search unit 134 checks the status information of the found record and searches for a record for which “usable” is set (step S44). If there is one or more records (Yes in step S45), the search unit 134 moves the process to step S46, and completes extraction (one or more) of the incoming side users who will be the current conversation partner (step 1). S46), the process ends. On the other hand, if none is found (No in step S45), the search unit 134 moves the process to S47.

  In the above description, the search processing is performed in the order of steps S40 → S42 → S44. However, this order may be changed. Furthermore, the search process may be performed at a time.

  In the above description, the additional information is described as the current position. However, the present invention is not limited to this, and as described above, the additional information includes the current position, the name of the currently traveling road, and the intersection at the current position. The name, the position of the destination, the moving speed of the user, the traveling direction of the user, the route to reach the destination, or a combination of two or more selected from these groups may be used. In such a case, the search unit 134 refers to the additional information, and the second part that is near the current position, near the destination, and / or near the route to the destination of the voice interaction device 11 (the user on the transmission side). Search the database for users. In addition, when the voice interactive device 11 is mounted on the in-vehicle navigation device, it is determined whether the vehicle is traveling in the city or in the suburbs, and the search area is narrowed if the vehicle is urban and the search area is expanded if the vehicle is traveling in the suburbs. Can do. Based on the scale information of the map currently displayed on the vehicle-mounted navigation device, the search unit 134 can also search for a called user who exists in the area displayed on the screen. Further, when the number of the called users found in the current search is small, the search unit 134 searches for the called users from a wide area in the next search, and conversely, the number of called users obtained by the search. When there are many, the search part 134 can also reduce a search area gradually and can search.

  It is assumed that the number of records obtained this time by the search processing as described above is two, and that the voice interactive devices 12a and 12b are installed in the store to which the records are assigned.

  Next, the grouping unit 135 groups and holds the voice interaction device 11 that has sent the message on the outgoing side this time and the voice interaction device 12 that has been searched this time (sequence S16).

  Furthermore, the dialogue control unit 136 establishes a connection for performing PTT-based voice communication between the voice dialogue device 11 grouped in step S16 and the voice dialogue device 12 (sequence S17 to S20). Thereafter, the calling-side user starts a voice conversation using the voice interaction device 11 and the called-side user uses the voice interaction device 12 (sequence S21). Under the above assumption, a connection is established between the voice interaction device 11 and the voice interaction devices 12a and 12b, and voice communication starts between them.

  When the connection is established, the grouped voice interaction devices 12 are in voice communication. Therefore, the dialogue control unit 136 stores the record assigned to the voice interaction device 12 that has started voice communication in the database. Set status information to Unusable. Further, as described above, the dialogue control unit 136 can monitor the voice communication between the voice dialogue device 11 and the voice dialogue device 12, and when the voice communication is finished, the status information of the target record is “available”. To "".

  If no record can be searched in the search process of step S14, the voice interaction control device 13 may connect to a predetermined connection destination, for example, a concierge service center. At this time, the connection destination and the voice interaction device 11 are grouped.

  Whether or not the voice conversation control device 13 performs a voice conversation with the called-side user obtained by the search with respect to the calling-side user after the grouping by the grouping unit 135 and before the connection is established. You can contact us. Thereby, the transmission side user can refuse voice conversation. In this case, the voice conversation control device 13 does not perform connection establishment processing (sequences S17 to S20).

  The voice communication performed between the voice dialogue apparatus 11, the voice dialogue control apparatus 12, and the voice dialogue apparatus 13 after performing the connection establishment process as described above will be described. FIG. 7 is a sequence diagram showing operations of the voice interaction device 11, the voice interaction control device 13, and the voice interaction device 12 shown in FIG. 1 after the connection establishment process.

  First, as described above, the calling-side user inputs the voice representing the search content to the microphone 112 while operating the dialogue start unit 114 of the voice dialogue apparatus 11. The voice representing the search content is exemplified by “I want to eat Chinese food”. In response to this, the voice interaction device 11 creates the first caller message as described with reference to FIG. Further, the voice conversation control device 12 accumulates the search contents included in the caller message in the message accumulating unit 137, and further uses the caller message as described with reference to FIG. The called user who should interact with the user this time is searched, and a connection is established between the voice interaction apparatuses 11 and 12.

After that, as shown in FIG. 7, the voice interaction control device 13 transmits digital audio information representing the search contents stored in its message storage unit 137 to the target voice interaction devices 12a and 12b (sequence S22). )
Upon receiving such digital voice information, the receiving user of the voice interaction device 12a and the receiving user of the voice interaction device 12b each respond independently. As for the contents of each response, the incoming message is transmitted from each of the voice interactive devices 12a and 12b, and the incoming message sent from the voice interactive device 12a is transmitted to the voice interactive devices 11 and 12b (sequence S23). . Also, the incoming message from the voice interaction device 12b is transmitted to the voice interaction devices 11 and 12a (sequence S24).

  In addition, the receiving side user inputs a response to the voice interactive apparatus 11 in response to each incoming side message. In response, voice interaction device 11 generates a further caller message based on the input response and transmits it to voice interaction devices 12a and 12b (sequence S25). Thereafter, the same sequence is repeated until the voice conversation is finished (sequences S26 to S29).

  As described above, according to the voice dialogue system according to the present embodiment, the voice dialogue control device 13 on the network 14 is included in response to the first caller side message from the voice dialogue device 11. The called user is searched using the search contents and the additional information, and the called user and the calling user obtained by the search are grouped on behalf of the voice interactive device 11 and held. Therefore, it is not necessary for the calling-side user to group and register the called-side users who are conversation partners in the voice dialogue apparatus 11 in advance. Furthermore, since it is not necessary to register the called user in the voice interaction apparatus 11, it is not necessary to waste the storage capacity of the voice interaction apparatus 11.

  Further, by realizing the voice interaction device 11 as an in-vehicle terminal device, it is possible to reduce the operation load of user information search or to search without knowing store address information.

  In addition, since a receiving side message from a certain voice interactive device 12 reaches another voice interactive device 12, the receiving side user does not always interact with the calling side user and may be confused with each other. Conceivable. For this reason, control may be performed so as to prohibit the exchange of messages on the called side between the voice interaction devices 12.

  In addition, when the voice interaction device 11 is mounted on the in-vehicle navigation device, the in-vehicle navigation device searches for a route from the current position to the store of the incoming user by acquiring the location information of the incoming side message, The calling user can be guided to the receiving user's store.

  In the above embodiment, the voice interaction device 12 has the same configuration as the voice interaction device 11 and performs the same processing. However, the present invention is not limited to this, and the voice interaction device 12 is represented by Bluetooth. When the short-range wireless communication protocol is implemented, a plurality of headsets can respond to voice conversation via the voice dialogue apparatus 12. In this case, status information for each headset that can access the voice interactive device 12 is added to the target record in the database.

  In addition, one of the input devices (not shown) provided in the voice interaction device 11 is connected to any one or more of the voice interaction devices 12 to which a connection is currently established in response to a user operation on the caller side. On the other hand, a release request function that outputs a release request signal for requesting release of the connection is preferably assigned. In this case, when the calling user operates the target input device, as shown in FIG. 8, the communication unit 111 of the voice interaction device 11 sends a release request signal from the input device to the network 14, and the voice It is transmitted to the dialogue control device 13 (sequence S31). In the voice dialogue control device 13, the communication unit 131 further receives a release request signal transmitted via the network 14, and the dialogue control unit 136 communicates with the voice dialogue device 12 designated by the received release request signal. Are released (sequences S31 and S32). As a result, the voice interaction device 11 continues the voice communication only with the remaining voice interaction devices 12 that were not targeted for the current connection release (sequence S33).

  In addition, one of the input devices (not shown) provided in the voice interaction device 11 can be applied only to one of the voice interaction devices 12 to which a connection is currently established in response to the operation of the originating user. It is preferable that a maintenance request function for outputting a maintenance request signal for requesting connection maintenance is assigned. In this case, when the originating user operates the target input device, the communication unit 111 of the voice interaction device 11 sends a maintenance request signal from the input device to the network 14 and is transmitted to the voice interaction control device 13. The In the voice interaction control device 13, the interaction control unit 136 releases a connection with a device other than the voice interaction device 12 specified by the maintenance request signal received this time. As a result, the voice interaction device 11 continues voice communication only with the remaining voice interaction devices 12 that were not targeted for connection release this time.

  With the above two methods, the calling user can terminate the voice conversation with the called user who no longer needs to talk.

  Furthermore, since the voice interaction control device 13 monitors the voice communication between the voice interaction devices 11 and 12, when there is no digital voice data flowing between the two, that is, when a silent section of a certain time occurs. In this case, it may be determined that the conversation has ended and the connection is automatically released.

  Further, when the voice interactive device 11 is mounted on the in-vehicle navigation device, the connection with the target voice interactive device 12 is maintained until the destination set from the store information of the voice interactive device 12 is reached. The connection may be automatically released upon arrival at the destination.

  On the other hand, since the voice interaction device 11 may be a mobile object, a situation occurs in which the voice interaction device 12 belonging to the group becomes unsuitable as a conversation partner of the calling user as time elapses, that is, with movement. For example, if the search content is “I want to eat Chinese food” and the additional information is the current location, the calling user has moved far away from the current location at the time of the search as a result of the movement. The situation is such that the voice conversation is continued with the called users grouped around the current location at the time of the search. In order to avoid such a situation, the voice dialogue control device 2 periodically sends additional information from the voice dialogue device 11 while the voice dialogue is being performed. Connections established between the devices 11 and 12 can be released and regrouped. As a result, it is possible to maintain a group that can satisfy the search request of the caller user according to the position information of the voice interaction device 11 that changes from moment to moment.

  Moreover, the control part 138 performed according to the computer program previously stored in ROM as mentioned above. However, the present invention is not limited to this, and the above processing may be realized by hardware. The computer program may be distributed in a state where it is recorded on a storage medium such as a CD-ROM. In addition, the computer program may be stored in a server device connected to the network so that the terminal device can be downloaded or stored.

  In the above embodiment, since the PTT-based voice interaction device 11 has been described, the interaction start unit 114 has been described as a button. However, the present invention is not limited to this, and the dialogue start unit 114 is a voice input device if the dialogue is started by voice input.

  As mentioned above, although this invention was demonstrated in detail, the said description is an illustration in all the meanings, and is not restrictive. It will be appreciated that many other modifications and variations are possible without departing from the scope of the invention.

  The voice interaction system according to the present invention is particularly useful for a PTT-based voice interaction device or the like that requires a technical effect that the user is less likely to feel the troublesome registration operation.

The schematic diagram which shows the whole structure of the voice dialogue system 1 which concerns on embodiment of this invention. FIG. 1 is a block diagram showing a detailed configuration of the voice interaction apparatus 11 shown in FIG. The block diagram which shows the detailed structure of the voice dialogue control apparatus 13 shown in FIG. The schematic diagram which shows the structure of the record which comprises the database stored in the database storage part 133 shown in FIG. The sequence diagram which shows each operation | movement of the voice dialogue apparatus 11, the voice dialogue control apparatus 13, and the voice dialogue apparatus 12 until the voice dialogue in the calling-side user and the called-side user starts in the voice dialogue system according to the present embodiment. The flowchart which shows the detailed process sequence of the search process by the search part 134 shown in FIG. The sequence diagram which shows each operation | movement of the voice dialogue apparatus 11, the voice dialogue control apparatus 13, and the voice dialogue apparatus 12 which are shown in FIG. FIG. 1 is a sequence diagram relating to the connection release of the voice interactive apparatuses 11 and 12 shown in FIG. 1 after the connection establishment process.

Explanation of symbols

1 Spoken Dialogue System 11 Spoken Dialogue Device (Sender)
12 Spoken dialogue device (incoming side)
13 Spoken Dialogue Control Device 14 Network

Claims (11)

A spoken dialogue system,
The voice dialogue system includes a voice dialogue device, and a dialogue control device that controls a dialogue between a first user who uses the voice dialogue device and a second user who uses another device,
The voice interaction device
A dialogue start unit operated by the first user to start a dialogue;
After the dialogue start unit is operated, at least a voice representing search content is input by the first user, and a microphone that outputs voice data representing the input search content;
An acquisition unit that acquires additional information necessary for searching for the second user together with audio data output from the microphone;
A user-side transmission unit that transmits the audio data and additional information acquired by the acquisition unit to a network;
The voice interaction control device
A receiving unit for receiving voice data and additional information transmitted from the voice interaction device via the network;
A storage unit for storing a database configured to be able to search for the second user using the audio data and additional information received by the receiving unit as search keys;
A search unit that searches the database stored in the storage unit using the audio data and additional information received by the reception unit, and identifies a second user that the first user can interact with at this time;
The second user specified by the search unit is grouped into one group, and at least the second user registered in the group is held until the dialogue between the first and second users is ended. A voice dialogue system comprising a grouping unit.   The voice interaction system according to claim 1, wherein when the voice interaction device implements a push-to-talk function, the interaction start unit is a button operated by the first user. When the voice interactive device can be mounted on a vehicle, the additional information is at least one selected from the group including the current position, traveling direction, moving speed, destination, and route to the destination of the voice interactive device. Including
2. The search unit according to claim 1, wherein the search unit searches for the second user near the current position of the voice interactive apparatus, near the destination and / or near the route to the destination with reference to the additional information. Voice dialogue system. The database includes status information indicating the current status of the other device,
The spoken dialogue system according to claim 1, wherein the search unit searches for the second user with reference to status information included in the database. The status information indicates whether the other device is in use,
The spoken dialogue system according to claim 4, wherein the search unit searches for a second user of a device that is not currently in use by referring to status information included in the database. The voice interaction control device
An audio storage unit for buffering audio data received by the receiving unit;
A dialogue control unit that establishes a connection between the device used by the second user grouped by the grouping unit and the voice dialogue device;
When a connection is established by the dialog control unit, the audio data buffered in the audio storage unit is sent to the network for transmission to a second user device grouped by the grouping unit. The spoken dialogue system according to claim 1, further comprising a control device-side transmission unit. The voice interaction device sends a release request signal for requesting release of a connection to one of the devices currently connected to the voice interaction device in response to the operation of the first user. It further includes a release request part for outputting,
The user side transmission unit further sends a release request signal from the release request unit to the network,
The receiving unit further receives a release request signal transmitted from the voice interaction device via the network,
The voice dialogue system according to claim 6, wherein the dialogue control unit further releases the connection with the device specified by the release request signal received by the receiving unit. The voice interaction apparatus is a maintenance request signal for requesting only one of the devices that are currently connected to the voice interaction apparatus to maintain the connection in response to the operation of the first user. Is further provided with a maintenance request unit that outputs
The user side transmission unit further sends a maintenance request signal from the maintenance request unit to the network,
The receiving unit further receives a maintenance request signal transmitted from the voice interaction device via the network,
The voice dialogue system according to claim 6, wherein the dialogue control unit further releases a connection with a device other than the device designated by the maintenance request signal received by the receiving unit. A dialogue control device for controlling a dialogue between a first user who uses a voice dialogue device and a second user who uses another device,
The voice interaction device
A dialogue start unit operated by the first user to start a dialogue;
After the dialogue start unit is operated, at least a voice representing search content is input by the first user, and a microphone that outputs voice data representing the input search content;
An acquisition unit that acquires additional information necessary for searching for the second user together with audio data output from the microphone;
A transmission unit that sends the audio data and additional information acquired by the acquisition unit to a network;
The voice interaction control device
A receiving unit for receiving voice data and additional information transmitted from the voice interaction device via the network;
A storage unit for storing a database configured to be able to search for the second user using the audio data and additional information received by the receiving unit as search keys;
A search unit that searches the database stored in the storage unit using the audio data and additional information received by the reception unit, and identifies a second user that the first user can interact with at this time;
The second user specified by the search unit is grouped into one group, and at least the second user registered in the group is held until the dialogue between the first and second users is ended. A dialog control device comprising a grouping unit. A dialog control method for controlling a dialog between a first user using a voice dialog device and a second user using another device,
An interaction start step executed by the voice interaction device in response to an operation to start an interaction by the first user;
Voice input that is executed on the side of the voice interaction device and after the dialogue start step is operated, at least a voice representing the search content is inputted by the first user, and voice data representing the inputted search content is output. Steps,
An acquisition step for acquiring additional information necessary for searching for the second user together with the voice data output in the voice input step and executed on the voice interaction device side;
A transmission step executed on the side of the voice interaction device and transmitting the voice data and additional information acquired in the acquisition step to a network;
A reception step executed on the side of the voice interaction control device and receiving voice data and additional information transmitted from the voice interaction device via the network;
The database stored in the voice interaction control device is searched using the voice data received in the receiving step and the additional information as a search key, and the second user who can interact with the first user is specified. A search step;
The second user, which is executed on the side of the voice interaction control device and specified by the search step, is grouped into one group, and at least until the dialogue between the first and second users is completed, And a grouping step for holding a registered second user. A computer program for controlling a dialogue between a first user using a voice interaction device and a second user using another device,
The voice interaction device
A dialogue start unit operated by the first user to start a dialogue;
After the dialogue start unit is operated, at least a voice representing search content is input by the first user, and a microphone that outputs voice data representing the input search content;
An acquisition unit that acquires additional information necessary for searching for the second user together with audio data output from the microphone;
A transmission unit that sends the audio data and additional information acquired by the acquisition unit to a network;
The computer program is
A reception step executed on the side of the voice interaction control device and receiving voice data and additional information transmitted from the voice interaction device via the network;
The database stored in the voice interaction control device is searched using the voice data received in the receiving step and the additional information as a search key, and the second user who can interact with the first user is specified. A search step;
The second user, which is executed on the side of the voice interaction control device and specified by the search step, is grouped into one group, and at least until the dialogue between the first and second users is completed, A computer program comprising: a grouping step for holding a registered second user.

Images