JP6992957B2 - Agent dialogue system - Google Patents

Description

The present invention relates to an agent dialogue system, particularly, for example, a novel agent dialogue system in which a plurality of agents and at least one person participate.

With the expansion of the robot field, robot research is shifting its focus to research on robots that work in everyday situations, and the development of robots that can interact with humans in the environment in which humans live is drawing attention. ..

With the development of voice recognition technology in recent years, robots that have the function of interacting with humans in voice language have been developed, but no matter how good the voice recognition technology is, robots and people In the dialogue, it is not easy to continue to give the person who interacts with the robot the same feeling as the "sense of dialogue (feeling of participating in the dialogue)" that a person has when participating in the dialogue with a person. I didn't.

Patent Document 1, which is an example of the background technique, discloses that in a dialogue between a robot and a human, the robot infers the emotion of the human and determines the utterance of a response sentence and the action accompanying the utterance.

Patent Document 2 attempts to realize a sustainable and natural interaction in a dialogue system between a human and a robot by synchronizing the two.

Japanese Unexamined Patent Publication No. 2004-90109 [B25J 13/00…] Japanese Unexamined Patent Publication No. 2012-181697 [G06F 3/16 ...]

In both the technique of Patent Document 1 and the technique of Patent Document 2, there is a limit to the processing based on voice recognition, and it is not easy for a person to maintain the above-mentioned "dialogue feeling".

Therefore, the main object of the present invention is to provide a novel, agent dialogue system.

Another object of the present invention is to provide an agent dialogue system capable of sustaining a person's sense of participation in dialogue.

The present invention has adopted the following configuration in order to solve the above problems. The reference numerals and supplementary explanations in parentheses indicate the correspondence with the embodiments described in order to help the understanding of the present invention, and do not limit the present invention in any way.

The first invention is an agent dialogue system in which two agents, each having a speech function, interact with each other according to a script at the dialogue place, and a person sits at the dialogue place.
Whether there is a sensor manager who certifies sensor signals such as human actions and state changes at the dialogue place as the intention expressed by the person, regardless of the actual intention of the person, and whether there is an intention to be approved by the sensor manager. A non-language that approves a person's intention to two agents according to the first approval dialog loaded from the script into the dialog queue when the first judgment means and the first judgment means decide that they have the intention to approve. When it is determined that the first approval dialog execution means for executing the operation, the second judgment means for determining whether or not the phase is for confirming the human intention according to the script, and the second judgment means for confirming the human intention, the person's intention is confirmed. When it is determined that the third judgment means for determining whether or not the intention can be confirmed and the third judgment means can confirm the intention of the person, the two agents are assigned to the two agents according to the second approval dialog loaded from the script into the dialog queue. The second approval dialog execution means is provided with a second approval dialog execution means for executing a dialogue for approving the intention, and the second approval dialog execution means causes one of the two agents to make utterances that are in sync with or out of sync with the intention of a person. It is an agent dialogue system that causes the other one of the two agents to make an utterance that is out of sync with the intention of the person .

In the first invention, the agent dialogue system (10: a reference reference numeral exemplifying the corresponding portion in the embodiment; the same applies hereinafter) has two bodies , each having a speech function (46) at the dialogue place (12). Agents (R1 and R2) are placed, and a person (H) sits at the dialogue place (12). Agents interact with each other according to a script. The sensor manager (18) has, for example, an audio signal acquired by the microphone (14), an image signal captured by the camera (16), an input signal (affirmative, negative) detected by the intention display device (17), and a biological signal. The biological sensor that detects the above, and the sensor signal such as the action or state change of the person detected by the motion capture system are recognized as the intention expressed by the person regardless of the actual intention of the person. The first determination means (20, S75) determines whether or not the sensor manager (18) has certified and has the intention to approve. When it is determined that the first approval means has an intention to approve, the first approval dialog execution means (20, S77) attaches a person to the two agents according to the first approval dialog loaded from the script into the dialog queue. Perform a non-verbal action that approves the intention. After the non-verbal operation by the first approval dialog execution means, it is determined whether the second determination means (20, S79) is in the phase of confirming the intention of the person according to the script. When it is determined that the second determination means is in the phase of confirming the intention of the person, the third determination means (20, S81) determines whether or not the intention of the person can be confirmed. Then, when the second approval dialog execution means (20, S83) determines that the third judgment means has confirmed the intention of the person, the second approval dialog loaded from the script into the dialog queue is followed by the person to the two agents. Have a dialogue to approve the intention of . Specifically, the second approval dialog execution means causes one of the two agents to make an utterance that is in sync with or out of sync with the person's intention, and the other one of the two agents is made to speak the person's intention . Perform an approval dialogue by having an untuned or tuned utterance. For example, by the second approval dialog execution means, for example, two agents are made to make their respective approval utterances. As an example, as shown in FIGS. 16 and 18-20, one agent (one of R1 and R2) has an utterance that synchronizes with the intention of the person (H) (for example, "That's right", " Have the other agent (the other of R1 and R2) make utterances that do not match the person's intention (for example, "Yes,""Eh,different," etc.). Thereby, it is possible to determine whether the intention expressed by the person is "affirmation or consent or denial or disagreement".

According to the first invention, since the first approval dialog execution means executes the non-verbal operation for approval and the second approval dialog execution means executes the approval dialogue, the person participates in the dialogue between the agents. Not only can it create a feeling of being (“dialogue”), but it can also give a person the feeling of speaking, which is the premise of the approval dialogue.

The second invention is subordinate to the first or second invention, and the second approval dialog execution means is an agent dialogue system in which two agents are made to execute an approval utterance after expanding and interpreting a person's intention. be.

According to the second invention, the second approval dialog execution means expands and interprets the intention of the person, and then causes the two agents to execute the approval dialogue for the intention of the person. It is possible to strongly give a person the feeling of speaking with an interpretation.

The third invention is a program executed by a computer of an agent dialogue system in which two agents, each having a speaking function, interact according to a script at the dialogue place, and a person sits at the dialogue place. , A sensor manager who certifies sensor signals such as a person's actions and state changes in a dialogue place as an intention expressed by a person regardless of the actual intention of the person , and a sensor manager has certified and has an intention to approve. When it is determined that the first judgment means and the first judgment means have the intention to approve, the two agents approve the intention of the person according to the first approval dialog loaded from the script into the dialog queue. When it is determined that the first approval dialog execution means for executing the language action, the second judgment means for judging whether or not the phase is for confirming the person's intention according to the program, and the second judgment means for confirming the person's intention, the person When it is determined that the third judgment means for determining whether or not the intention of the person can be confirmed and the third judgment means can confirm the intention of the person, the person is assigned to the two agents according to the second approval dialog loaded from the program into the dialog queue. It is a program of the agent dialogue system that functions as a second approval dialog execution means for executing a dialogue that approves the intention of .

According to the present invention, even in a dialogue between agents, the person participates in the dialogue by approving the action of the person during the dialogue as a manifestation of the intention of the person, regardless of the actual intention of the person. It is possible to give a feeling (“dialogue feeling”) to a person and to give a person a feeling as if he / she spoke with the approval.

The above-mentioned object, other object, feature and advantage of the present invention will be further clarified from the detailed description of the following examples made with reference to the drawings.

FIG. 1 is a schematic diagram showing an outline of a robot dialogue system according to an embodiment of the present invention. FIG. 2 is a block diagram showing an example of the configuration of the sensor manager in the embodiment of FIG. FIG. 3 is a block diagram showing an example of the configuration of the group manager in the embodiment of FIG. FIG. 4 is a schematic view showing an example of the robot in the embodiment of FIG. FIG. 5 is a block diagram showing an example of a configuration of a robot controller that controls the robot of FIG. FIG. 6 is a schematic diagram showing an example of a script read from the dialog database shown in FIG. 1 to the group manager. FIG. 7 is a flow chart showing an example of the operation of the group manager shown in FIG. FIG. 8 is a flow chart showing an example of the operation of the robot controller shown in FIG. FIG. 9 is a schematic diagram showing an example of a dialogue sequence in the embodiment of FIG. FIG. 10 is a schematic diagram showing an example of a specific dialogue scene in the embodiment of FIG. FIG. 11 is a schematic diagram showing another example of the specific dialogue scene in the embodiment of FIG. FIG. 12 is a schematic diagram showing another example of the specific dialogue scene in the embodiment of FIG. FIG. 13 is a schematic diagram showing another example of the specific dialogue scene in the embodiment of FIG. FIG. 14 is a schematic diagram showing another example of the specific dialogue scene in the embodiment of FIG. FIG. 15 is a flow chart showing an example of the operation of the human intention processing module shown in FIG. FIG. 16 is a schematic diagram showing an example of a specific dialogue scene for processing a person's intention display in the embodiment of FIG. FIG. 17 is a schematic diagram showing another example of a specific dialogue scene for processing a person's intention expression in the embodiment of FIG. FIG. 18 is a schematic diagram showing another example of a specific dialogue scene for processing a person's intention expression in the embodiment of FIG. FIG. 19 is a schematic diagram showing another example of a specific dialogue scene for processing a person's intention expression in the embodiment of FIG. FIG. 20 is a schematic diagram showing another example of a specific dialogue scene for processing a person's intention expression in the embodiment of FIG. FIG. 21 is a schematic diagram showing another example of a specific dialogue scene for processing a person's intention expression in the embodiment of FIG. FIG. 22 is a schematic diagram showing another example of a specific dialogue scene for processing a person's intention expression in the embodiment of FIG. FIG. 23 is a schematic diagram showing another example of a specific dialogue scene for processing a person's intention expression in the embodiment of FIG. FIG. 24 is a schematic diagram showing another example of a specific dialogue scene for processing a person's intention expression in the embodiment of FIG. FIG. 25 is a schematic diagram showing another example of a specific dialogue scene for processing a person's intention expression in the embodiment of FIG.

With reference to FIG. 1, the first robot R1 and the second robot R2 and one person H are present at the dialogue place 12 of the robot dialogue system 10 of this embodiment. However, the number of robots may be 3 or more, and the number of people may be 2 or more. In the following, when it is not necessary to distinguish between the first robot R1 and the second robot R2, it may be simply referred to as a robot R.

When a person speaks, if the other person's utterance is not generated in response to the utterance, it tends to be difficult to continue the dialogue itself, so robots with imperfect voice recognition and natural language processing capabilities are humans. It is not easy to continue to speak appropriately for the utterance of, and the dialogue is liable to break down.

On the other hand, when observing a dialogue between others by the side, human beings can continue to listen to the dialogue while partially understanding or agreeing, even if the content is not completely understood or agreed. In addition, by doing so, it may be possible to later understand or agree with the part that was not understood or agreed. However, it is not easy to continue to concentrate on the dialogue without the feeling of participating in the dialogue (sense of dialogue).

Therefore, in this embodiment, when the robot speaks, it is not always made to respond to a human, but another robot is made to sit in the place of the dialogue, and an appropriate response to the robot's utterance is given. Avoid collapse by letting the robots speak, that is, letting the robots interact with each other, while letting people hear the dialogue between the robots and sometimes expressing the intention of human actions, such as "agreement" and "disagreement". ("Affirmation", "Negation"), etc. are regarded as received, and the intention of the person is approved and confirmed by continuing dialogue between robots or displaying non-verbal cautions by multiple robots. Makes humans feel a sense of dialogue.

The robot dialogue system as in this embodiment can be used, for example, as a communication service robot, an information providing service robot, an advertising robot, or the like for a single living person.

In this embodiment, a microphone 14 as an auditory sensor and a camera 16 as a visual sensor are provided in order to sense the situation of the dialogue place 12 of the robot dialogue system 10. The microphone 14 is for acquiring environmental sounds including voices spoken by the robot R and the human H, and two or more microphones 14 may be installed if necessary. The camera 16 is also a camera that captures the situation of the dialogue place 12, particularly the facial expression and movement of the person H, and captures a moving image or a still image. Two or more cameras 16 may be installed if necessary.

Further, in addition to the microphone 14 and the camera 16 described above, as sensors, although not shown, a living body such as a wearable posture sensor, an acceleration sensor, a heartbeat state, a breathing state, and a body movement (body movement) state. A biosensor for detecting a signal, a motion capture system, or the like may be provided.

Further, the dialogue system 10 of this embodiment is provided with an intention display device 17 as a means for expressing the intention of the person H. As an example, the intention display device 17 has an affirmative button 17a for inputting an affirmation or an agreement, and a negative button 17n for inputting a negative or disagree. However, the affirmative button 17a and the negative button 17n may be configured by a hardware switch such as a push button switch, or may be configured by a software switch (virtual button) on the touch screen. Further, the affirmative button 17a and the negative button 17n may be formed as individual switches or buttons.

A person detected by an audio signal acquired by the microphone 14, an image signal captured by the camera 16, an input signal (affirmative, negative) detected by the intention display device 17, a biosensor that detects the biometric signal described above, and a motion capture system. Sensor signals such as actions and state changes are input to the sensor manager 18. The sensor manager 18 acquires these sensor signals, determines the status of the dialogue location 12, and outputs the determination result as sensing data to the group manager 20 and the human intention processing module 23, which will be described later.

With reference to FIG. 2, the sensor manager 18 includes a CPU (central processing unit) 18a, and a communication device 18c is connected to the CPU 18a via an internal bus 18b. The communication device 18c includes, for example, a network interface (NIC), and the CPU 18a can communicate with the group manager 20 and the like via the communication device 18c, and data can be exchanged between them.

Further, the memory 18d is connected to the CPU 18a via the internal bus 18b. The memory 18d includes a ROM and a RAM. For example, a sensor signal including an audio signal from the microphone 14 (FIG. 1) and an image signal from the camera 16 (FIG. 1) is input through a sensor I / F (interface) 18e composed of a DSP (Digital Signal Processor). .. Then, the memory 18d temporarily stores the sensor signal.

The sensor manager 18 is a kind of determination device, and the CPU 18a determines the state of the dialogue place 12 and the intention of the person H based on the sensor data stored in the memory 18d. For example, it is detected by a voice signal acquired by the microphone 14, an image signal taken by the camera 16, an input signal (affirmative, negative) detected by the intention display device 17, a biosensor that detects the biometric signal described above, and a motion capture system. Sensor signals such as a person's actions and state changes can be recognized and processed as input of the intention expressed by the person H, regardless of the actual intention of the person H. Then, the sensor manager 18 sends data indicating the determined state to the group manager 20 and the human intention processing module 23.

Further, a program (OS, sensor signal acquisition program, etc.) required for the sensor manager 18 is stored in the memory 18d. The sensor manager 18 operates according to the program stored in the memory 18d.

A keyboard and a display (not shown) may be further attached to the CPU 18a.

The group manager 20 shown in FIG. 3 controls the speech operation (verbal operation) and behavior (nonverbal operation) of each of the two robots R, for example, according to the flow diagram of FIG. 7 described later.

The group manager 20 includes the CPU 20a, and the communication device 20c is connected to the CPU 20a via the internal bus 20b. The communication device 20c includes, for example, a network interface (NIC), and the CPU 20a can communicate with the sensor manager 18, the robot R, and the like via the communication device 20c, and data can be exchanged between them.

Further, the memory 20d and the input device 20e are connected to the CPU 20a via the internal bus 20b. The memory 20d includes a ROM and a RAM. A script (dialog) is read from the dialog database 22 through the memory I / F 20f, and the script (dialog) is temporarily stored in the memory 20d. However, "dialog" means a series of utterances and non-verbal movement commands that should be made during the dialogue, and the dialog database 22 is a set of dialogs (for example, what is important in dialogue about what you like about bananas and what is important in robot care). Includes a series of instructions for dialogue on each topic, such as dialogue. The "script" is a character string representing the sequence of the command, and the script data is a character string representing the command. Therefore, a series of script data becomes a script. Such a script is transmitted from the dialog queue 23c to the robots R1 and R2 by the group manager 20. However, the script data includes not only linguistic data but also non-linguistic data. The linguistic data is script data instructing the utterances of the robots R1 and R2, and the non-linguistic data is, for example, the movement of the robots R1 and R2, looking at the person H, nodding, shaking the head, bending the head, and the like. Script data that indicates non-linguistic operation. Further, a description indicating that it is a phase (intention confirmation phase) for confirming the intention (explicit or implicit) of the person H is added to such script data as necessary.

Further, a program (OS, sensor signal acquisition program, etc.) required for the group manager 20 is stored in the memory 20d. The group manager 20 operates according to the program stored in the memory 20d.

As described above, the CPU 20a of the group manager 20 controls the operation, that is, the behavior of each robot, and the history of the behavior is stored in the memory 20d described above and provided to the sensor manager 18 as needed. do.

Further, in the robot dialogue system 10 shown in FIG. 1, a human intention processing module 23 is provided. The human intention processing module 23 may be an independent computer, but in this embodiment, it is configured as one program of the above-mentioned group manager 20. Therefore, the human intention processing module 23 is controlled by the CPU 20a.

As will be described in detail later with reference to FIG. 15, the human intention processing module 23 does not necessarily depend on the actual intention of the person H, but the intention of the person H recognized by the sensor manager 18, for example, affirmation or denial. The script data (dialog) for the robots R1 and R2 to speak in order to deal with the intention of the certified person H is read from the dialog database 22 and stored in the pool 23a, which is an area of the memory 20d, respectively. Let (load).

The out-of-context response dialog pool 23b is loaded into the context or future set of dialogs or dialog queues 23c loaded into the next dialog candidate pool 23a, and is an exceptional or out-of-context that was not planned for the ongoing dialog. It is a storage area for storing a group of scripts that should perform an utterance or a non-verbal action in order to deal with the intention of the person H. The out-of-context intention display means, for example, the above-mentioned intention display device 17 that repeatedly hits the affirmative button 17a and / or the negative button 17n, and ignores the intention forever even though the robot makes an utterance waiting for a person's intention display. It is conceivable that it will not be displayed.

The dialog queue 23c is also, for example, an area in the memory 20d, and the dialog data 23c contains script data loaded in the next dialog candidate pool 23a and script data loaded in the out-of-context response dialog pool 23b. Can then be loaded in the form of a queue for immediate execution by robots R1 and / or R2.

For example, the execution time (tnext) of the script data included in the script is written in the script as a header, and in the dialog queue 23c, the script data is always sorted by the execution time (tnext), and the group manager 20 executes the script. The script data is sent to each robot controller 24 so that the script data having the same time can be executed at the same time. Therefore, for example, the robots R1 and R2 can perform the same operation at the same time, for example, looking at the person H, and the same robot R1 or R2 can see the other robot or the person H at the same time as speaking, for example.

With reference to FIG. 4, FIG. 4 shows the appearance of the robot R of the embodiment, and the robot R is provided on the table 30 so as to be able to rotate back and forth and left and right with respect to the table 30. That is, two degrees of freedom are set for the body 32.

From the left and right positions corresponding to the shoulders of the person on the torso 32, the right arm 34R and the left arm 34L are rotatably provided back and forth and left and right by shoulder joints (not shown), respectively. That is, two degrees of freedom are set for each of the right arm 34R and the left arm 34L.

A neck 36 is provided at the center of the upper end of the body 32, and a head 38 is provided above the neck 36. The neck 36, that is, the head 38 is attached to the body 32 so as to be able to rotate back and forth and left and right. That is, the neck 36, that is, the head 38 is set with three degrees of freedom: a roll angle (tilt left and right), a pitch angle (tilt back and forth), and a yaw angle (rotation left and right).

The right eye 40R and the left eye 40L are provided on the front surface of the head 38, that is, the surface corresponding to the human face, and the eyeballs 42R and 42L are provided on the right eye 40R and the left eye 40L. The right eye 40R and the left eye 40L can close and open the eyelids, and the eyeballs 42R and 42L can rotate up, down, left and right, respectively. That is, one degree of freedom is set for the right eye 40R and the left eye 40L, that is, the eyelid, and two degrees of freedom are set for the eyeballs 42R and 42L.

The face is further provided with a mouth 44, which can be closed and opened. That is, one degree of freedom is set in the mouth 44.

At the position of the human chest on the body 32, a speaker 46 for making an utterance to be heard by the person H in the dialogue system 10 and a microphone 48 for listening to the utterance voice of the environment, particularly the person H, are provided.

A camera 50 capable of shooting a moving image or a still image is built in a portion of the head 38 corresponding to the forehead of the face. The camera 50 can shoot a person H facing the other, and the camera signal (video signal) from the camera 50 is sent to the CPU 18a via the sensor I / F of the sensor manager 18 as in the environment camera 16 (FIG. 1). , May be entered.

FIG. 5 is a block diagram showing a robot controller 24 that is built in the robot R and controls the operation (utterance, gesture, etc.) of the robot R. With reference to FIG. 5, the robot controller 24 includes a CPU 24a, and a communication device 24c is connected to the CPU 24a via an internal bus 24b. The communication device 24c includes, for example, a network interface (NIC), and the CPU 24a communicates with the sensor manager 18, the group manager 20, an external computer, another robot (both not shown), and the like via the communication device 24c. You can exchange data between them.

Further, the memory 24d is connected to the CPU 24a via the internal bus 24b. The memory 24d includes a ROM and a RAM. The control data and script data sent from the group manager 20 are temporarily stored in the memory 24d.

Further, a program (OS, sensor signal acquisition program, etc.) required for robot control is stored in the memory 24d. The robot controller 24 controls the operation of the robot R according to the program stored in the memory 24d.

That is, an actuator control board 24e configured by, for example, a DSP is further connected to the CPU 24a of the robot controller 24, and the actuator control board 24e is provided in each of the above-mentioned parts of the robot R as described below. Controls the operation of.

The movement of the fuselage 32 with two degrees of freedom, that is, the rotation in the front-back and left-right directions is controlled by the CPU 24a controlling the fuselage actuator 52 through the actuator control board 24e.

The movement of the right arm 34R and the left arm 34L in two degrees of freedom, that is, the rotation in the front-back and left-right directions is controlled by the CPU 24a controlling the arm actuator 54 through the actuator control board 24e.

The movement of the neck 36, that is, the head 38 with three degrees of freedom, that is, the rotation of the front, back, left, and right is controlled by the head actuator 56 through the actuator control board 24e.

The right eye 40R and the left eye 40L, that is, the eyelid opening / closing operation, are controlled by the CPU 24a controlling the eyelid actuator 58 through the actuator control board 24e. The movement of the eyeball 42R and the eyeball 42L in two degrees of freedom, that is, the rotation in the front-back and left-right directions is controlled by the CPU 24a controlling the eyeball actuator 60 through the actuator control board 24e. The opening / closing operation of the mouth 44 is controlled by the CPU 24a controlling the mouth actuator 62 through the actuator control board 24e.

The speaker 46 of the robot R shown in FIG. 4 is connected to the CPU 24a of the robot controller 24. The CPU 24a is given by the group manager 20 and, if necessary, utters (utters) from the speaker 46 according to the script data stored in the memory 24d.

With such a robot controller 24, the head and arms of the robot R make necessary movements when necessary in the dialogue system 10, for example, when a non-verbal movement is required by a script. Specific control of the actuator and the like can be easily inferred from the above description, and is not necessarily described.

As shown in FIG. 1, each robot R1 and R2 is provided with a built-in robot sensor 26 similar to the robot controller 24. The robot sensor 26 includes an attitude sensor, an acceleration sensor, and the like for detecting the state of each movable component of the robots R1 and R2, and sensor signals from those sensors are input to the sensor manager 18. Therefore, the sensor manager 18 can sense the states of the robots R1 and R2 based on the sensor signal from the robot sensor 26.

The microphone 48 and the camera 50 of the robot R shown in FIG. 4 are input to the sensor manager 18 via the robot sensor 26. The sensor manager 18 stores the voice data captured from the microphone 48 in the memory 18d (FIG. 2), and executes the voice recognition process as needed. The sensor manager 18 also processes the camera signal from the camera 50 to sense the situation at the dialogue location 12.

Although only one sensor manager 18 is shown in the embodiment of FIG. 1, two or more sensor managers may be provided, and in that case, each sensor manager is a sensing item. Can be shared.

Similarly, if necessary, two or more group managers 20 may be used, and conversely, the sensor manager 18 and the group manager 20 may be realized by one computer.

Further, the robot R used in the dialogue system 10 of the embodiment of FIG. 1 is not limited to the robot described above with reference to FIG. 4, and may have at least a function of speaking according to a script.

FIG. 6 is a schematic diagram showing an example of a script in which the robot R interacts accordingly. In this script, first, the part where the robots R interact with each other (time zone) and the part where the robot R asks the person H (question or request for consent) (time zone) are clearly distinguished. In the embodiment, the flag for that purpose is "question flag Fah (FLAG_ASK_HUMAN)".

When the question flag Fah is reset, that is, when "Fah = 1" is not set, the robots R output the utterance sentence predetermined by the script data from the speaker 46 (FIGS. 4 and 5). Let person H, who is a participant, hear the content of the dialogue. When the question flag Fah is set, that is, when "Fah = 1", the robot R outputs an utterance sentence predetermined by the script data from the speaker 46 (FIGS. 4 and 5) and participates. Speak to person H, who is a person.

In the example of FIG. 6, the upper part (A) and the lower part (C) are script data for dialogue between the robots, and the middle part (B) asks the person H from the robot R. Script data for.

In the first line (a1), second line (a2), and third line (a3) of the script of the part (A), "tnext = XX" is defined, and here "XX1", "XX2", "XX3" specifies the time or time when each script data should be executed. "RR" describes speaking from robot to robot, and subsequent "SPEAKER = R1 (R2)" and "LISTNERR = R2 (R1)" are the speakers of the first robot R1 ( Alternatively, it is the second robot R2), and the listener is defined as the second robot R2 (or the first robot R1).

The description following "TEXT" in the script data is the utterance sentence that the robot should speak at that time. The description following "NONVERBAL" defines a non-verbal operation to be performed by the first robot R1 and / or the second robot R2, such as looking at the other robot or looking at the person H.

In part (B), the question flag Fah is set, that is, "Fah = 1", so either one of the two robots R1 and R2, in this example, from the first robot R1 to the person H. , The utterance of the question is output. Therefore, in the first line (b1), following the description "tnext = XX4" that defines the time, "RH: SPEAKER = R1, LISTNER = H" indicating that the first robot R1 speaks to a person is added. Will be described. "TEXT" and "NONVERBAL" are as described above.

Under the first line (b1) of the part (B), there are script data of condition 1, condition 2, and condition 3, but these are humans in response to the question from the robot R in the first line (b1). It stipulates that when H speaks, the content (reply) spoken by the robot next changes according to the conditions.

Condition 1 is script data in which robots talk to each other about an utterance from a person H, for example, "it is interesting". Therefore, following the description "tnext = XX5" that defines the time, the second robot R2 is the second. 1 “RR: SPEAKER = R2, LISTNER = R1” indicating that the robot speaks to R1 is described.

Condition 2 and condition 3 are both settings in which the robot reacts to the utterance of the person H, and both of them follow the description "T = XX5" that defines the time, and then the second robot R2 or the first robot. "RH: SPEAKER = R2, LISTNER = H" or "RH: SPEAKER = R1, LISTNER = H" indicating that R1 speaks to a person is described.

The utterance sentence of the condition 2 or the condition 3 is determined by the group manager 20 according to the voice recognition of the voice from the microphone 14 by the sensor manager 18 and the result of the voice recognition. The content of the utterance sentence is a script. Predefined as data. For example, when person H utters "I like XX" in response to the question from the robot "What do you like?" In the first line (b1), the robot says. Replies, "I like XX," or "I like △△ more than XX." In each case, the utterance is the result of the sensor manager 18 recognizing "○○" included in the utterance of the person H. Therefore, in both condition 2 and condition 3, the utterance sentence excluding "○○" is specified in advance.

Like part (A), part (C) is a part where robots talk to each other because the question flag Fah is reset, that is, it is not "Fah = 1", but basically part (B). ) Progresses independently of the speech from person H. Since the content of the script data is as explained in Part (A), no duplicate explanation is given here.

The time or time tnext that determines the utterance timing in the embodiment of FIG. 6 is updated (shifted) by the group manager 20 based on the instruction of the sensor manager 18, for example, when there is an utterance interruption (interrupt) from the person H. ) May be necessary. For example, the sensor manager 18 manages a flag (not shown) indicating the necessity of updating the utterance timing, and the group manager 20 rewrites T when the flag is set (when the flag is "1"). It should be.

In the robot dialogue system 10 of FIG. 1, the two robots R1 and R2 of the dialogue place 12 proceed with the dialogue according to the script as illustrated in FIG. 6, and control the utterances of these robots R1 and R2 in an integrated manner. It is the group manager 20 that does.

The flow chart shown in FIG. 7 shows the operation of the CPU 20a (FIG. 3) of the group manager 20. The operation of FIG. 7 is repeatedly executed at a speed of, for example, a frame rate.

In the first step S1, the CPU 20a executes initialization, for example, reading script data as illustrated in FIG. 6 from the dialog database 22 (FIG. 1).

In the following step S2, the CPU 20a updates the sensor data sent from the sensor manager 18 (FIGS. 1 and 3).

In the sensor manager 18, based on the sensor signals from the sensors such as the microphone 14 and the camera 16 shown in FIG. 1, the non-verbal parameter "NONVERBAL_PARAMETERS", for example, a person's "eyes are restless" or "yawned". , Sensing. Further, as described above, the sensor manager 18 can also receive a signal indicating a person's intention from the intention display device 17 and certify what kind of intention the person has expressed.

Further, in the sensor manager 18, based on the sensor signal as well, "is there a human voice", "is (someone) busy", "is it interrupted by language", or "is the robot responding non-verbally?" Is the person generating the behavior that should be done? "

In addition, the information necessary to determine the detailed parameters of the behavior of the robot (for example, when looking at the person H, the position of the person, and when nodding to the person, the intonation or the end point of the person's utterance. ) Needs to be sensed.

In addition, it is necessary to sense how individual robots behaved, such as "speaking," waiting "," listening "," sitting ", etc., but the behavior of these robots is It can be sensed by input from software that manages the history of robot behavior (for example, from group manager 20).

In step S2, the sensing result data (sensing data) from the sensor manager 18 is updated.

In the next step S3, the CPU 20a determines whether the dialog queue 23c has been updated. If "NO" in step S3, the process proceeds to step S5. If “YES”, the current time (t) is set to “0” and the execution time (tnext) is set to “-1” in step S4.

When "NO" is determined in step S3, in the next step S5, the CPU 20a of the group manager 20 determines whether or not there has been an interrupt due to the speech of the person H, based on the sensing data updated in step S2.

When "YES" is determined in this step S5, the CPU 20a determines whether or not the utterance recovery flag IS_RECOVER = Frc is set in the next step S7, that is, whether or not Frc = 1. Here, "YES" means that the robot R has recovered from the utterance collision caused by the utterance of the person detected in step S5, and the robot R needs to insert some utterance. Therefore, in step S9, the script data for the return phase for returning from the dialog interruption state due to the interruption due to the utterance of the person H read in step S1 is loaded, and in step S11, the script data is broadcast to the robot R, that is, the robot controller 24. Therefore, after that, the robot R speaks according to the script data for the return phase.

However, as described in FIG. 6, the script data to be loaded in this step S9 (the same applies to steps S17 and S19 described later) is "ID_SPEAKER" for the speaker, "ID_LISTNER" for the listener, and the attendant (participation). Person) includes information such as "ID_BYSTANDER", speech time "ONSET_TIME_SPEECH", and non-verbal operation "NONVERBAL_PARAMETERS".

For example, in response to a human speech interruption, the first robot R1 speaks "Sorry" to a person, the second robot R2 speaks "So?" To the first robot, or the first robot. From R1 to the second robot R2, "Oh yeah (the next utterance determined by the script)" is spoken. In other words, an utterance that responds lightly to a person's utterance interruption is inserted. Human utterances are not ignored, but the robot does not respond exactly to the utterance, but only the response determined by the script.

Then, after step S11, the process returns to step S2.

If "YES" in step S13, that is, when it is determined that the next utterance time predetermined by the script has come, the CPU 20a determines in step S15 whether or not the above-mentioned question flag Fah is set, that is, to the person H. Judge whether to speak a question to the question.

If "YES" in step S15, the question phase script data stored in the dialog queue 23c is loaded from the beginning in step S17, and it is broadcast in step S11.
In other words, human responses are not ignored.

Then, when "NO" is determined in step S15, that is, when it is determined that it is not the time to utter a question to a person, the CPU 20a determines script data (non-questioning script data) that is not for questioning in the next step S19. ) Is loaded.

In step S21, the CPU 20a determines whether the script still remains. That is, it determines whether the dialogue flag IS_CONVERSATION has been reset. If "YES", the process proceeds to step S11, but if "NO", a broadcast to that effect is performed in step S23, termination processing is performed in step S25, and this routine ends.

Both the robots R1 and R2 are controlled by the robot controller 24 according to the flow chart of FIG. The routine of FIG. 8 is also repeatedly executed, for example, at a frame rate, like the routine of FIG.

In step S31, the CPU 24a of the robot controller 24 performs initialization, and in the next step S33, the flag or data “IS_CONVERSATION” received from the sensor manager 18 and broadcast from the group manager 20 as shown in FIG. ”,“ IS_ONSPEECH (who is speaking) ”,“ CAN_BE_RECOVERED (whether it was turned from an utterance collision) ”,“ IS_INTERRUPTED_VERBAL (is there an interrupt by language) ”,“ IS_INTERRUPTED_NONVERBAL (is there an interrupt by non-language)) , "ID_SPEAKER", "ID_LISTNER", "ID_BYSTANDER", "SPEECH_TEXT", "ONSET_TIME_SPEECH", "NONVERBAL_PARAMETERS" Parameters: Parameters that specify non-verbal movements such as when and who to look at, nod, etc.) ”.

Then, in step S34, the CPU 24a refers to the data "IS_CONVERSATION" and determines whether or not the dialogue is still continued. If "YES", in the next step S35, the CPU 24a of the robot controller 24 determines whether or not the data broadcast from the group manager 20 received in step S33 includes script data for non-language operation. Then, if "YES" is determined in step S35, the CPU 24a issues an operation command to the actuator control board 24e of the corresponding robot R1 and / or R2 so as to execute the non-verbal operation command in the next step S36. give. Therefore, robots R1 and / or R2 perform broadcast non-verbal movements.

Then, if "NO" in the previous step S35, in the next step S37, the data "IS_ONSPEECH" is referred to to determine whether or not someone is speaking. If "YES", in step S39, the speaker at that time determines whether or not he / she is (the robot controlled by). In other words, refer to the data "ID_SPEAKER" to determine if you are speaking.

If you (the robot controlled by) are speaking, the result is "YES" in step S39, and the process proceeds to step S41. In step S41, the CPU 24a determines whether or not there is an interrupt (utterance cutoff) by the language with reference to the data “IS_INTERRUPTED_VERBAL”. If "YES" in step S41, it means that there is an interaction by language. Therefore, in the next step S43, the CPU 24a refers to the data "CAN_BE_RECOVERED" and determines whether or not the utterance may be resumed.

If "YES" in step S43, the robot is made to resume utterance in step S46, but if "NO", the robot is made to wait without making utterance in step S46.

When "NO" is determined in step S41, that is, when there is no language interaction, the CPU 24a determines in step S47 whether or not it was interrupted by non-language operation. For example, it is determined whether or not the person has irregular non-verbal movements, for example, the person H does not move significantly.

"When the robot decides that it should speak while generating a nonverbal response, such as when the person H is performing irregular nonverbal movements or is showing a listening movement such as nodding to the robot's talk. Is "YES" in step S47, and in the following step S48, the CPU 24a causes the robot R1 or R2 controlled by the CPU 24a to respond by a non-verbal operation based on the data "NONVERBAL_PARAMETERS". When "NO" in step S47, the process by the CPU 24a returns to step S33.

When "NO" in step S39, that is, when oneself (the robot controlled by the robot) is not instructed to speak, the data "ID_LISTNER" is referred to to determine whether or not oneself is designated as a listener.

If "YES" in step S49, the CPU 24a controls the robot R controlled by the CPU 24a so that it becomes a listener in step S51.

If "NO" in step S49, the data "ID_BYSTANDER" indicates that the robot R1 or R2 is a seated person, so that the robot controller 24, that is, the CPU 24a operates as a seated person with respect to the robot R1 or R2. To do.

When "NO" is determined in step S37, that is, when no one is speaking, in step S55, the CPU 24a determines whether or not the speaker is the speaker by referring to the data "ID_SPEAKER". If "YES", the robot R1 or R2 controlled by the robot is controlled to speak according to the content of the utterance sentence TEXT in step S57.

If "NO" in step S55, the robot is neither a speaker nor a listener. Therefore, the CPU 24a controls the robot to operate as a seater in step S59.

If "NO" in step S34, this routine is terminated after the termination process is executed in step S61.

As described above, all of the robots R only utter the contents determined in the script at the timing determined in the script broadcasted from the group manager 18, and do not utter spontaneously.

On the other hand, the robot R does not interpret the utterance from the person H. Even if there is an utterance from the person H, the robot R only strikes an aizuchi and does not respond to the utterance from that person. The utterance also speaks as decided in the script.

By using such a dialogue system of the robots R1 and R2 according to the script by the robot controller 24, a dialogue sequence is established as shown in FIG.

For example, you can clearly distinguish between a robot talking to a person and a robot talking to a robot, and at the same time, ignoring the human being or the robot who will be the seat in the dialogue. Based on the history of the dialogue and the estimation of the intentions of the participants, the robots generate utterances, gaze distribution, and gestures for asking questions to humans so that they do not look like they are. That is, since the robot R can listen to the utterance of the person H, it is possible to ask the person H a question and have the person H speak. This makes it possible for a person to feel as if he / she is also participating in the dialogue between the robots. As a result, it is possible to impress humans with the contents included in the dialogue between robots.

Specifically, the scenes shown in FIGS. 10-14 can be intentionally made to appear by a script. However, the specific example of FIGS. 10-14 is not directly related to the intention display of person H.

In the example of FIG. 10, when the first robot R1 and the second robot R2 switch from the topic A to the topic B when the dialogue is progressing on the topic (Topic or theme) A while repeating the utterance. , On the script, control the robots R1 and / or R2 to ask the person H a question. The content of the question at this time may be, for example, a topic A that has already been completed, or a topic B that is about to begin. It may be related to both. Then, when there is an answer from the person H, the robots R1 and / or R2 utter an aizuchi such as "I'm sorry" according to the script.

That is, the robot does not ask the person H during the dialogue between the robots, and the robot speaks to the person H at the timing of switching the topic. Then, the person H can maintain the state with the consciousness of participating in the dialogue between the robots.

However, the robot may interrupt the response from the person H. For example, robot R2 "○○ is XX, so it is △△.", Robot R1 "Eh, but isn't ○○ □□?", Robot R2 "Oh, but (enter topic B). It is also possible to interrupt such as.

In the example of FIG. 11, when the robots R1 and R2 are repeating a question or an answer and proceeding with a dialogue on some topic, the same question that the robot asks the robot is uttered to the person H. For example, the robots R1 and / or R2 make the person H ask a question such as "What do you think?". Then, there is an answer from person H. The robots R1 and R2 utter the same aizuchi as in FIG. 10 to the robots R1 and R2.

In the example of FIG. 12, contrary to the case of FIG. 11, when the robots are interacting with each other, the same question that the robot R1 or R2 asks the person H is asked to the other robot R2 or R1. In this case, the person H will be able to have a "sense of dialogue" by including the same question as the question to himself in the dialogue between the robots.

In the example of FIG. 13, when robots R1 and R2 are proceeding with a dialogue on some topic while repeating questions and answers, for example, the second robot R2 explains something to the first robot R1 and starts from the first robot R1. Make an utterance to express your impressions. Just as the person H agrees with the impression, the robot R1 or R2 asks the person H for the consent, for example, "ne". In this case as well, if the story is convincing to some extent, it is possible to utter consent and encourage consent in the heart, so human H should be aware that he / she is participating in the dialogue between robots. Can be sustained.

In the example of FIG. 14, for example, the second robot R2 to the first robot R1
Ask or explain to the person, and ask the person that the first robot R1 intentionally made a mistake, misunderstood, or did not understand, and asked the person, "This is not the case." Encourage them to make corrections and supplementary utterances. However, if there is no utterance from a person, the second robot R2 makes corrections or supplements instead and sends the utterance of the first robot. People actively participate in the dialogue, which enhances the sense of dialogue.

When there is an intention display by the person H who is a seated person detected by the intention display device 17, the person intention processing module 23 follows the flow diagram shown in FIG. 15 from the dialog database 22 to the next dialog candidate pool 23a or out of context. A dialog (script data) is loaded in the response, and a dialog (script data) is loaded from the next dialog candidate pool 23a into the dialog queue 23c.

Specifically, in the first step S71 of FIG. 15, the CPU 20a that controls the human intention processing module 23 executes initialization and reads a dialog (script data) from the dialog database 22 into the next dialog candidate pool 23a. That is, it is decided which topic or theme dialog is to be read, and the dialog of the topic or theme group is read into the next dialog candidate pool 23a. At this time, the constraint conditions such as the dependency of the dialog, for example, which dialog should not be executed before this dialog, and this dialog should be executed at the same time as a specific dialog are also read. At the same time, in step S71, the necessary dialog is loaded into the dialog queue 23c from the next dialog candidate pool 23a.

The dialog at this time is set, for example, so that one robot R1 consistently speaks or operates in favor of the topic and the other robot R2 consistently speaks or operates in the opposite position. Therefore, each robot shall not change its position from beginning to end.

In the next step S73, the CPU 20a updates the sensor data from the sensor manager 18.

Then, in the next step S75, the CPU 20a determines whether or not the intention of the person to be approved is detected in the updated sensor data.

When "YES" is determined in step S75, that is, when the robots R1 and / or R2 have a person's intention to approve, in step S77, the CPU 20a is an approval dialog (mainly non-) for the person's expression of intention. Language operation) is loaded into the dialog queue 23c.

However, if the approval dialog interferes with each robot's scheduled dialog (including non-verbal movements), the approval dialog will not be loaded. For example, when one of the robots is scheduled to see the person H next time, if the approval dialog is "the robot sees the person H", the approval dialog cannot be put in the dialog queue 23c. .. Furthermore, if one of the robots is scheduled to see another robot and the approval dialog is "the robot sees person H", then the robot sees the other robot and immediately sees the person. You have to look at H, and the robot behaves unnaturally (interference). Therefore, even at this time, the approval dialog is not loaded into the dialog queue 23a.

That is, in step S77, only the dialog that does not interfere with other dialogs is loaded into the dialog queue 23c as an approval dialog that approves the intention of the person. Therefore, according to the approval dialog funneled to the dialog queue 23c, the intention of the person H detected in step S75 is approved by a non-verbal operation such as "both robots R1 and R2 see the person H".

When "NO" is determined in step S75, that is, when the robot does not detect the intention display of the person to be approved, the process proceeds to step S79 as it is.

In step S79, it is determined whether or not the phase is to confirm the intention of the person. For example, if the previous utterance is in the form of a question or negative sentence to person H, set the script of the question or negative sentence to confirm the person's intention and the timing to confirm (for example). In the question flag Fah) of FIG. 6 and step S79, it is determined whether or not the timing is reached.

Then, in step S81, it is determined whether or not the intention of the person H can be recognized. For example, if it is clear that the person H operates the affirmative button 19a or the negative button 19n of the intention display device 17 to input the intention, the CPU 20a can recognize the intention of the person. On the other hand, when the person H does not operate the intention display device 17 and ignores the question sentence, the CPU 20a cannot definitively recognize the intention of the person.

In such a case, although not shown in FIG. 15, the human intention processing module 23 loads the dialog necessary for confirming the intention of the human H into the dialog queue 23c, and one of the robots R speaks. Depending on the non-verbal movement, or both, the person H may be encouraged to confirm his / her intention. For example, confirm the intention of human H by non-verbal movements such as robots R1 and / or R2 looking at human H, and / or by utterances such as "what do you think" of robots R1 and / or R2. can do. Then, after executing such a decision determination dialog, the process may be returned to the previous step S73 to determine whether or not the intention can be recognized again.

However, when there is no expression of intention from the person H for a predetermined time, for example, 5 seconds, the intention of the person H can be regarded as either "affirmation" or "denial".

When the intention of the person H can be definitively recognized in this way, the process proceeds from step S81 to the next step S83. Then, in step S83, the CPU 20a controls the human intention processing module 23 and selects a dialog (script data) for approving the intention from the next dialog candidate pool 23a based on the intention of the authorized person H. And load it into the dialog queue 23c. Therefore, at this time, the robots R1 and R2 execute a dialogue for approving the intention of the person H according to the approval dialog funneled in the dialog queue 23c. As for the dialog at this time, as described above, different dialogs are loaded in the dialog queue 23c depending on whether the person H has made a positive intention display or a negative intention display.

In step S85, since the CPU 20a loaded the dialog into the dialog queue 23c in step S83, the next dialog candidate is read from the dialog database 22 and loaded into the next dialog candidate pool 23a.

In the previous step S79, if it is not in the human intention recognition phase, "NO" is determined, and in the next step S87, the CPU 20a determines whether or not the sensor manager 18 has detected an uncontextual intention of the person H that cannot be ignored. do. If there is no indication of intention outside the context of such person H, the result is “NO” in this step S87, and the process proceeds to the next step S89.

In step S89, it is determined whether to continue the dialogue. For example, when there are no more available dialogs in the dialog database 22, or there is no more time, it is determined as "NO" in step S89, and the operation of the human intention processing module 23 of FIG. 15 ends.

When it is determined as "YES" in step S89, that is, when the dialogue should be continued, in step S91, it is determined whether or not the dialog to be executed remains in the dialog queue 23c, that is, whether or not the dialog is terminated.

If "YES" in step S91, as in step S71, in step S93, a group of dialogs (script data) of another topic is selected from the dialog database 22 and read, and loaded into the next dialog candidate pool 23a. At the same time, the dialog loaded in the next dialog candidate pool 23a is loaded in the dialog queue 23c for the next dialogue between the robots R1 and R2. After that, the process proceeds to step S85, and the dialog is further loaded from the dialog database 22 into the next dialog candidate pool 23a.

If "NO" is set in step S91, the process returns to the previous step S73 if the dialog is not completed.

Further, when it is determined as "YES" in step S87, that is, when there is an intention indication outside the context of person H. In step S95, the CPU 20a selects an appropriate dialog among the dialogs loaded in the out-of-context response dialog pool 23b shown in FIG. 1 based on the intention of the person H outside the context at that time, and the selected dialog is selected. Is loaded into the dialog queue 23c. Therefore, the robots R1 and / or R2 perform utterances and / or non-verbal movements in response to the above-mentioned intentional expression of the person H outside the context.

Then, in the next step S97, the CPU 20a deletes or suspends the contents of the current dialog queue 23c and the next dialog candidate pool 23b based on the will of the person. That is, the dialog that is considered unnecessary to be executed based on the intention of the person H at that time is deleted from the dialog queue 23c and the next dialog candidate pool 23b, or the execution of the dialog that is considered unnecessary to be executed is suspended.

As described above, in the human intention processing module 23, the robots R1 and R2 utterances and non-languages for approving the intention of the person H who is a seated person, for example, expressed by using the intention display device 17. By executing the movement, the human H clearly participates in the dialogue (dialogue feeling) in the dialogue system 10 including the robots R1 and R2, even if the human H does not speak anything. To reach.

Hereinafter, various examples of the approval dialog for the intention display of the person H will be described with reference to FIGS. 16 to 24. However, it should be noted that all dialogues are executed according to the dialogs (script data) preset in the dialog database 22 on the premise of the following.

The "Topic" circled in an oval is the "topic" described in FIGS. 10-14, and means that Topic-1, Topic-2, ... Are different topics. Topic indicates the utterance in the position of support, and "Anti-Topic" surrounded by a shaded oval indicates the utterance in the position opposite to Topic. In addition, "Follow-Topic" indicates an utterance in a position that follows Topic. Further, in the person H, the “affirmation” indicated by the white circle means that the sensor manager 18 has detected the affirmative intention display of the person H, for example, when the affirmative button 17a of the intention display device 17 is operated. do. On the contrary, the “denial” indicated by the shaded circle means that the sensor manager 18 has detected the negative intention of the person H. In the explanation of the following specific example, the "indication of a person's intention" is not only the intention that the person H independently or implicitly expresses, but also the actual intention of the person H. However, as mentioned above, it should be understood that it refers to the intention to be certified by the sensor manager 18.

In the example of FIG. 16, the robot R1 speaks in favor of Topic-1 and the robot R2 speaks in the opposite position to the robot R1 with the initiative of the dialogue.

According to the dialog after uttering the question sentence included in Topic-1 (Anti-Topic-1), the CPU 20a of the group manager 20 causes the robots R1 and R2 to perform a non-verbal operation of seeing the person H together. The script data (instruction of the non-language operation) held in the queue 23c is broadcast. Therefore, the person H understands that his / her intention is required.

After the person H makes an affirmative indication in response to such non-verbal movements of the robots R1 and R2, for example, in step S83 of FIG. 15, the robots R1 and R2 are shown by a horizontally long rectangular frame in FIG. Run the approval dialog. In this example, since the person H has made an affirmative intention, the robot R1 utters, for example, "That's right" for approval of the intention. On the other hand, in the robot R2, since the person H has made an affirmative intention even though he / she is in the opposite position, for example, in order to approve the intention display, the person H has an affirmative intention. Make an utterance that blames the display.

That is, here, the robots R1 and R2 execute utterances in which the robots R1 and R2 approve the affirmative intention of the person H in their respective positions.

The utterance in this approval dialog may be the utterance of either the robots R1 or R2 first, and the utterances of the robots R1 and R2 may be executed at the same time. Further, in this approval dialog, the dialogue between the robots R1 and R2 may be executed not only once as in the embodiment but also a plurality of times. It should be pointed out in advance that these things also apply to the following examples of FIGS. 17-20.

After that, the robot R1 having the initiative speaks a connecting utterance to return to the dialogue in the same Topic-1 or another Topic-2, for example, "Yeah, that's why". After that, the robots R1 and R2 continue Topic-1 (or 2) (Anti-Topic-1 (or 2)).

In the example of FIG. 16, the robot R1 approves the affirmative intention of the person H by making an utterance that synchronizes with the affirmative intention of the person H, and the robot R2 makes an utterance that does not synchronize with the affirmative intention of the person H. By doing so, the person H's affirmative intention is approved.

In the example of FIG. 17, the robot R1 speaks in favor of Topic-1 and the robot R2 speaks in a position of following the robot R1 with the initiative of the dialogue. After the utterance of Topic-1 (Follow-Topic-1), as in the case of FIG. 16, the robots R1 and R2 both perform a non-verbal operation of looking at the human H, thereby prompting the intention display of the human H. ..

Then, when the person H indicates an affirmative intention, for example, in step S83 of FIG. 15, the robots R1 and R2 execute an approval dialog as shown by a horizontally long rectangular frame in FIG. In this example, since the person H has made a positive intention statement, the robot R1 utters, for example, "That's right" to approve the intention statement. On the other hand, since the robot R2 is in a position to follow the robot R1, he utters, for example, "That's right" in order to approve the affirmative intention of the person H.

That is, here, the robots R1 and R2 execute utterances in which the robots R1 and R2 approve the affirmative intention of the person H in their respective positions.

After that, the robot R1 having the initiative makes a tie utterance to return to the dialogue in the same Topic-1 or another Topic-2, for example, a synchronized utterance such as "Yeah, that's why". After that, the robots R1 and R2 continue Topic-1 (or 2) and Follow-Topic-1 (or 2).

In the example of FIG. 17, the robot R1 approves the affirmative intention of the person H by making an utterance in synchronization with the affirmative intention of the person H, and the robot R2 in the position of following the robot R1 is also the affirmative of the person H. Approve the affirmative intention of person H by making an utterance in sync with the intention.

In the example of FIG. 18, the robot R1 speaks in favor of Topic-1 and the robot R2 speaks in the opposite position to the robot R1 with the initiative of the dialogue. After the utterance of Topic-1 (Anti-Topic-1), as in the case of FIG. 16, the robots R1 and R2 both perform a non-verbal operation of looking at the human H, thereby prompting the intention display of the human H. ..

After the person H indicates a negative intention, for example, in step S83 of FIG. 15, the robots R1 and R2 execute an approval dialog as shown by a horizontally long rectangular frame in FIG. In this example, since the person H has made a negative intention, the robot R1 rebukes the negative intention of the person H in order to approve the intention, for example, "Oh, no". On the other hand, in the robot R2, since the person H has expressed a negative intention even though he is in the opposite position, the person who utters, for example, "Isn't it?" Approve the negative intention of person H.

That is, here, the robots R1 and R2 execute utterances in which the robots R1 and R2 approve the negative intention display of the person H in their respective positions.

After that, the robot R1 having the initiative makes a connecting utterance to return to the dialogue in the same Topic-1 or another Topic-2, for example, a turning utterance such as "No, but". After that, the robots R1 and R2 continue Topic-1 (or 2) (Anti-Topic-1 (or 2)).

In the example of FIG. 18, the robot R1 approves the negative intention of the person H from the standpoint of approval by making an utterance that does not synchronize with the intention of the person H, and the robot R2 makes an utterance that does not synchronize with the intention of the person H. By doing so, he approves the negative intention of person H from the opposite standpoint.

In the example of FIG. 19, the robot R1 speaks in favor of Topic-1 and the robot R2 speaks in the opposite position to the robot R1 with the initiative of the dialogue. After the utterance of Topic-1 (Anti-Topic-1), as in the case of FIG. 16, the robots R1 and R2 both perform a non-verbal operation of looking at the human H, thereby prompting the intention display of the human H. ..

After the person H has expressed affirmative intention, for example, in step S83 of FIG. 15, the robots R1 and R2 execute an approval dialog as shown by a horizontally long rectangular frame in FIG. In this example, since the person H has made a positive intention statement, the robot R1 utters, for example, "That's right" to approve the intention statement. On the other hand, in the robot R2, since the person H has made an affirmative intention even though he / she is in the opposite position, for example, in order to approve the intention display, the person H has an affirmative intention. Make an utterance that blames the display.

That is, here, the robots R1 and R2 execute utterances in which the robots R1 and R2 approve the affirmative intention of the person H in their respective positions.

After that, the initiative of the dialogue is transferred to the robot R2, and the robot R2 utters a connecting utterance for returning to the dialogue in Topic-1 (or 2), for example, "No, that's why". After that, the robots R2 and R1 continue Topic-1 (or 2) (Anti-Topic-1 (or 2)).

In the example of FIG. 19, the robot R2 approves the affirmative intention of the person H from the standpoint of approval by making an utterance in synchronization with the affirmative intention of the person H, and the robot R2 becomes the affirmative intention of the person H. Approve the affirmative intention of person H from the opposite standpoint by making disagreeable utterances.

In the example of FIG. 20, the robot R1 speaks in favor of Topic-1 and the robot R2 speaks in the opposite position to the robot R1 with the initiative of the dialogue. After the utterance of Topic-1 (Anti-Topic-1), as in the case of FIG. 16, the robots R1 and R2 both perform a non-verbal operation of looking at the human H, thereby prompting the intention display of the human H. ..

After the person H indicates a negative intention, for example, in step S83 of FIG. 15, the robots R1 and R2 execute an approval dialog as shown by a horizontally long rectangular frame in FIG. In this example, since the person H has made a negative intention statement, the robot R1 utters, for example, "Oh, no" to approve the intention statement. On the other hand, in the robot R2, although the person H is in the opposite position, the person H has made a negative intention, so for the approval of the intention, for example, "It's right", the person H's denial. Make an utterance that is in sync with the expression of intention.

That is, here, the robots R1 and R2 execute utterances in which the robots R1 and R2 approve the negative intention display of the person H in their respective positions.

After that, the initiative of the dialogue is transferred to the robot R2, and the robot R2 utters a connecting utterance for returning to the dialogue in the same Topic-1 (or 2), for example, "No, but ne". After that, the robots R2 and R1 continue Topic-1 (or 2) (Anti-Topic-1 (or 2)).

In the example of FIG. 20, the robot R2 approves the negative intention of the person H from the standpoint of approval by making an utterance that does not synchronize with the negative intention of the person H, and the robot R2 becomes the negative intention of the person H. Approve the negative intention of person H from the opposite standpoint by making a synchronized utterance.

In the example of FIG. 21, the robot R1 speaks in favor of Topic-1 and the robot R2 speaks in the opposite position to the robot R1 with the initiative of the dialogue. After the utterance of Topic-1 (Anti-Topic-1), as in the case of FIG. 16, the robots R1 and R2 both perform a non-verbal operation of looking at the human H, thereby prompting the intention display of the human H. ..

After the person H makes an affirmative intention, the robot R1 executes a dialog for expanding and interpreting and approving the affirmative intention of the person H, for example, in step S83 of FIG. First, since person H expresses affirmative intention and he is in a position of support, he utters, for example, "That's right" so as to be in tune with person H. Then, an utterance that expands and interprets the affirmative intention display of the person H is executed for the robot R2.

For example, if Topic-1 is a topic on the theme of luxury items, when Robot R1 utters "Banana is delicious, isn't it?" In Topic-1, when Human H makes a positive intention, Robot R1 First speaks to person H, "That's right." After that, he expands and interprets the affirmative intention of person H to the robot R2 and utters, for example, "I like fruits". On the other hand, the robot R2 speaks to the robot R1 to reinforce and approve the expanded interpretation of the robot R1, such as "Well, do you like fruits?".

That is, when the person H makes an affirmative intention for the dialogue between the robots R1 and R2 in the Topic, the robot R1 having the initiative of the dialogue expands and interprets the affirmative intention of the person H and approves it. By executing the dialog and subsequently executing the dialog in which the robot R2 approves the expanded interpretation, it is felt that the person H has actually spoken the expanded interpretation. In the above example, person H has the feeling that he actually said "I like fruits".

After that, the robots R1 and R2 execute an approval dialog as shown by a horizontally long rectangular frame in FIG. 21. In this example, the robot R2 approves the expanded interpretation of the robot R1 for the affirmative intention of the person H, so that the robot R1 responds to the approval of the expanded interpretation, for example, "I'm sure it's that type." The robot R2 utters to the robot R2, and since the person H has made an affirmative intention, he utters, for example, "terrible" that rebukes the affirmative intention of the person H from the opposite position. ..

That is, here, the robots R1 and R2 execute utterances that approve the affirmative intention display of the person H on the premise of the expanded interpretation from their respective positions.

The utterance in this approval dialog may be the utterance of either the robots R1 or R2 first, and the utterances of the robots R1 and R2 may be executed at the same time. Further, in this approval dialog, the dialogue between the robots R1 and R2 may be executed not only once as in the embodiment but also a plurality of times, or may be omitted. It should be pointed out in advance that these things also apply to the following examples of FIGS. 22-25.

The robot R1 with the initiative then speaks a tether to initiate a claim in the same Topic-1 or another Topic-2, such as "Yeah, that's it." After that, the robots R1 and R2 continue Topic-1 (or 2) (Anti-Topic-1 (or 2)).

In the example of FIG. 21, after the robots R1 and R2 approve the expanded interpretation of the affirmative intention of the person H, the robot R1 makes an utterance synchronized with the person H, and the affirmative intention of the person H is in favor of the robot R1. The display is approved, and the robot R2 approves the affirmative intention of the person H from the opposite position by making an utterance that does not synchronize with the affirmative intention of the person H.

In the example of FIG. 22, the robot R1 speaks in favor of Topic-1 and the robot R2 speaks in a position of following the robot R1 with the initiative of the dialogue. After the utterance of Topic-1 (Follow-Topic-1), as in the case of FIG. 16, the robots R1 and R2 both perform a non-verbal operation of looking at the human H, thereby prompting the intention display of the human H. ..

After the person H makes an affirmative intention, the robot R1 executes a dialog for expanding and interpreting and approving the affirmative intention of the person H, for example, in step S83 of FIG. First, since the person H expresses a positive intention and is in a position to follow him, he utters, for example, "That's right" so as to synchronize with the person H. Then, an utterance that expands the affirmative intention display of the person H is executed for the robot R2.

For example, if Topic-1 is a topic on the theme of luxury items, when robot R1 utters "Banana is delicious, isn't it?" In Topic-1, when person H makes a positive intention, the robot R1 first speaks to person H, "That's right." After that, he expands and interprets the affirmative intention of person H to the robot R2 and utters, for example, "I like fruits". On the other hand, the robot R2 makes an utterance to the robot R1 to reinforce and approve the expanded interpretation of the robot R1, such as "I understand the story".

That is, when the person H makes an affirmative expression for the dialogue between the robots R1 and R2 in the Topic (Follow-Topic), the robot R1 having the initiative of the dialogue expands the affirmative intention of the person H. By executing the dialog for interpreting and approving, and subsequently executing the dialog for the robot R2 to approve the expanded interpretation, it is felt that the person H has actually spoken the expanded interpretation. In the above example, person H has the feeling that he actually said, "I like fruits."

After that, the robots R1 and R2 execute an approval dialog as shown by a horizontally long rectangular frame in FIG. 22. In this example, since the robot R2 approves the expanded interpretation of the robot R1 for the affirmative intention of the person H, the robot R1 responds to the approval of the expanded interpretation, for example, "I feel like that type. "It was terrible," he said to the robot R2, and since the robot R2 had a positive intention expression by the person H, he was in a position to follow the robot R1 and synchronized with the affirmative intention expression of the person H. For example, say "That's right."

That is, here, the robots R1 and R2 execute utterances that approve the affirmative intention display of the person H on the premise of the expanded interpretation from their respective positions.

After that, the robot R1 having the initiative makes a tethered utterance to return to the dialogue in the same Topic-1 or another Topic-2, for example, "Yeah, that's it". After that, the robots R1 and R2 continue Topic-1 (or 2) (Anti-Topic-1 (or 2)).

In the example of FIG. 22, after the robots R1 and R2 approve the expanded interpretation of the affirmative intention of the person H, the robot R1 makes an utterance synchronized with the person H, and the affirmative intention of the person H is in favor of the robot R1. The display is enlarged and interpreted, and the robot R2 also approves the affirmative intention of the person H from the standpoint of following the robot R1 by making an utterance in synchronization with the affirmative intention of the person H.

In the example of FIG. 23, the robot R1 speaks in favor of Topic-1 and the robot R2 speaks in opposition to the robot R1 with the initiative of the dialogue. After the utterance of Topic-1 (Anti-Topic-1), as in the case of FIG. 16, the robots R1 and R2 both perform a non-verbal operation of looking at the human H, thereby prompting the intention display of the human H. ..

After the person H makes a negative intention display, the robot R2 executes a dialog for expanding and interpreting and approving the negative intention display of the person H, for example, in step S83 of FIG. First, since the person H expresses a negative intention and he is in the opposite position, he utters, for example, "That's right" so as to be in tune with the person H. Then, an utterance that expands the negative intention display of the person H is executed for the robot R2.

For example, if Topic-1 is a topic on the theme of luxury items, when robot R1 utters "Banana is delicious, isn't it?" In Topic-1, when person H expresses a negative intention, the robot R2 first speaks to person H, "That's right." After that, he expands and interprets the negative intention of person H to the robot R1 and utters, for example, "I hate fruits." On the other hand, the robot R1 speaks to the robot R2 to reinforce and approve the expanded interpretation of the robot R2, for example, "Do you hate fruits?"

That is, when the human H expresses a negative intention to the dialogue between the robots R1 and R2 in the Topic (Anti-Topic), the robot R2 opposed to the robot R1 expands and interprets the negative intention display of the human H. By executing the dialog for approving the robot R1 and subsequently executing the dialog for the robot R1 approving the expanded interpretation, it is felt that the person H has actually spoken the expanded interpretation. In the above example, person H has the feeling that he actually said "I don't like fruits".

After that, the robots R2 and R1 execute an approval dialog as shown by a horizontally long rectangular frame in FIG. 23. In this example, the robot R1 approves the expanded interpretation of the robot R2 for the negative intention of the person H, so the robot R2 responds to the approval of the expanded interpretation, for example, "I'm sure it's that type." And the robot R1 made a negative intention, so the robot R1 does not agree with the negative intention of the human H because he is in a position of being opposed to the robot R2. It's terrible. "

That is, here, the robots R1 and R2 execute an utterance that approves the negative intention display of the person H on the premise of the expanded interpretation from their respective positions.

After that, the robot R1 having the initiative speaks a tie to return to the dialogue in the same Topic-1 or another Topic-2, for example, "No, but". After that, the robots R1 and R2 continue Topic-1 (or 2) (Anti-Topic-1 (or 2)).

In the example of FIG. 23, after the robots R1 and R2 approve the expanded interpretation of the negative intention of the human H, the robot R2 makes an utterance synchronized with the human H, and the negative intention of the human H is in the opposite position. The display is enlarged and approved, and the robot R1 approves the negative intention of the person H by making an utterance that does not synchronize with the negative intention of the person H.

In the example of FIG. 24, the robot R1 speaks in favor of Topic-1 and the robot R2 speaks in the opposite position to the robot R1 with the initiative of the dialogue. After the utterance of Topic-1 (Anti-Topic-1), as in the case of FIG. 16, the robots R1 and R2 both perform a non-verbal operation of looking at the human H, thereby prompting the intention display of the human H. ..

After the person H makes an affirmative intention, the robot R1 executes a dialog for expanding and interpreting and approving the affirmative intention of the person H, for example, in step S83 of FIG. First, since person H expresses affirmative intention and he is in a position of support, he utters, for example, "That's right" so as to be in tune with person H. Then, an utterance that expands the affirmative intention display of the person H is executed for the robot R2.

For example, if Topic-1 is a topic on the theme of luxury items, when robot R1 utters "Banana is delicious, isn't it?" In Topic-1, when person H makes a positive intention, the robot R1 first speaks to person H, "That's right." After that, he expands and interprets the affirmative intention of person H to the robot R2 and utters, for example, "I like fruits". On the other hand, the robot R2 speaks to the robot R1 to reinforce and approve the expanded interpretation of the robot R1, such as "Well, do you like fruits?".

That is, when the person H makes an affirmative intention for the dialogue between the robots R1 and R2 in the Topic, the robot R1 having the initiative of the dialogue expands and interprets the affirmative intention of the person H and approves it. By executing the dialog and subsequently executing the dialog in which the robot R2 approves the expanded interpretation, it is felt that the person H has spoken the expanded interpretation. In the above example, person H has the feeling that he really likes fruits. "

After that, the robots R1 and R2 execute an approval dialog as shown by a horizontally long rectangular frame in FIG. 21. In this example, the robot R2 approves the expanded interpretation of the robot R1 for the affirmative intention of the person H, so that the robot R1 responds to the approval of the expanded interpretation, for example, "I'm sure it's that type." The robot R2 utters to the robot R2, and since the person H has made an affirmative intention, he utters, for example, "terrible" that rebukes the affirmative intention of the person H from the opposite position. ..

That is, here, the robots R1 and R2 execute utterances that approve the affirmative intention display of the person H on the premise of the expanded interpretation from their respective positions.

After that, the initiative is transferred to the robot R2, and the robot R2 utters a tie to start the claim in Topic-1 (or 2), for example, "No, but". After that, the robots R2 and R1 continue Topic-1 (or 2) (Anti-Topic-1 (or 2)).

In the example of FIG. 24, after the robots R1 and R2 approve the expanded interpretation of the affirmative intention of the person H, the robot R2 makes an utterance synchronized with the person H, so that the affirmative intention of the person H is in the opposite position. The display is enlarged and approved, and the robot R1 approves the affirmative intention of the person H by making an utterance that does not synchronize with the affirmative intention of the person H.

In the above-described embodiment, the intention display of the person H is input according to, for example, the operation of the person of the intention display device 17. However, for example, the image signal from the camera 16 may be processed so that, for example, when the person H nods, there is an indication of "affirmation", and when he shakes his head, there is an indication of "denial".

In the example of FIG. 25, the robot R1 speaks in favor of Topic-1 and the robot R2 speaks in opposition to the robot R1 with the initiative of the dialogue. After the utterance of Topic-1 (Anti-Topic-1), as in the case of FIG. 16, the robots R1 and R2 both perform a non-verbal operation of looking at the human H, thereby prompting the intention display of the human H. ..

After the person H makes a negative intention display, the robot R2 executes a dialog for expanding and interpreting and approving the negative intention display of the person H, for example, in step S83 of FIG. First, since the person H expresses a negative intention and he is in the opposite position, he utters, for example, "That's right" so as to be in tune with the person H. Then, an utterance that expands the negative intention display of the person H is executed for the robot R2.

For example, if Topic-1 is a topic on the theme of luxury items, when robot R1 utters "Banana is delicious, isn't it?" In Topic-1, when person H expresses a negative intention, the robot R2 first speaks to person H, "That's right." After that, he expands and interprets the negative intention of person H to the robot R1 and utters, for example, "I hate fruits." On the other hand, the robot R1 speaks to the robot R2 to reinforce and approve the expanded interpretation of the robot R2, for example, "Do you hate fruits?"

That is, when the human H expresses a negative intention to the dialogue between the robots R1 and R2 in the Topic (Anti-Topic), the robot R2 opposed to the robot R1 expands and interprets the negative intention display of the human H. By executing the dialog for approving the robot R1 and subsequently executing the dialog for the robot R1 approving the expanded interpretation, it is felt that the person H has actually spoken the expanded interpretation. In the above example, person H has the feeling that he actually said "I don't like fruits".

After that, the robots R2 and R1 execute an approval dialog as shown by a horizontally long rectangular frame in FIG. 23. In this example, the robot R1 approves the expanded interpretation of the robot R2 for the negative intention of the person H, so the robot R2 responds to the approval of the expanded interpretation, for example, "I'm sure it's that type." And the robot R1 made a negative intention, so the robot R1 does not agree with the negative intention of the human H because he is in a position of being opposed to the robot R2. It's terrible. "

That is, here, the robots R1 and R2 execute an utterance that approves the negative intention display of the person H on the premise of the expanded interpretation from their respective positions.

After that, the initiative is transferred to the robot R2, and the robot R2 utters the utterance of the connection of the start of the claim in Topic-1 (or 2), for example, "No, but". After that, the robots R2 and R1 continue Anti-Topic-1 (or 2) (Topic-1 (or 2)).

In the example of FIG. 25, after the robots R1 and R2 approve the expanded interpretation of the negative intention of the human H, the robot R2 makes an utterance synchronized with the human H, so that the negative intention of the human H is in the opposite position. The display is enlarged and approved, and the robot R1 approves the negative intention of the person H by making an utterance that does not synchronize with the negative intention of the person H.

In the specific example shown in FIGS. 16 to 25, when the intention of the person H is recognized, the intention is approved by the dialogue between the robots R1 and R2, but step S83 in FIG. 15 is an example of the execution means. In the approval dialog shown by, the intention of the person H may be approved by the utterance of at least one agent (at least one of the robots R1 and R2 in the embodiment).

Further, in order to approve the intention of the certified person H, the non-verbal operation in step S77 of FIG. 15 may be executed. In this case, step S77 also constitutes the execution means.

In the above embodiment, the robots R1 and the dialog database 22 are stored in advance. However, instead of this database 22, necessary script data may be sequentially supplied to the group manager 20 from the net, for example.

Further, the above-mentioned embodiment is a robot dialogue system using a robot which is a physical agent, but the present invention is not limited to such a physical agent, but also an avatar displayed on the screen of a display, for example. Alternatively, an agent-to-agent dialogue system such as a character may be applied. In this case, the robot controller 24 and the robot sensor 26 in FIG. 1 are replaced with a display controller (not shown) for displaying such avatars and characters. That is, the present invention is an agent dialogue system.

10… Robot dialogue system 12… Dialogue place R1, R2… Robot 17… Intention display device 18… Sensor manager 20… Group manager 22… Dialog database 23… Human intention processing module 24… Robot controller

Claims (3)

It is an agent dialogue system in which two agents, each having a speech function, interact according to a script at the dialogue place, and a person sits at the dialogue place.
A sensor manager that recognizes sensor signals such as the person's actions and state changes at the dialogue place as the intention expressed by the person regardless of the person's actual intention.
The first judgment means, which is certified by the sensor manager and determines whether or not there is an intention to approve,
When the first determination means determines that there is an intention to approve, a non-verbal operation is executed to approve the intention of the person to the two agents according to the first approval dialog loaded from the script into the dialog queue. First approval dialog execution means,
A second judgment means for determining whether or not the phase is to confirm the intention of the person according to the script.
When the second judgment means determines that the phase is to confirm the intention of the person, the third judgment means for determining whether or not the intention of the person has been confirmed, and
When it is determined that the third determination means can confirm the intention of the person, the second agent causes the two agents to execute a dialogue to approve the intention of the person according to the second approval dialog loaded from the script into the dialog queue. 2 Equipped with approval dialog execution means
The second approval dialog execution means causes one of the two agents to make utterances that are synchronized or out of sync with the intention of the person, and the other one of the two agents is made to speak the person. An agent dialogue system that makes utterances that are out of sync with the intentions of . The agent dialogue system according to claim 1 , wherein the second approval dialog execution means causes the two agents to execute the approval utterance after expanding and interpreting the intention of the person . A program executed by a computer of an agent dialogue system in which two agents, each having a speaking function, interact according to a script at a dialogue place, and a person sits at the dialogue place.
A sensor manager that recognizes sensor signals such as the person's actions and state changes at the dialogue place as the intention expressed by the person regardless of the person's actual intention.
The first judgment means, which is certified by the sensor manager and determines whether or not there is an intention to approve,
When the first determination means determines that there is an intention to approve, a non-verbal operation is executed to approve the intention of the person to the two agents according to the first approval dialog loaded from the script into the dialog queue. First approval dialog execution means,
A second judgment means for determining whether or not the phase is to confirm the intention of the person according to the script.
When the second judgment means determines that the phase is to confirm the intention of the person, the third judgment means for determining whether or not the intention of the person has been confirmed, and
When it is determined that the third determination means can confirm the intention of the person, the second agent causes the two agents to execute a dialogue to approve the intention of the person according to the second approval dialog loaded from the script into the dialog queue. 2 Make it function as an approval dialog execution means
The second approval dialog execution means causes one of the two agents to make utterances that are synchronized or out of sync with the intention of the person, and the other one of the two agents is made to speak the person. A program of an agent dialogue system that makes utterances that are out of sync with the intentions of .

Images